1600185-02-85-02 - Dados clínicos, Papers

Documentação F006-02

Projeto: Data: Código:

OXYMAG VENTILADOR DE TRANSPORTE 20/10/2010 1600185 Elaborado por: Verificado por: Aprovado por: Folha:

Marcelo Onodera Toru 21/10/2010 Tatsuo 21/10/2010 1/6 Objeto / Título do Documento: Revisão:

85 – Dados clínicos, Papers 02

MAGNAMED TECNOLOGIA MÉDICA LTDA

1. Objective

The objective of this document is to demonstrate the equivalence of essential characteristics of the transport ventilator Oxymag with the devices or functions, which is the subject of published reports listed in the annex.

2. References

Directive MDD 93/42/EEC – Annex X MEDDEV 2.7.1 - Evaluation of clinical Data

3. Method

This document will provide a brief comparative between the Oxymag transport ventilator ventilation modes and the articles listed in the annex.

4. Comparatives

4.1. Volume Controlled Ventilation (VCV) mode Doc 1 -Volume-Controlled Ventilation: Article cited by Anesthesia UK: Draeger Medical. Shows the principles of the VCV ventilation mode which is similar to the Oxymag Ventilator as described in the instruction manual. This article describes a possible risk of lung injury due to high pressures that can be reduced by using pressure-limited ventilation. Anaesthesia UK is an educational site with training resources for anaesthetic professionals. It provides interactive practice questions, journal abstracts and reference articles for the Primary FRCA, Final FRCA, Irish FCARCSI, European Diploma of Anaesthesiology, American Board examinations, and currently receives over 30,000 page views daily. The site currently has over 22,000 registered clinicians. Doc 2 - VCV Mode Application : Article cited by Department of Anesthesiology and Pain Medicine and Pain Research Institute, Yonsei University College of Medicine, Korea. Shows the clinical application of the VCV Mode and its benefits. 4.2. Pressure Controlled Ventilation (PCV) mode Doc 3 - Pressure-Controlled Ventilation: Article cited by Anesthesia UK: Draeger Medical. Shows the principles of the PCV ventilation mode which is similar to the Oxymag Ventilator as described in the instruction manual. Volume-controlled and pressure-controlled ventilation are compared. Anaesthesia UK is an educational site with training resources for anaesthetic professionals. It provides interactive practice questions, journal abstracts and reference articles for the Primary







FRCA, Final FRCA, Irish FCARCSI, European Diploma of Anaesthesiology, American Board examinations, and currently receives over 30,000 page views daily. The site currently has over 22,000 registered clinicians. Doc 4 - Pressure-Control Ventilation: Shows the clinical application of the PCV Mode. Doc 5 - Comparativo e Aplicação Clínica do VCV e PCV: Show a comparative study and its clinical application of the VCV and PCV Ventilation Mode using the TAKAOKA´s Anesthesia Machine, model Nikkey. TAKAOKA is the largest manufacturer of anesthesia machine in Brazil. The principles of the VCV and the PCV modes described of this article is the same as described on the instruction manual of the Oxymag Ventilator. In this study, involving healthy children submitted to the general anesthesia using two modes of mechanical ventilation, any interference was observed in the cardio respiratory stability along the surgical period. 4.3. (Volume and Pressure) Synchronized Intermittent Mandatory Ventilation (V-SIMV and P-SIMV) mode Doc 6 - Effect_of_Synchronized_Intermittent_Mandatory Ventilation: Article that shows the clinical study of the SIMV Ventilation mode which is described on the instruction manual of Oxymag Ventilator. According to this article, SIMV, pressure support ventilation, assist control ventilation, have been widely used. These types of ventilation are reported to provide good patient–ventilator interaction and to give good results when weaning from mechanical ventilation. This article was published in Anesthesiology - American Society of Anesthesiologists. The American Society of Anesthesiologists is an educational, research and scientific association of physicians organized to raise and maintain the standards of the medical practice of anesthesiology and improve the care of the patient. 4.4. Pressure Support Ventilation (PSV) mode Doc 7 - Pressure Support: Article that shows the impact of the Pressure Support Ventilation on Anesthesia Practice. PSV Mode is implemented in the Oxymag Ventilator as described in the instruction manual. This article shows that the PSV mode is an invaluable addition to the practice of anesthesia. The use of PSV allows patients to breathe spontaneously while reducing the patient’s work of breathing. This can be a clinical benefit in both outpatient and same day surgical anesthesia.PSV provides a new and clinically useful ventilation strategy that was only common in the intensive care units and for the extremely ill pulmonary patient. With PSV in anesthesia, a larger patient population can be served. This article was written by Datex-Ohmeda that is the world’s leading supplier of anesthesia systems, equipment, and services and an emerging leader in critical care.







Doc 8 - Performance_Characteristics of Five_New_Anesthesia Ventilators: Shows a comparative study of five anesthesia machines in PSV Ventilation Mode. The characteristics described on table 1 of the study are similar to the PSV Mode described in the instruction manual of the Oxymag Ventilator. 4.5. Continuous Positive Airway Pressure (CPAP) mode Doc 9 – CPAP Principles: The Oxymag ventilator provides a ventilation mode called CPAP (Continuous Positive Airway Pressure). One of the applications of CPAP ventilation mode is to treat patients with OSAHS (Obstructive Apnoea-Hypopnoea Syndrome). In the first article written by Dr. R. Farré from Barcelona University references that at least 5% and 2% of the adult male and female population respectively are suffering from OSAHS. OSAHS is characterized by recurrent obstructions during sleep caused by an abnormal increase in the collapsibility of the upper airway, which is triggered by several factors, including anatomical alterations and obesity. The short-term symptoms described by OSAHS patients are related to alterations in normal ventilation (choking, gasping or dry mouth) and disruption of sleep architecture caused by recurrent arousals (excessive sleepiness, lack of attention and irritability). Patients with OSAHS have an increased risk of traffic accidents, probably as a result of somnolence. Moreover, the nocturnal events chronically experienced by OSAHS patients contribute to the development of long-term comorbidities, such as cardiovascular and cerebrovascular diseases and inflammatory, metabolic, cognitive and mood alterations. The article follows describing the principle of functioning of CPAP device based on a blower and an exhalation port (intended leak orifice) and practical issues regarding this system. Furthermore, the article presents the principle and practical issues of auto-adjusting CPAP devices. Doc 10 – CPAP Benefits: shows a survey from 204 patients that used a CPAP therapy. A questionnaire was sent to the patients, and they had to answer about use of CPAP, sleepiness and road traffic incidents before and after CPAP, changes in nocturnal and daytime function, problems with CPAP therapy, and weight change. This study documents experience and perceptions of CPAP in a large sample of unselected CPAP users with a wide range of illness severity. Although necessarily limited by its use of mainly self-reported and retrospective information, the study provides evidence of patient-perceived, CPAP-induced improvement across a wide range of function, including sleepiness, driving competence, cognitive function, work efficiency, well-being, and nocturnal symptoms. 4.6. Noninvasive Ventilation (NIV) Doc 11 - Noninvasive Ventilation for Critical Care: Oxymag ventilator provides a functionality called NIV (Noninvasive Ventilation) where the mechanical ventilation is done by mask instead of intubation. Being an article from CHEST, one of the most worthy references in critical care medicine, this article represents the best practice or reflects the “state of the art” related to this subject. The







article presents the recommendation of using NIV for different types of patient. This ventilatory assistance without an artificial airway, has emerged as an important ventilatory modality in critical care. This has been fueled by evidence demonstrating improved outcomes in patients with respiratory failure due to COPD exacerbations, acute cardiogenic pulmonary edema, or immunocompromised states, and when NIV is used to facilitate extubation in COPD patients with failed spontaneous breathing trials. NIV for acute respiratory failure: the strongest level of evidence, including multiple randomized controlled trials, supports the use of NIV to treat exacerbations of COPD as a first choice to treat acute respiratory failure. Similarly strong evidence supports the use of noninvasive positive pressure techniques to treat acute cardiogenic pulmonary edema. Another NIV application supported by multiple randomized trials is to facilitate extubation in COPD patients. Immunocompromised Patients: the use of NIV is also well supported for immunocompromised patients who are at high risk for infectious complications from endotracheal intubation, such as those with hematologic malignancies, AIDS, or following solid-organ or bone marrow transplants. In a randomized trial of patients with hypoxemic respiratory failure following solid-organ transplantation, NIV use decreased intubation rate and ICU mortality compared with conventional therapy with oxygen. Asthma: several uncontrolled series and one randomized trial support the use of NIV for acute asthma. Postoperative Respiratory Failure: either NIV or CPAP may be helpful in averting postoperative respiratory failure by preventing atelectasis and/or improving gas exchange as suggested by three randomized controlled trials in patients undergoing different surgical procedures. When used appropriately, NIV improves patient outcomes and the efficiency of care. Although it is still used in only a select minority of patients with acute respiratory failure, it has assumed an important role in the therapeutic armamentarium. With technical advances and new evidence on its proper application, this role is likely to expand. 4.7. Airway Pressure Release Ventilation (APRV) mode Doc 12 - APRV Theory and Practice: The Oxymag ventilator provides a ventilation mode called APRV when the adjustment of I:E ratio in inverted (i.e. inspiratory phase greater than expiratory phase). Airway Pressure Release Ventilation has been described as continuous positive airway pressure (CPAP) with regular, brief, intermittent releases in airway pressure. The release phase results in alveolar ventilation and removal of carbon dioxide (CO2). Airway pressure release ventilation, unlike CPAP, facilitates both oxygenation and CO2 clearance. APRV is consistent with lung protection strategies that strive to limit lung injury associated with mechanical ventilation. The article shows the principle of functioning of this ventilation mode presenting the curve Airway Pressure x Time of the conventional volume targeted ventilation and the curve of APRV. Besides the article presents a table establishing the comparison among other conventional ventilation modes.







The author presents the history of mechanical ventilation in a brief chapter specially showing studies how the mechanical ventilation has been contribute with acute respiratory distress syndrome (ARDS). In this chapter the author describes the goals of the ventilation and how the APRV mode can reach these goals. After this chapter the author present the terminology of the ventilation mode, in other words, how we can adjust this mode in the ventilator. Then the article describes the indication, advantages and disadvantages of this ventilation mode, as described below in a brief table: Advantages:

1. Lower Paw for a given tidal volume compared with volume-targeted modes, e.g., AC, SIMV 2. Lower minute ventilation, i.e., less dead space ventilation 3. Limited adverse effects on cardio-circulatory function 4. Spontaneous breathing possible throughout entire ventilatory cycle 5. Decreased sedation use 6. Near elimination of neuromuscular blockade use

Potential Disadvantages:

1. Volumes change with alteration in lung compliance and resistance 2. Process of integrating new technology 3. Limited access to technology capable of delivering APRV 4. Limited research and clinical experience

The article describes an application of APRV in a case of acute lung injury and finishes it describing how the ventilation mode can be adjusted in weaning and the conclusions. According to his conclusion, the presence of APRV in the Oxymag will contribute to the increase of clinical practice of this mode and the comparison with other modes. The author shows that there is a lack of weaning consensus, in other words, the use of Oxymag in the weaning can shoot researches. 4.8. Pressure Limited Ventilation (PLV) mode Doc 13 – PLV: The pressure limited ventilation (PLV) is a ventilation mode that will be provided by Oxymag for neonates. This article shows the cardiopulmonary effects of Pressure Limited Ventilation (PLV) during Acute Lung Injury (ALI), evaluating the gas exchange and hemodynamic effects. This evaluation consisted on measurements of several parameters (right atrial, pulmonary artery, left atrial, arterial, lateral pleural and pericardial pressures, Paw, ventricular stroke volume, mean expired CO2, and arterial and mixed venous oxygen contents, airway resistance and static lung compliance) in seven male mongrel dogs in laboratory. The protocol consisted of observing the effects of various types of positive-pressure ventilation during control and ALI conditions. Hemodynamic variables were averaged over the entire ventilatory cycle and a minimum of three breaths was used to derive mean pressures and stroke volume. Muscle paralysis was induced at the beginning of each condition to abolish spontaneous movement.







Results: After ALI, static lung compliance, PaO2, and pH decreased, whereas airway resistance and PaO2 increased. For a constant lung volume, pericardial and pleural pressures were not different between control and ALI. Both absolute dead space and intrapulmonary shunt fraction increased after ALI. Ventilation did not alter hemodynamics during ALI. Conclusions: Changes in lung volume determine pericardial and pleural pressure. PLV strategies do not alter hemodynamics but result in less of an increase in Vd/Vt (dead space / tidal volume) than would be predicted from the obligatory decrease in tidal volume. 5. Conclusion The technical literature used in this report shows that the ventilation modes of the Oxymag Transport Ventilator are already validated and widely used worldwide. All ventilation modes of the equipment were evaluated using the literature and demonstrated their equivalences. There were no side effects identified and based on the Technical File of the Oxymag Ventilator, specially the Risk Management File and the Safety and Performance Tests register, we can consider that all risks are acceptable when weighed against the intended benefits of the device.




Marcelo Onodera Toru Tatsuo 1/1 Objeto / Título do Documento: Revisão:



Doc 1 –

Volume-Controlled Ventilation

Search our site

Go

Advanced Search

Home | News | Exam dates | Contact us | About us | Testimonials | Username: Password: Login Forgot?

map

You are in Home >> Resources >> Physics and equipment >> Ventilation

Volume-controlled ventilation Created: 18/2/2005

Email this page

Add this article to my examination home page

Print friendly page

Volume-controlled mechanical ventilation is delivered with a constant inspiratory flow, resulting in increasing airway pressure through inspiration. To maintain this fixed rate of gas flow the pressure must rise through inspiration. The actual preset tidal volume remains constant as lung compliance and resistance change. The inspiratory flow rate alters the velocity with which gas flow is delivered (Figure 1). Ventilation with a high inspiratory flow delivers the pre-selected tidal volume more quickly. If the ventilator is time cycled between inspiration and expiration and the tidal volume has been delivered before all the time allowed for inspiration has elapsed, an inspiratory pause occurs and the pressure drops below the peak inspiratory pressure. There is no fresh gas flow during this inspiratory pause. High inspiratory flow during volume-controlled ventilation has detrimental effects on lung ventilation. Therefore, low inspiratory flow rates should be used to keep the peak ventilatory pressure as low as possible. This ensures more homogeneous ventilation.

Figure 1

The risk of lung injury can be reduced by using pressure-limited ventilation (Figure 2). In older ventilators, pressure limitation stops the inspiratory flow, resulting in a reduction in target tidal volume. In more modern ventilators, once the pressure limit is reached, the flow decelerates to maintain the peak pressure at the pressure limit for the rest of the breath. This ensures the tidal volume delivered is as close to the target tidal volume as possible for the set pressure limit. A pressure limit of 30–35 cm H2O is appropriate in adults.

Insert Figure 2

© 2003 The Medicine Publishing Company Ltd

ArticleDate:20050218 SiteSection: Article

This section is sponsored using an educational grant from Draeger Medical

Login Status

You are not currently logged in.

UK/Ireland Registration

Overseas Registration

Forgot your password?

Sponsorship

This section is sponsored using an educational grant

from Draeger Medical

Exam home

UK register

Non-UK register

Abstracts

Image library

Clinical

Exams

Resources

Intensive Care

Courses

Books

Handheld

How to..

Podcasts

Software

Forum

Poll results

Links

Refer a friend

Site Map

Página 1 de 2Anaesthesia UK : Volume-controlled ventilation

17/07/2009http://www.frca.co.uk/article.aspx?articleid=100420

vp

All rights reserved © 2009. Designed by AnaesthesiaUK.

We subscribe to the HON Code principles. {Site map} {Site disclaimer} {Privacy Policy} {Terms and conditions}

Página 2 de 2Anaesthesia UK : Volume-controlled ventilation








Doc 2 –

VCV Mode Application







Doc 3 –

Pressure-Controlled Ventilation

Search our site

Go

Advanced Search

Home | News | Exam dates | Contact us | About us | Testimonials | Username: Password: Login Forgot?

map

You are in Home >> Resources >> Physics and equipment >> Ventilation

Pressure controlled ventilation Created: 19/2/2005

Email this page

Add this article to my examination home page

Print friendly page

Pressure-controlled mechanical ventilation rapidly achieves a fixed pressure throughout the breath by delivering a decelerating inspiratory flow pattern (Figure 3). The result is a tidal volume that varies with lung compliance and resistance. For example, if there is an increase in airway resistance, or reduction in lung compliance, the delivered tidal volume decreases and hypoventilation results. Pressure-controlled ventilation is usually closely monitored with alarms set for a minimal acceptable tidal and/or minute volume. Volume-controlled and pressure-controlled ventilation are compared in Figure 4.

Figure 3

Figure 4

Interaction between ventilated breaths and the patient’s inspiratory efforts –

The common modes are as follows.

Controlled mandatory ventilation (CMV) with no allowance for spontaneous breathing is the most common mode used in the operating theatre during routine anaesthesia.

In synchronized intermittent mandatory ventilation (SIMV) controlled breaths (volume- or pressure-controlled) are delivered to a preset respiratory rate separate from the spontaneous breaths.

Login Status

You are not currently logged in.

UK/Ireland Registration

Overseas Registration

Forgot your password?

Sponsorship

This section is sponsored using an educational grant

from Draeger Medical

Exam home

UK register

Non-UK register

Abstracts

Image library

Clinical

Exams

Resources

Intensive Care

Courses

Books

Handheld

How to..

Podcasts

Software

Forum

Poll results

Links

Refer a friend

Site Map

Página 1 de 2Anaesthesia UK : Pressure controlled ventilation


In assist-controlled ventilation, triggered spontaneous breaths are assisted identically to the controlled breaths.

In pressure-support ventilation, spontaneous patient breaths trigger a set amount of pressure to assist the breath.

Biphasic positive airway pressure (BIPAP) is a mixture of spontaneous breathing and time-driven, pressure-controlled ventilation (Figure 5). This system alternates between two adjustable pressure levels of continuous positive airway pressure (CPAP). Spontaneous breaths are possible at both pressure levels at all times. Cycling between the two levels produces gas flow and a resulting mechanical breath.

Figure 5

© 2003 The Medicine Publishing Company Ltd

ArticleDate:20050219 SiteSection: Article

This section is sponsored using an educational grant from Draeger Medical

vp

All rights reserved © 2009. Designed by AnaesthesiaUK.

We subscribe to the HON Code principles. {Site map} {Site disclaimer} {Privacy Policy} {Terms and conditions}

Página 2 de 2Anaesthesia UK : Pressure controlled ventilation








Doc 4 –

Pressure-Control Ventilation

22 AARC Tımes February 2000

LifeforVentilation

he application ofmechanical ventilation is apractice in which the respira-tory clinician may face widelyvarying conditions in a patient’spulmonary compliance.Decreased compliance in thepulmonary system can lead tosignificant problems inproviding mechanicalventilatory support. Theclinician must try to obtain abalance in obtaining acceptableventilation and oxygenationparameters and theconsequences of the neededsupport on the patient’s lungparenchyma. To use a famousquote credited to Hippocrates:“As to diseases, make a habit oftwo things — to help, or at leastdo no harm.”

Pressure-control ventilationis most often prescribed forpatients with severe adultrespiratory distress syndrome(ARDS). ARDS is initiallycharacterized by the formationof noncardiac pulmonary edemaand, in later stages, by theformation of hyaline membraneand pulmonary fibrosis. Theseconditions lead to reducedpulmonary compliance,

progressive atelectasis, andimpaired gas exchange,especially oxygenation.1 Sincesevere physiologic shunting isthe cause of this hypoxemia, itdoes not respond well to oxygentherapy. This leads to a classicclinical hallmark of ARDScalled refractory hypoxemia.2

This type of patient requireshigh peak airway inspiratorypressures to deliver preset tidalvolumes in the traditionalvolume-control modes of venti-lation. A significant problemwith ARDS is that the lung isnot uniformly affected. While amajority of the lung fields mayhave low compliance due toatelectasis, hyaline membraneformation, surfactant deficiency,and pulmonary fibrosis, otherlung fields may have normal

compliance. By utilizingconventional volumeventilation, the majority of thedelivered tidal volume is routedto these normal compliance lungfields. This can lead to over-distension of these areas andfurther insult to the lungparenchyma.

Pressure-control ventilationis an alternative mode ofcontrolled ventilation. When apatient is placed on pressure-control ventilation, the cliniciansets the rate, inspiratory time,positive end-expiratory pressure(PEEP), and, most importantly,the peak airway pressure limit.The ventilator acts as a constantpressure generator, limited to apreset value set by the clinician.Once this pressure is reached, itis held at that level till a limiting

S t e v e n S i t t i g , r r t

T

Pressure-Control Ventilation: “Primum non nocere” — Options in Limiting Pulmonary Barotrauma

RTs must always be aware of the

potential harm that may be caused

while providing mechanicalventilation in any situation.

Ve n t i l a t i o n f o r L i f e

24 AARC Tımes February 2000

factor (such as the end of theinspiratory cycle) occurs. A majoradvantage in utilizing pressure-control mode is that the patientcan receive as much inspiratoryflow as needed. These inspiratoryflows can be as high as 120 to 200L/min. dependent on the flow-limit capabilities of theventilator.

By limiting the deliveredpeak airway pressure, theclinician helps limit thepulmonary barotraumadelivered to the lung. The peakalveolar pressure, which is amajor factor in ventilator-induced injury, can climb nohigher than the preset pressure.

Some researchers havesuggested that, due to therelationship of lung injury tohyperinflation, the termbarotrauma should be replacedby the term volutrauma. These

researchers feel this volutraumais not caused by the peak airwaypressure or the PEEP level butthe difference found across thealveolus known as the trans-alveolar pressure. In a normalsubject, a transalveolar of 30–35cm H2O achieves the alveolarsize associated with total lungcapacity. Therefore, it is notsurprising to associate traumato alveolar membranes withrepeated exposure to tidaltransalveolar pressures greaterthan 35 cm H2O. Thismechanical ventilation-inducedtrauma produces alveolarmembranes that are susceptibleto becoming permeable towater and protein.

In i t ia l sett ings

When the decision is madeto institute pressure-controlventilation, many importantclinical decisions must be madeto ensure the new ventilatorsettings are not doing moreharm than good. Whenpressure-control ventilation isinstituted, the clinician must setthe maximal delivered peakpressure, PEEP level, inspira-tory time, and rate.

Remember that in pressure-control ventilation, deliveredtidal volume, minute volume,and alveolar ventilation is aproduct of the set peakpressure, inspiratory time, andcompliance of the respiratorysystem. Determining thecorrect initial set peak pressurewhen converting to pressure-control ventilation fromvolume-control ventilation canbe a challenge; the clinicianmust try to use a set peakpressure that will deliverapproximately the same tidalvolume that was being delivered

during volume ventilation. Thispeak pressure level can becalculated by determining thedifference in pressure betweenthe end inspiratory plateaupressure and the set PEEP levelduring a volume-controlledbreath. As stated earlier, setpeak pressures greater than 35cm H2O should be avoided ifpossible to help decrease insultto the alveolar membranes.

Due to the heterogenousnature of the lung injury seen inARDS, a minimal level of PEEPneeds to be applied to preventtidal end expiratory collapse ofthe edematous airway tissue.When this tissue collapses,there is a decrease in respiratorysystem compliance; and distinctpoints of inflection are noted onthe static pressure volumecurve. Application of PEEPlevels above this lower point ofinflection on the pressurevolume curve known as the“Pflex” is thought to bebeneficial, but determining anaccurate value for the deflectionpoint has been reported asimprecise when estimating it offstatic pressure volume curves.3

Providing this minimal levelof PEEP support, it is felt,prevents the recurring processof alveolar tidal recruitmentand subsequent collapse. It isalso felt that this minimalPEEP level may help limitprogressive microatelectasisand inactivation of surfactant.4

A minimum level of 7–12 cmH2O has been suggested as agood starting point. If higherlevels of PEEP are imple-mented, it is recommendedthat a pulmonary arterycatheter be placed so thatcardiac function may beaccurately assessed.

Ve n t i l a t i o n f o r L i f e

AARC Tımes February 2000 25

The next parameter that isvital in determining deliveredtidal volume is the set inspiratorytime. A lot of the literature onpressure-control ventilationconcerns the use of prolongedinspiratory times and inverseinspiratory-to-expiratory-timeratios (I:E). This facet ofpressure-control ventilationhelps increase delivered meanairway pressure (Paw) and helpsimprove oxygenation.5 Thedownside to this is that it is veryuncomfortable for the patient.Most have to be sedated andparalyzed in order to maximizethe benefits.

The possible need forprolonged pharmacologicalparalysis can have severe long-term consequences, almost assevere as ARDS itself.

I had the privilege of hearing alecture by Marshall L. Post,RRT, about pressure-controlventilation at the AARCInternational RespiratoryCongress in New Orleans in1997. In his lecture he describedhow he and his staff usedgraphics to optimize the setinspiratory time. By using thisreal-time diagnostic tool, theywere able to maximize the setinspiratory time and deliveredtidal volume without developingauto-PEEP. The set respiratoryrate and inspiratory time weremanipulated to allow bothinspiratory and expiratory flowto reach baseline. It allowedthem to pressure ventilate thepatient without subjecting thepatient to the possible long-termharm that could have beencaused by prolonged paralysisand sedation.

Newer-generation ventilatorsnow offer a combination ofpressure-control and volume-

control modes of ventilation. Inthis combined pressure-controlvolume-regulated mode, thedelivered peak pressure islimited; but a set tidal volumeand minute volume can bedelivered. Then, even ifcompliance changes, set minuteventilation is guaranteed andpeak pressure is limited.

For those of you who have readthis article and have wonderedwhat the phrase “primum nonnocere” means, it is a tenetcredited to Hippocrates and usedin the Hippocratic oath. Itmeans, “First Do No Harm.” Weas respiratory clinicians mustalways be aware of the potentialharm that may be caused whileproviding mechanical ventilationin any situation. •

Steven Sittig is a pediatric respiratorytherapist at Mayo Clinic in Rochester,MN.

REFERENCES1. Luce, J.M. (1998). Acute lung injuryand the acute respiratory distresssyndrome. Critical Care Medicine,26(2), 369-376.2. Fulkerson, W.J., MacIntyre, N.,Stamler, J., & Crapo, J.D. (1996).Pathogenesis and treatment of theadult respiratory distress syndrome.Archives of Internal Medicine, 156(1),29-38.3. O’Keefe, G.E., Gentilello, L.M.,Erford, S., & Maier, R.V. (1998).Imprecision in lower “inflection point”estimation from static pressure-volume curves in patients at risk foracute respiratory distress syndrome.Journal of Trauma-Injury Infection &Critical Care, 44(6), 1064-1068.4. Verbrugge, S.J., Sorm, V., &Lachmann, B. (1997). Mechanisms ofacute respiratory distress syndrome:Role of surfactant changes andmechanical ventilation. Journal ofPhysiology and Pharmacology, 48(4),537-557.5. Gore, D.C. (1998). Hemodynamicand ventilatory effects associated withincreasing inverse inspiratory-expiratory ventilation. Journal of

Trauma-Injury Infection and CriticalCare, 45(2), 268-272.

ADDITIONAL READINGTobin, M.J. (Ed.). (1994). Principlesand practice of mechanical ventilation.New York: McGraw-Hill.







Doc 5 –

Comparativo e Aplicação Clínica do VCV e PCV

��

��

��

��

��

��

��

��

��

��

�� !"#�"$��%&'!$�())*

��

��

��

��

��

��

��

��

��

�� !��"��#�

��$�� %��&��'��(�%�)(��"

�%+�"&", %-.��/"& %�/'&&"� %-.��0�"1'#2,3'%�/��#�%4%�/" � �5&��%/6%-.� � "4 � �'2,3'%& � �0/'3%&� � �"/'% �'%$ � /%��,'7"�&'/%/" � �"/"�%! � /� � �'� � ��%,/" � /� � �6!$ � +%�%��8 ",-.��/�� 9 6!��/"��"& �":

�� !"#�"$��%&'!$�())*

�*+,';,<��,$��%8'%,%��-&��./��"-��"#��%��0��"�-��"��"��

&��1��'��#��%�./�'��%��#�%2-��#��%�./�'��%��&��""/�3(�4��5�6��7�� !��"��#�7'�$�� %��&��'��(�%�)(��" 8,99:

=>�?:��'!:�3�!��:�

'&&"� %-.��@4"& �%/�A�B��,'7"�&'/%/"��"/"�%!��'��%,/"�/��6!: ��%36!/%/"�/"��"/'3',%:��#�%4%�/"��5&��%/6%-.��"4��'2,3'%&��0/'3%&��"/'% �'%:�� !"#�"$��$�())*:

C: � �"&+'�%-.� � %� '?'3'%! � (: � ��'%,-% � D: � �"&+'�%-.� � 3�4�+�"&&.��+�&' '7%�E:��", '!%-.��/"�%! %�?�"FG2,3'%�H:�,"& "&'%�#"�%!��:��'7%$��"??"�&�,��"/��:�� '&$��!%',"�+%�"3'/%��"!'1��:��9 6!�:

� ��I��EE)�% %!�#%-.��'8!'� "3%��J��

��

��

� ��

�� !"

��

��

�� !��" #�� $��

��%�� &

�'�� #��(��)�� *�

+��,� ,�� &

�� -�& �.�� /�� 0�� 1��

1��2��&

�� &�'�

#��&�-��&�" ��#��3�4��

�� &�'�5��&06��"��3��)��#��#7��)��8)##)9��

�� :��&�

��+�3��3�� ;��,�� <�33�� /��

� �� =��

�� &

��" #��

��!�� 7��

��71��&

��-�&�>��#��#�,��-��&�: ��

. �1�.�� &�

:��,��1��

�2��& � 5�� ,��

��&� ��,��

��&�/��%��

"#�$��!

%�" ��&��#��"

%�" �� '�%�"�

%�" ��'��(�� #��"�

��"#�!

�'" ��

)�� !�#*+! ?

,��(�"+!��%� �� #�� @

@&?&�0�� /�6��"��7�� @

@&@&�0�� /�6��)�� A

@&B&�0�� /�6��+��C��,� D

@&E&�0�� /�6�� .�� F

@&G&��/��/�6��0�� 7��+��)�� ?G

@&A&�� :1�� 0 0��# 0 ?H

-��.#" �&�� (��!��" #�!� @I

/��!'.� �(!" @@

0��1�� "��2 !�!" @E

�G&?&�- ��:�� @E

�G&@&�J�� K� ��:�� @E

�G&B&�5 �� @E

�G&E��0��,��:�� @G

�G&G&�5�L=��#�� BI

�G&A&��2��0�� /�6�� BB

�G&D&�/��2�� M��"��6��

� �K��7��+��7�� BA

�G&H&�:��N�� BA

�G&F&�� :�� BH

�G&?I&��O�� BH

� 3��&��4��"�'�'%�!��$&��" EI

5�� ! EH

K�� EH

�� GI

C�� G@

/;�� GB

K� �� GH

-�� AG

K��=�� D@

6��!��%#"7�"� DH

� ��8!"

�<:P%�?��,� ��#�;�%��7��

�<:P%�@�� J�,��:��

�<:P%�B��,� ��+��7��

�<:P%�E��,� ��#7��%��7��

%�" ��&��#��"

��

&�9��)��C��0�� Q� ��Q�

�� & B

&�9��,��.�� 1��1��,�� ,� ��

0 0�8�9��,�� # 0�8�9& ??

&�9��-��5�� 1��8�� 9�� & B@

&�9��/�� <�33��N�+�3��3�®�85��#� �� 9& BE

%�" �� '�%�"

�:�� &�

GF

�:��0� ��;��,��,��6��

AI

��,�� 6��E��,��&�

�:��0� ��;��,��

A?

��,�� 6��E��,��&�

�:��(��0� ��;��,��,��6��7��

AB

��E��,��&�

�:��(��0� ��;��,��,��;��

AE

B��,��

%�" ��'��(�� #��"

5�� ,��

��!(� ��;��<��=��

�"� ��

'� �>�� :��?:��>��@

�� A� ��<� ��

��B� � ��C��=�� D�� ?��E)F,!@

�!, ��D>�� : � �?��F9@

�� C��;�� D�� ?��E)F,!@

��<��1( ��<�� <��< ��?��<�� C��@

��& ��

�(11 ��G9�� A� ��

&� &��HIC��E�:��E)

&�!, &��A� �� >�9C��

� &��HIC��D��?�� E)@

F��!- '��: �� D�� 9��?��H�%E)@

F�1� F ��1 �� 9��

��( (��A� ��D��

�( (�� <��

%1� %�� 9��

%1�( %�� J��<��A� ��=��

� � ��

��

!, !>�9C��

1 1��

1��!, #�� %@�� 8��"�9

1�� 1�� G��?��F9@

1�� 1�� <��<�

1�!, 1�� !,�� 9��?��F9@

1�!,�K&�!, ��A� �� !,�� 9��A� ��

>�9C�� L�� >�9��A�

1�( (��A� �� ?��

1��1 1�� <�� >��A� �?��F9@

1� �!, 1�� !,��>��D��

1�G�� 1�� G��<��G��?��F,!@

�F �F�� 9��

11� 1�� D��?��F,!@

11%� 1�� M�?��F,!@

��N ��C��D�� ?��F,!��E)�E)�@

��A� ��O� ��A� �� D�� O�� >��D��

"�!, "��A� �� 9� :�� 9��?P@

"�� "�� 9Q��D��9��

� ��

�� >��D��

�� : �� H��

�� 9��?R�@

�� D��

# �1 #�� <��1��;��

( ( ��?�%@

(S &��> �?%��E)@

(�K( ��A� ��<��A� �� A �� D9�� < ��

(�( (��A� �� < ��?�� @

(� (��A� ��=��

( ( �� ?�%@

��"#�!

.��<��!:T��< �O� ��M��6��7��

��$��;��

�� ,�� 6�� ,� ��

80 09�� 8# 09��

� �� & ��G� � O �:��

�� 2 ��@IIB��2��@IIG��,� ,��8��G��9�

��$��;��=��"�� #��

� ��&�<��0 0��,� ��?I� &3��? ��1��&�<��

# 0��7��?I�� &3��?��1��&�<��

��#::#�G��"@%�� C�:��?�@&�%��

��6��,�� 7��E��

��@��B��& �%�� <%0��R��

�� & �� O �BD��;� ��

��=�� ?H �� 0 0� �?F �� # 0& �%��,��

��,� ��1�� 0+��/?�

S??F�� /E��S??B�� 8�TI�IB9��GU��

0+�� &�%��!��,��M��8�2��L=��7��9��

�� !�� & �� ,��,��

��& �� U��O� �� ,��

,�� 7�� ,�� 7��

��!��&

1��<��E��<�O�0�� /�6��V�0�� ,� �V�0��

�� V��#��

�'" ��

!:T��<��O � �+�� * ��

��1��*��

�� 4�� ,�� ,� �� ,��

80 09 � ,�� ,�� 8# 09� � �� *� � ��

��*��,��4 �� & �� O� �K��

�� > *��@IIB��>��@IIG��,� ,�� 8��A��

��G�*��9��*�� "��

�� #�� &�C��0 0�� ,� ��?I� &3��?�4��1�&��C��

�� # 0�� 3 � ��* � �� ?I �� &3��?� � 4�� 1�& � C� � �� 4��

,�� ,��#::#��G��"@%�� C�:�?�@&� �+��

��4�� ,�� *��

��1��E��*�4��@��B��&��5��

+�� <%0�� R�� & ��O� � BD�

�� 4�� *��?H��0 0�

��?F��# 0&�K��1�� ,� �� *��

�� 4�� ,�� 0+� /? �S??F �� 4�� /E �4�� S??B �� 8�TI�IB9��

�� GU� �� 0+� �� *& � �+��

��,��8��2��*�

?D

��*9�� 4�� 4��&��K��,��

��& �� O� C��*��,� ,�� *��

��4�� ,�� 4��,��*��*��

��*�� *�� &

�V�B�N ��O��<�� B�� W�< �� <�� W��

<�� W��

?H

�� !�#*+!�

/� �M��,�� 6��

�� & �� ;�� ,�� 7�� ,� �� 80+9 ��

?I� &3��?��,�� 1��;�� ?&�%��

��7��;��M��

�� & � ��

,�� 7�� ,� ��&�<��

��1��1�� 7��

��L=��,��?�@&

�� ,�� 6�� ,��,��

�� 7�� ,��,� � �� ;��

�� & � �� ;��

,�� 7�� ,�� 7�� , � ��

,�� ;��1��&

��(�"+!��%� �� #��

(��A� ��=��O�F��D��

��,�� 6��80/9�;�� $��

�� ,�� & �� ,��

��7��,� ��=��+��8K��C�?D�D��K��

CC�E�BE9��&�

%��,�� 6�� -��3��5��4��

��?F@F��,��,��1��1�� M�&�:��

�� 1��

� � �� =��;��& �:��?FB?� �� W��

��2��-��3�� ,�� ;��,��

�� &�:��?FBB��,��B&

%�� ,��M��=��;��

�� ?FGI�?&�9��)@&�

�� ,�� ,��

�� ,��

��,��B�G&

�� &�9��)��C��0��

Q� ��Q�� &

<��:X��;��AI��

�� & �#��;�� ,�� M� �� ,��

�� ,�� M� � � ��& � <�� ;��

,�� ,� ��,� ,��B&

��7��,�� Y��:��X��

��?FAB�� >�� >��.&�W��*��

@?

�� =�� 7�� 7� � ��

��&��

�� !� �� =��,��

� � �� & � <�� ,�� ,�� 6��

,�� ,� ��L=��

��7�� ,��L��,�� M��&�%��

��&�-��;��AI�� HIU&��

��=�� ,�� ,��

��0/��&��71��;��1��

��,�� B�E& �)��*� �� ?FD?��

��,�� 1��8#::#9��,��;��5��

-��K��7��K�;��<��8/��"�� 9��

�� 7�� 1�� B& �:��?FD@��W��*��

� ��,��,�� 1��

�� *��&�:��;��

�� 6�� ,�� 7��

,�� 7�� 8C/09& � :�� ,�� ,�� ,��

��M��7�� L��&�:��

�� =�� 7�� ,� ��

�� 7 � �� M� � � � �� E�G& � %��

,�� 5��C0�?II��85��C�� 9��;��,��

��(�� 1��;��2��

�� B&�

@@

<��;��HI�� ,��

� ��,�� 7�� 1��7��

��,��,�� 7��,�� 8# 09��&�

�� ,�� ,� ��80 09�� ,�

�� 1�� 7�� 8��9� �� # 0�,��

��,�� 1��7��

,��,��;��D& �%�5�,��

FII � 85��: �� 5��9 � ; � � � �� ,��

�� 1��

��;��H&

:�� $��,�� ;��;��FI��,�� (��

��6�� 8,�� (�� Z��9 � �� ,� � � ��

��F&�� 1��

��,�� 6��7��&�:�� 6��

� �� ,� � � �� ,�� ,��

�� ,� � � �� H& � � � ��

��,;��,� ��1��

��,��,�� &�

"�2 � ��,�� ,�� 2� � ��

�� ,��,�&

/��,��,�� 6��

,�� =��,��

�� 7��A�)I&

@B

(��A� ��=��

�� ,��=��=��

��6�� &�:��

��,�� 2�,��,��

��& �%� � ��;�� ,�� 7�� 71�� 8 %@9 � �

�� ;�� ,�� ,��#� %@&� -�� 1��

�� 8 K.9� � �� ,��

�� 1��

%@ �8.�%@9& � -�,��

��,� ,�� L��,� ��

� ��& �� K.�� M� �� 6��

� ��,�� Y��D�?I&

:�� L=��

�� ; � � ��& �� !��

�� ,�� 6��

��D&��,�� 6��;��7��&�5�� 7��

�,��

?�@&� �� 7�� 7��

�� ,�� ! �� ; ��, ��,� ,� ��;��

,�� , ��=�� M��2�,��&�:��

�� M� � � � �� M� � �� ,�� 6�� 7��

��L��&�#��, ��,��

@E

��;��,�� 7��

,��,��7��7��G&

�(��A� ��=�� <�

��,�� 6�� X�� +�� C��,� � #�� 8X+C#9 � ��

� ��,�� ,�� 7��& �:1��

�� 7��;��,�� 7��,�&�<��

�� ,��,��=��

��B�D&

X��,�� =��,��,��

�� ,� ��?I3��

��,� �� 8[�?II�� 9��G&

� ;�� ,� ��Y��;��

��,��

��,� ��,�&� ��,��

��,�� ,��

,� �� ,�� H�??& �<�� ,��,�� 2� �;�

��&�

%�� ,��$�� ,��

8��9��&�<��7�� ,��

@G

�� ,�� ,�� ,� ��

��,� �� Y,� ��&

#�� ,�� 7��&�%��

� � �� 7�� M� �� ,��

�� M� ��,& � "�2� � �� ,��

�� ,�� 1��

��1��&�#��1�� ;��,��,�� 7��

�� 7�� , � �� ,�� ,� �� 80 09& �:��

�� 5��3�� ,� ��,��,��

��8# 09��, �?B&��7��# 0�

��,��2�� ,�� ,� ��

,� ��L��0/�� M��?B&�%��

��# 0��=��2��&�K ��

��,�� ,��,��,�� L=��

��;�� ,��,��,�� 1��2��

?B�?E&

�� ,�� 6��8J#C09�;��

�� ,�� 6��& �.�� 5��

��!�� K��7�� 85�K�9� � � � � �� $ � �;�� ,��

��,� ��&�%�� ,��

$�1��,�� ,� ��1��,��,� ��$��

�� 1��8#::#9�?E&�+�� 5�K��

��,� ��;��7�� ;��, ��

@A

��<&�� $�,�� 6��1��,�� ,� ��

�� , � ��

��7��1��(��6��@&�0� �� M��

�� M� � � �� 7��1� � � �� M� � ��

�� ,�� M�� 7�� ,� ��&�:��

�� 7�� ;��@�<��<��

�?E�?G&�<��J#C0��M��7��

�,� �� =�� @�?A�

?H&�X��&�:�� 1��

��,�� ;��

�� M�� 0+� ?I� &3��?� ��#::#&�:��

�� 7�� , � �� 1�� M��

�� 7�� ?&

(��A� ��=��O�� &��

:�� 7��,�� ,�� M��

��8��7��9��,�� M��8��1��7��9&��

��7�� 1��;��

��7��7��1��

��1��7�� =��,;��

@D

��,�� ?F�@@&�

� � � �� 7�� 8�C9 � � �� 1��7�� 8�:9 � ��

�� 1��7��1��7��&��%�� 1��7��;��

��,� �� ,;� ��C��;��1��

�� =��=�� =�� 7��@B�@E&��%��

��,�� 1��

�� ,�� 1��

�� ?A�@E�@G&�"�� # 0�� 1��;��

�� 1��;��V��0 0�� 1��;�

��?&�9��,@&��

��,�� ,� �� 1��@A&

@H

&�9��,� N�.�� 1��1��,��

,� ��0 0�8�9��,�� # 0�8�9&�

?�� 1��7��V�@�� 1��7��V�B�

� 1��1��7��V�E�� 1�� 1��1��7��,��&

�

F l u x o

Te m p o

1

2

3

4

I N S P

E X P

1 2 3 4

�

F l u x o

Te m p o

1

2

3

4

I N S P

E X P

1 2 3 4

�

;

@F

� �� 1�� ,��;��

� �M�� ;��8##C9��7��1��

��1��

�� ,;� ��@I&��##C�,��,��,��

��,�� =�� =��,��;��

�� 1�� 7��& �/��,� �� 7��

��=�� =�� ,�� 1��7��

�� & �:��M� �� 6�� 7�� ##C � ��

��?G��@I��"@%�� &�%�� 1��

BI��BG��"@%�@I�@D&�

<��M�� ,�� ,��;��

�� 1��;��& ��,��;��

�� 7��;�� (�8##J�+9�@H&

-��1�� 1��;�� ,;� ��;��

�� 1 � � � 1��& � C�� M� � � ,� �� $ � ��

��;��,� ��$��;��8��,�9��

��=��$�� 1��1��7��&��,��,��;��

1��;��#::#�@E&

%��2�,��,�� ,��

�� ,�� ,�� M�&�:��

�� ;�� 7��1&��1��

�� M�� =�� &�

��,��

BI

�;��,��;��8#�;��9&�%�� #�;��

�� 1��7��##C�� C�:��7��1��7��

��, �� 1��&�%�� 1�� C�:��#�;��

��L��M��,�� @F&

��9��A��9��9�� (�(��1�(��

��<�� <;<�� A� �� (�(��

< �� G��GE�>�� T�� HIC��<��D��A� ��O��!�

<�� 9��> �H��G�� A� ��"��>��

�� 9��A� ��D�� > �� G��

��D�� 9�� D�� G�� D��?��X��D��@��

� ��9��>��D��H�� <��

�� < ��K�� !�� ;��J�� H�� < ��

�� 9�� < �� 9�� !��T�� :��

� �� > ��D�� 1�� ( ��A� ��O��H��

�� <�� :�� > ��D��

��<��9�� < ��GE�>�� ?< �� @��A� ��

� ��HI��<�A� �� > ��D�� < �� <�J�� U�� :��

<��G��AU��:��>�� C��9�� :��A� �

��H�� > �� <G � �� <��11��

��;�� C�� =�� AU�� :��

��C�� C��,6�

� �� 1�(�� =�� GE�>�� G�� <��G��?11�@��

!�� =�� T�� A� ��O��!�� 1�(��

B?

<��9��H�� D�� T��

�� A� ��A� ��11��,Y�-Z�� > �

��;�� <�� 9��=��D��

��,6��!��> �G�� 9�� 9��<��G��

�� U��<�� [��C��<��G��

�� D�� 9��G�� > �G��;>�� J��

H�� T��9��\��H�� < ��

9�� > �� A� ��

�� ?�� @��]�� G��9�� > ��

9��<��?��> �� @��

��D��!�( ��;�<��;<�� ;��C�� 9��

� ��<��AU�� C��C��,Y��<��G��H�� :��

�C�� C��:�� H�� > ��G�� A��

�H��:�� GE�>�� H��<��G��

��C�� ]�� <��G��:��<��U��

�� A�� H��:�� GE�>�� > ��;��

��H�� <��G�� C��<�� :��> ��!�

��> �� U�� G�H�� H��:�� 11��

�� ?�� > �� 9��J�� @� ��G�H�� D��

�T�� D��H�� H��:��

��<� ��2��H�� D�� T��H��

�H��:�� 11�� <��G�� 1��<��H��

� �� >��D�� <�� :��<��<��> �8�� 8.��@9��

B@

�� <�� >�� <��1��1�

��GE�T��H�� <�� < �� D�� 2��E

��H�� > ��>��D�� A�� <�� J�� H�� >��A��

H�� HIC��9��A� ��1��1�� E1��1��

��U�� =��T;<��-Z�-)�

� �� J�A� ��=��(��D��

��7��2��\��2��

�;��2��,�� M��

��2��,��,�� 6��

��6��2��,�� M�]�B@&

-�� 7��

��,� ��M��,��M��=��

,�� 7��$��,�� M�&

-�� 7��,�� 6��

�� ,�� 9 �� V � �9 ��

��6�� ,�� 7�� B@& � �� ,� ��

�� ,��,� �� &

�� M��1��,��

� � �� ,�� %@� ��

��;�� @&��1��

BB

�� %@� 8#�%@� ��"�9 ��

�1�� ,� �� %@& � ��

,�� ,� �� (�� 8#� %@��

��"�9�� %@��,;�� %@�

1� �� 1��& � :�� ,� ��

��1�� %@��,� ��

��B@&�

:�� $��6��,�� 7��,� ��M��,� �� 1��

� �� ,�� L�� ;��

��Y�� ,�� 6��&� ��

�,� ��

�� M�� ;��,� �� &�/�� 1��

�,� �� 7��7��

�� =��=��2��,��1� ��

�� ,�� M� � ��,��

��!��B@&

�� =�� ;�� ,��,� ��

� � �� ,� � �� &��"@%�?& ��

�� =��7�� T�0+Y#� ��

#::#�??&

O�� 6��7��;��

� ;��,�� M��,�� 2��

��

BE

��,��2�� =��

��,��,�� B@&�

�� =��6�� 8 �*�9��,��,� ��

�� 1�� M� � ��

��7��8�� 1��9�� 1��&�� *��

�� *��T�0+Y##C�#::#�??&

<�� *��;�� ?I��@IU&�X�� *��

� ��7��,�� ,��Y��&�5��

�� 1�� *� ��, � ��

��=��,��& �#��1��

�*�� & �:�� ; ��71�� ,� �� ,��

�� ;��$��=�� 1��7��??&

O � �� ,� �� ;�� ,��

��,�� 1��

�� =��&�-��,�� 6��M��,�� 7��

�� =��,��&

%�� ,� �� ,��

,� ��& �� 0 0��

��7�� &�>��,� ��

��,�� # 0�� ,��,� ��1��

�� ,� ��,� ��&�

-�� M�� 7�� =��

��,��L�� M��

BG

�� 1��&�#��6��

,�� ;��

,�� 7��&�:��, ��,�� M��6��

��,�� &

�� :�:1�� 0 0��# 0

�

�� 0 0��# 0�� 1��

��7��& �%� � �� ,� �� 1� �� 6��

,�� 7��,�� 6�� ,��,��

##C��#�;��##J�+��#�%@��#� %@��BB&

%�� ;��AI��

�,� �� ,�� 1��

��,�� BE�BG&�� ;��FI��0 0��

# 0�� 1��

�� ,�� 1��BD&�

:�� BH&

%�� ,�� 6��

�� ,�� 7��&�+��

�� ,� ��#�%@��,�� # 0��0 0��

�� ,��

# 0�� 7�� ;��7��& �-��

BA

�� ,�� 1��

��1�� ,� ��BF&�

:�� $�� =��0 0��# 0��

�� # 0 � �� #�%@Y.�%@ �?A�EI& �%��

��,�� 7��E?�EE&

�� $� �� %@�� 7��# 0��

��,��#� %@� ��,��

�� ,� ��&�:��,��

��0 0��# 0�BF�E?&�

��: �� T��HI��H�� :��

�� =�� > ��D�� A� ��

<��D�� (�(��1�(�� <�� H��

.#" �&�� (��!��" #�!

:��,�� 6��

� ��$�,�� ,� �� &�%��

��,�� 7��

��

� �� 1��&

X��;��;�� ,��

��6�� L=�� =�� =�� ,� � � ��

��&

:��;��

��,� ,�� 1�� & � +��;� � ��

��=�� ; �� 7�� 1� � ��

��,��;�� &

5�� ,� ,��

��

BH

��!�� ,� ,�� ,�� 6��

��6�� &

%��;��=�� ,��&�C��

�� =�� ,��$��;��B�

�� 6��

� �� 7�� ,��

� ��6��&

��!'.� �(!"

!:T��< ��

��M�� 7�� ,�� 6��

�� ,� ��$��

��;�� &

!:T��< ��

?& ��,� �� M� � ��6��

,�� 6�� ,� ��$��;��

�� &

@& ��,� �� M� ��

,�� 6�� ,� ��$��;��

�� &

B&��,� ��M��7��,;��,��

��6�� 7�� ,��

��6�� ,� ��$��;��

�� &

1�� "��2 !�!"�

��

:�� &

% ��J�A� ��

%�� !��"�� #�� 8" #�9�� &

"��A� ��

��;��

.�� 8A��G��9��1��

��;��=�� !��"�� #��

2 ��@IIB��2��@IIG�� 5��C��CC&��

��;��1� ��

�& #�� 7�� ,� � ��

! �� B �� 7�� ,�� ;�� ! ��

��V

�& C��X��+��C��,��,�� 6��

1�� V

�& -�� =��V

�& � ��;�� V

& -�� V�

�& K��, �� &

(��;<��

-��,� ��;��;��

� C��8��9V

� 51��8�� Y��9V

� #��83� ��9V

� ��8��9V

� 5�� 8�@9��,;��7�� ^8��1�E9�[D_�Y8��[FI9V

� �"��?��A� �� <�� G��

�� :�:�� :�� " ��B� �

�� 9��E��"�@�<��H�� )W

EB

]�� )� ��5��*��

��?��,

��@��=��

��B��=��,�� ,��

��E��=��,��

��G��,��@E��

%��&��:�� ;��=��&�

%��7�� A& `�/��

/��#�� ,��/��EG&

� >2��8��9��,� ��! ��,� ��;��;��V

� � ��8�� 7��1��

� �� 19V

� -��1��8�,��, �M��

7��&&&9V

� X��! ��=��8��7��7��

��;��&&&9&

-��,� ��;��;��

� <��8��7��9V�

� #��83� ��9V

� C��=��8��Y��9V

� �� M��7��8��Y��V��,�� 6��9&

EE

-��,� ��;��;��,��;��

� �� V

� <��,(��V

� K� 1��V

� R�1��7��! ��A��8��;�� 6��9&

-��,� ��;��;��

� +�� &

-��7��

/��+��7��8��1��CCC9

� /�� "��6�� ,��,��& �.��

�� ,��,� � � �� 6��

+�3��3�®�85��#� �� 9��

� .��L=�� 8. 9 � � �� ,;� ��

��,��-CC�� V

� #�� ,��,��8#�<C9 �→�/��7��

�� 7 ��7 ��;�� ;�� ;��V

� 5�� 1��=�� 1�� 85�%@9� →� �,� ��

�� 1��

�� & �� ;��,� �� ;��

�� V��

EG

� #�� %@�� 1��8#+ %@9�→�/��

��1��(��8��9��,��

�� %@�� 8��9&�:��

�� ,� � �� 1��

�� ;��7��V�

� �� ;�� →� /��

��;�� 1� ��

�� 7��V

� +��8+��9�→�/��

��&

� /�� 0�� N ��

��

� 0� � � �� 1�� 80+9 �→� <�� ,�,��

,� �� ,� ��AI��,��

,� ��=��8��9V

� .��L=��K��7��8�K9V

� K ��7��1��7��8K�C�:9&

� /��/�6��0�� 7��

�� 0+��,��,� ��=��&� ��

��/�� %�� <�33��\+��

��M�� 1��1�� &�%��

EA

��,�� 1��,� �� 1��M��,��

�� =��=�� 2�� 1�]�EA&

� #��,��;��8##C9V

� #��;��,��;��8#�;��9V

� #�� (�8#� ��(9V

� #�� 1�� 8#::#9V

� �� =��6��8 �*�9V

� �� =��8 ��9 �→�#��,� ��

��1��7��&�:��

�� K�� ,��, ��

��,� ��

�� V

� K��=�� ,��;��8K�49&

� /�� ,��,��+��)�� =��

�� 7��=��" #�� ;�� &

� �"V

� #�%@V

� #� %@V

� 5�%@V

� " %BV

� :1��8:�9&

ED

"�HIC��1� �� ^

1� � � � � ��G�� 1�� "�:�� [ � � ��A� � �� 1G � ��

� �9C��

��7� � ��,�� O�� /;��

C��;��=��$��

��;�� ;��

�� &�

X�� 8��1��C9��,� ��;��;��

��&�%��, ��

�� ,��

8��1��CC9&�%��;�� ,�� I�G��&3��C�

@I�BI�� &�

#��5��C��CC��A�

�� G�� 1��& �%�� ,��

�� ,;��

?��,�� 0 0��@��,�� # 0�� &�

<�� !��6��

� ��6��+�3��3�®� �� 7��

��,��,��1�� 7��(��&

��,�� 6��,�� 7��

�� 1�� ,��

<�33� ��+�3��3�®& �:�� ; � �� ,�� 1��

EH

��8�� 9��$�2�� ?&�9��-@&�

8,�� 9

&�9��-�^�� 1��8�� 9�� &

� ;��5�%@� �#+ %@�� ,� ��

��,;� � � � ��=� � ��& � �� L��

��;�� 1��!��@E�� ?��

�� 7��=��" #��

�;�� &�-�� =��&�%��

�;�� ,��,��?G�@I��&

%��$�� !�� , �@I�

BI��7��;��;��&��7��

�� 7��,�� 1��=��?IIU��

��,��;��1��a�@@��@E&�.�� @μ�&W��?�C0�

�� B��&3��?�C0�� &�%��

�� 7�� YE�[�B�G&�

EF

.�� ,��,��ED�EF&��1��

�� ,��1�� & �%��

�� ,��,�� 7��

�� ?I � � @I � ��"@% � � � � � ,�� ,� � � 1� �P � ,� � � ��

�� 6�� ,�� 7�� 6��

<�33�®&�>�� 6� ��(��

��,��

��$�� &�%� ��

�� @I��"�&��7��1��

��

�� ,�� ; � 8� �� , � ��

�� 9�� 7��7��&�X�� ,��I�@GU��

�� ?�@II&III��?� &3��?�� B��&3��

?&�� &�<��

��, �� ,��;��

�� , �� &� ��!��

�� ,��1��

��7�� 1��

��&

�� ;�� 7��

�� 1��=�� .�%@��I�G�

�� 1�� ;��@ &��?&�� ,� ��8 �/9��

��?� �/��1��8��,��?�B�?�D� �/9&�

GI

��7� � �� !�� I�G ��&3��?� � ��

�� @IU � �� BI �� 6��

,�� 7��& � +�� ;��

��<��8I�IB��&W��?9��8I�I?G��&3��?9��1��&�

��7��1��

��,�� 6��&�8��1��CCC9

�T��(��A� ��=��

%�,�� <�33��N�+�3��3�®�85��#� �� 9��

�2�� ,��,��,�� 7��0 0��

# 0�?&�9��/@&

G?

&�9��/�E�� <�33��N�+�3��3�®�85��#� �� 9&

<��L=��7��8�K9��2��1��

��&� ��A��?��K�@G�BI�� &��?� V�?��@��K��@I�@E�

�� &��?V�@�B��K�?A�@@�� &��?V��B�G��K�?E�@I�� &��?&�

�� 7��1��7��8K�C�:9��?�@��7��

G@

��@GU��0 0& �+�� #::#�G ��"@%& ��

��1��1��BG��"@I&

+�� ,�� 0 0��0+��

?I� &3��?�&�<�� 0 0��1��

� 1��&�

<��# 0��2��,��

�;��7��,� ��?I� &3��?&�X��,��,� ��

��1��,��M��,;��2��

�� ,��,��& �<� �� # 0� � �� 1� � �� ;��

��7��;��1��,;��,�� 1��P��

� 1�� &

�� ,�� 7�� 6��

��;�� /?��

��!�� G�?I��7�� ,�� 7�� V�/@��GI�

��7��V�/B��

0�� /�6��8�� 9V�/E��

�� 0�� /�6��& � +�� ,��

�� ,�� 7�� 8.�9&

��T ��AU��F�� =��D�� D��

GB

!��T ��AU�� D�� :�� GE

��:�� >��9� �� A� ��"�!,��0P�

� �<�� :�� <��1�!,�� 3Z��F9��T��<�E��&�!,��)�Z�

�� :��<�� J��E�� :�� <� ��

<�� M��,�� 8#� %@�b�GI��"��#+ %@�b�

EG��"�9��,��K��@IU��@��@�� &�<��8#� %@�c�

BI��"��#+ %@�c�@G��"�9��K��@IU��;��

��,� ��1��&

%��,�� M��

��,��M��1� ��

��&

%�� M��6��

��;�� 1��

��7��7��1��,� =��,��&

�H�� ^��

+��,�� <�33��+�3��3�

®& ��W&+�3��3��;��

,�� 1��& � ��

�� ,�� ,�� & � ��

GE

�� ,��

�� ;�� 6�� ,�� & � % � ,��

�� ,;��7�� 1��

��$�2�� &�� 1��,��,��

�� ,��6� �� 1�� ,��

��&��%�� 1�� ,��

�� &�� 3��

��&�� ,��

��6�� <�33�®� �� 0�� /��*�

��(��&�%��7��0�� /��*��;��

��2�� \d]��

� 1�& �� !�� (��

,�� 6�� ,�� 7�� 0��& � � � �� ,��

�� $��

,� ��M��2��&��,��,��

��&

��

GG

%�� 5#55 ��

,�� ?@&I@ �� ,��,��

1��;��,��,��,;��

�� ,;��<%0��&��7��

1��,;��R��&�<��

��,��,��

@IU��,� ��1�� I�IG��HIU&�

#��BI��7��&�

�� 2��

%��2��$��,�� O��#��5�!��

"�� &�#��2�� )##)��a�I@EBE&

% � �� , � � � � �� ,��

��7��6��

J�,��:�� &�8��1��CC9

%��1��

,��;��,�� 8,� �� 9��

�� &�-��1��

�� ,��M��,��,��

��1� � �� =��& �� !�� 1��M� � �� L��

�� 8��=�9��,�� 2��,��,� ��

GA

%@ �1��7�� 5�%@� � � ��

��;��&

��&��4��"�'�'%�!��$&��"

?& �� .� �- �� 5� � �� W�2�3�4� �%� �/�� +� �� +�� &�

/�� 0�� J��.�� *��3��#��:1�� /��

��:��1��&�� *�@IIGV�?I@��BB?�BF&�

@& Z��"/-��e�� >/ ��5��./-��0��5�� +��"��K��Z��,�5��"��

�� &�:�� 0�� K �� *��3��5*��

#��4��<�� # ��*�.��&�� *�@IIIV�FB��?E?B�?D&

B& +��3��C��C�� / )&�"��7��0�� # ��6��&�C<�� ,� ��Z��

"��/K�8��9&�0�� 6�� & �@�� &�5��

#� ��V�@IIE&��&B��??&

E& /��*��K&�K��,�� ,�� &�#��K��K,�

@II?V�?HE�?FE&

G& /�� / ��-K&�K��,�� ,�� & � ��

/��?FFAV�?D�AIB�A?@&

A& #�,��>��)��# K��;��5&�0�� /�6��#��C<��#�,��>#��)��

# �8��9&�/��C��,��#��&�K��>��K,��V�@IIG&��EHD�GIH&��

D& ��:�� C��)��.��#�� K��5��>��)�� )�� &�#��,�

K�� +�� #�� 0�� 0� ��

�� 0�� K-5`&� ��@IIIV�??D�?AFI�FA&

H& -�� -�� #&�/�� 0�� &�C<��W��#��:��-��<�>�8��9&�

/�� 0�� &�@��&�5�&J��/��*V�?FF@&��BEG�BDB&

F& "��->��5��:&�#��&�C<��"��->��5��:�8��9&�

<�� #��,� ��&�@��&�J�� V?FHA&�&???�??H&

?I& +��)��5"� �+�� / 3��.� �)�� :� �+�� :� �/ �� #)"��

J�� & � �:�� K�� /��,� � ��

�� *�@IIB��FH��?E�@@&

??& :��> &�/��7��&�C<��:��> �8�9&5��0�� 7��

��&�@��&�K��>��K,��V�?FFH&�&@�?G&

GF

?@& ��KJ&�#�� #��<�� #��/�� 0�� &�

K��*� ��?FF?V�BA�GAF�FG&�

?B& 5��3��#��-��*�W/��/��WJ��J��J��/��"��5/&�#��

,�� 4��,��&�#�� /��@IIEVG��?BB�?BH&

?E& ��/�%�� 50��/��-/��/�� K��5��)##��.� ��) �� &�

:�� , � � � ,�� * � �� * � �� * � ��

�*��&�<�:�� >�/�&�?FFHV�BBH��BED�BGE&�

?G& -�*��-��5��)&�:,��*� ��4��K-5�

��f�C��,� ��/��@II@V�@BI�@BE&

?A& /��)�&�#��,� �� ,�� &�#��

��@IIGV?G�ABI�ABD&

?D& 5��*� ��5� � �� 5+� � 532��* �K<� � .�� , �� -�& � #��

0�� +�� 0�� X�� C�� J�� /�� & ��

�� @IIIVF?�??EG�GI&

?H& 5��*��5��-��JW��:��-J��)�,��><��0�� +J��/�W��:-�� &� ��

K��*�/��C��X��"��5��*&�� @IIEVFH�EF�GG&

AI

?F& +��>/&� ��*��*��&�

:�� &�<�:�� >�/�&�@IIIVBE@�BEH&�

@I& +��>/&��,�� ,�� &�<�:�� >�/��@II?V�BEE�?FHA�?FFA&

@?& +�� K))�� ,� �� KK&�"��7��,�� 6��& � ��

/��C��,��@IIIV�H��?�@D&

@@& ��>&�#��,�� &�5;��

/��C��,��@IIIV�H�AF�?@E&

@B& )��J��#��,� ��J��K��.�& �#��0� ��,�� *�

�*��*�� *&��K,�K��-��?FHDV�

?BA�DBI�DBA&�

@E& #�� :� � ��,� �� KK& � 0�� 6�� & � C<��

��,� �� &�8�&9&�0�� /�6��&�5��#� ��V@III&�?@G�?BB&

@G& �� K& � :�� ,� � �� 6�� ,�� 7��

,�� ,� ��(��&�

8-�� 8/�� 99& �.�� /��

X��,��:�� #� ��Q>! ��/��.� ��Q��@II@&�?II��&

A?

@A& ��:��d��)KK&� ��*�� ,��

��,��*�� &� ��?FFIV�FH�?EEG�?EEF&�

@D& - ��K��)&�#� ��1��&� ��%�� @IIBV�?A��

AF�DB&

@H& � � � >% &0�� 6�� 7��7��& � C<�CC � ��

0�� /�6��&�>�#�� @IIIV@A��?B�?A&

@F& ��,� �� Z�� /�� /.& � 0�� # �� /�6�� ,��

<�� & � C<� � KK� & � 8:�&9& �0�� /�6�� & �5��#� �� & �@III&�

0�� # ��/�6�� ,�� <�� &��&?ED�?HH&

BI& /��3�#:��W��3��>&�#�� ,�� K-5�� &�

��?FFDV�??@V??I@�??IA&

B?& /��>>�� 3�#5��+�4��>-&�-�� ,��

�� &�>�� #�*�� ?FHFVADV?IH?�?IF@&�

B@& ��,� �� K&� �� ,�� 6��&C��CC� ��

0�� /�6��&�>�#�� @IIIV�@A��?A�@I

A@

BB& .�� :�.& � C�� =�� ,�� 7�� 1��

� �� 7�� ,�� ,� � � ,��

,�� 8��9&�.�� /��X��,��.��

K��)��5 ��#�� @IIB&

BE& �� <& � �& � :�� ,��*�� * � 4�,�� 1��&�

�� *?FADV�@H�BFI�BFG&

BG& ��<&��&�:��,��*��*�4�,��

�� ,�� &��K,�K��-��?FAFV?II�G?H�G@G&�

BA& /g��>��)��>��:�� >J��#� ��K-&�#�� ,��

�� ,�� 4�� *�� 4&� �� /��?FFBV�@?�??EB�??EH&

BD& J��/K��);��:��J��.��J&�:�� 4��

�� C�: � �� ,�� ,� �� ,�� * � ��

1�� *�� 4�� * � �� *��&�

�� *?�FFEV�HI��FHB�FF?&�

BH& /��3��/��J��4��3�� /��5,��<��W&�0�� 4��

��,�� *�� 4��1�� *��*�� &�

�� *�?FFAVHE�HH@�HHF&�

AB

BF& +�� /�� :��W��"��5��3�/��#��W��3��W&� ��%��

0� �� Z��#�� 0�� -��%��J��& ��>�

��?FFDVDF8B9��BIA�B?I&

EI& -�,��>W��K-�� K5��#��3��-+&� ��,� ��

�� ,�� C�� 4�4�,��f�+��2�� 2�*��

C�� ?FFAVE?�HIH�H?E&�

E?& /�� )�� J� �+�� >� � J�� .� �K�� & � ��* � ��

�� ,�� 4��4��,�� *��

�*��&� ��?FFBV?IE�HD?�HDG&

E@& /g��>��)��>:��:�� >J��#� ��K-&�#�� ,��

�� ,�� 4�� *�� 4&� ��&� ��/��?FFBV@?��??EB�??EH&�

EB& K��5"��5��K��d��)��Z��>�� #��:&�K��

��,�� ,��,� �� ,�� ,��*�

�� &� �� /��?FFEV@@�@@�B@&�

EE& /��"��W��3�K/��K��K��Z� ��K��W�� Z#&� ��*��

,� �� ,�� ,��CY:�� 2�*�� &�

��>�K�� & ��/��?FFGV?G?�DB?�DBA&

AE

EG& #��/�&��,� ��;��;��& �C��/��8��9& �� #��

+;��&�#�� /;��V�@IIE&��&B@B�BE?&

EA& �W&+�3��3�&�/�� %��& 7��/�� @IEI?IIHE&��@II?

ED& � � � � � � �. ��/J��5��:�� 5-�� &�:�� +�� #��C��

X�� &�� @IIBV�FD�?A?@�A&

EH& � � � � � � �W��"�� *�-"��W��3�K)�� &� ��

�� *�� &�� *�?FFDV�

HA�A@D�B?&

EF&� ��3��/��/*��KK&�/��4�*��4��, ��&�

��%�� @II@��?G�B@F�BBD&

�� !�

��

.��<��!:T��< �O� ��M��6��7��

�� $ � �� ;��

�� ,�� 6�� ,� ��80 09��

��8# 09��

�� & ��G� � O �:�� 2 ��@IIB��2��@IIG�

�,� ,��8A��G��9��$��;��=��

"�� #�� &�<��0 0��,� ��?I� &3��?��

��1��&�<��# 0��7��?I�� &3��? ��1��&�<��

��#::#�G��"@%�� C�:� �?�@& �%��

��6��,�� 7��E��

��@��B��&�%��<%0��R��

�� & �� O �BD�� ;� �� =��

��?H��0 0��?F��# 0&�%��,��

,� ��1�� 0+� ��/?�S??F��

��/E��S??B�� 8�TI�IB9��GU��0+��

��&�%��!��,��M��8�2��L=��7��9��

��!�� &�� ,��,��&�

� ��U��O� �� ,�� ,�� 7��

�� ,�� 7�� !��

��&

1��<��E��<�O �0�� /�6��V � 0�� ,� �V � 0��

�� V��#��&�

AD

�4"��'�

!:T��<��O � �+�� * ��

��1��*��

�� 4��,�� ,� �� ,��

80 09 � ,�� ,�� 8# 09� � �� *� � ��

��*��,��4 �� & �� O� �K��

�� > *��@IIB��>��@IIG��,� ,�� 8��A��

��G�*��9��*�� "��

�� #�� &�C��0 0�� ,� ��?I� &3��?�4��1�&��C��

�� # 0�� 3 � ��* � �� ?I �� &3��?� � 4�� 1�& � C� � �� 4��

,�� ,��#::#��G��"@%�� C�:�?�@&� �+��

��4�� ,�� *��

��1��E��*�4��@��B��&��5��

+�� <%0�� & ��O� � BD�

�� 4�� *��?H��0 0�

��?F��# 0&�K��1�� ,� �� *��

��4��,�� 0� � ��/?�S??F�� 4�� /E�4��S??B�� 8�TI�IB9��

�� GU� �� 0+� �� *& � �+��

��,��8��2�� *��*9 �� 4��

�� 4��&� �K��,�� & �� O� C��

��*��,� ,�� *�� 4��

�� ,�� 4�� ,��*� ��*� ��

AH

��*�� *�� &�V�B�N �� ,�� *��V�

,� �� ,�� V�� ,�� V�#��&

�� A�

AF

��,�� 6��80/9�;�� $��

,��,�� ?&��,��

��7��,� ��=��+��

8K�� C�?D�D� �K�� CC�E�BE9& ��,� �� 7��

�� 6��

,�� 7��=��

� ��,�� 6��8J#C09 �?�@��;��;��FI��,��

�� ,��,�� ,� ��,��

�� 1��&�K��,��

�� 2��

��,��B�E&�

%��,��,�� 6��

�� ,� �� =�� =��& �� ,��

�� 8# 09�;�� ,�� 1��7��

�� ,��;��&�:��

�� ,�� ; �� 2� � �� ;�

��L=��7��&�-��,��

�� ,� �� ,� �� 1�� 7�� ,��

�� 7�� =��

�� =��7��G�H&

+�� 1� � ��

�� 1� � ��

��,��;��# 0��

DI

�� ,�� G�D�H&��#��

�� ;��

� �� =��2�� ,��

� � �� ,� � � �� L�� 1�� F& � :��

��,��,��

��,�� 8# 09��,�� ,� ��

80 09�?I&

<�� M��6��7��

��$��;��

�� ,�� 6�� ,� ��

�� &

�G� � ��

:��2 ��@IIB��2��@IIG��

�� 8��9��

�� 1�� $ � �� ; � �� =��

��!��"�� #�� 8�� 9� ��

�� 5�� C � � CC& �.��1� �� ,V�

�� ! �� =��V ��7��,��;��

! ��V��X��+��C��,��,��

��6�� 1�� V � � �� ;��

D?

�� V�� ,�� , �

� ��&�%��,�� =��

O��#��"�� #�� ,��

��,�� ,�� &�

%��

�� ,;� �� ?� ��

,�� 0 0��@��,�� # 0��

��&

%��;�� ,�� I�G��&3��?�@I�BI�

��&�<�� !��,��,��,��

��,��-CC�� ,��,��8#�<C9��

�� 7 �� 8#�59� ��7 �� 8#�-9� � �;�� 8#�/9� � ��

�� 1�� 85�%@9� ��

1��7��71��8#+ %@9��

8+��9 � ��,;� � �� 6�� +�3��3�®� 85�� #� �� 9& � � ;� � ��

�� 5�%@� �#+ %@� � �,� �� ,;� �� =��

�� &

%�,� ��8��7��1��7��9��6��

,�� 7��8�� =��6��7��

� ��(��;��,��;��=��7��,��

�� 1��9��1��,��

��<�33��8+�3��3�®9&�

�� 7�� ,�� 1��=�� ?IIU��

D@

��,��,� , �� +��3�&��

,��;��1��a�@@��@E&�.�� @�μ�&W��?�

C0�� h�B��&3��?�C0�� &�

%�� 7�� YE�[�B�G�

??�?A&��7��1��

�� ,��;�8��, ��

� � � � �� 9 � �� 7��7��& � %� � ��

��!��

��1�&�<�� ,��

��&

� � �� 7�� ?�

�� ,� ��/��8 �/9��1��8��,��

?�B�?�D � �/9� � �� 1��=��

��.�%@��I�G�� 1�� ;��@ &��?&��7��

�� !��I�G��&3��?�� @IU��

� � �� BI � �� 6�� ,�� 7��& � ��

��!��1�� 1��

�� ,�� <�� 8I�IB ��&W��?9 � �� 8I�I?G ��&3��?9& � ��7� � ��

1��

,�� 6��&

DB

�T��(��A� ��=��^�1� � � � ��

%�,�� <�33��N�+�3��3�®�85��#� �� 9��

�2�� ,��,��,�� 7��0 0��

# 0&

<��L=��7��8�K9��2��1��

��A��?��K�@G�BI�� &��?V�?��@��K��

@I�@E�� &��?V�@�B��K�?A�@@�� &��?V��B�G��K�?E�@I�� &�

��?&�X�� 7��1��7��8K�C�:9��?�@��

��7��@GU��7��0 0&�%��

��#::#�G��"@%&��1�� 1��BG�

��"@I&

+�� ,�� 0 0��0+��

?I� &3��?�&�<�� 0 0��1��

� 1��&�

<��# 0��2��,��

�;��7��,� ��?I� &3��?& �X��,��

,� ��1��,��M��,;��

�2��,��,��&� �<��# 0�� 1�� ;��

��7��;��1��,;��,�� 1��P��

�� 1�� &

�� ,�� 7�� 6�� ;��

��/?��!�� G�?I��7��

��,�� 7�� V�/@��GI��7��V �/B��7��

DE

��0/�8�� 9V�/E��

�� 0/& �+�� ,��

�� ,�� 7�� 8.�9&

��T ��AU��F�� =��D�� D��

!��T ��AU�� D�� :�� GE

��:�� >��9� �� A� ��"�!,��0P�

� �<�� :�� <��1�!,�� 3Z��F9��T��<�E��&�!,��)�Z�

�� :��<�� J��E�� :�� <� ��

<�� M��,�� 8#� %@�b�GI��"��#+ %@�b�

EG��"�9��,��K��@IU��@��@�� &�<��8#� %@�c�

BI��"��#+ %@�c�@G��"�9��K��@IU��;��

��,� ��1��&

%� � �� ,�� =��

��2��1� ��&

%�� M��6��

�� ;� � �� 1��

��7��7��1��,� =��,��&

��

DG

��Z��4��

�� !�"!��#��$��%��&��

��'��&��&��#(�)#�

��#��$��*��%��

��&��+��,-��(��.��

��.��.��!"/��

�%��"'"0��1"/��'�

��2"��*��

��

:��2 ��@IIB��2 ��@IIG�� BD��;�

��=��?H�� )��0 0��?F��)��# 0&��

�;�� @B �� ?@�

3� ��V��AIU��1�� &�<��,��

��;��;��8+�� C9&�

%��!��?GH �[� E@��

��0 0��?EG�[�A@��# 0�8�TI�B@9�&

+�� C��

,��, 0 0�T?H

# 0�T?F �

C��8��9 @@�@�[�?@�? @B�H[?B�D I�AFD51��/�� 8U9 ?G�8HB9 ?@�8AB9 I�@AF#��83�9 ??�F�[�B�B ?@�?�[B�H I�FID ��8��9 HB�D�[?I�@ HI�E�[??�@ I�BGD5�� 8�@9 I�G��[�I�? I�G�[I�? I�H??

DA

�5��C�8U9 H�8ED9 H�8E@9 I�GIG>2��8��9 H�E�[?�I H�@E�[?�B I�AAB� ��`�8U9 B�8?D9 B�8?D9 I�AGG-�� 1��8U9 B�8?E9 D�8?A9 I�@AI ��#�;,��8U9 E�8?E9 ?I�8?D9 I�?EF#��8U9 G�8@F9 B8?D9 I�EEB#��<�� @�F�[?�I� @�F�[?�I I�H@BR�1��K��7��8U9 ??�8A?9 F�8ED9 I�A??"��7��U B@�F�[@�H BB�E�[B�@ I�A?I"�� Y� ?I�D�[?�? ?I�H�[?�I I�DDI+��.�� 8U9 ?B�8DD9 ?I�8GB9 I�@GGK� ��;��±�-�,��#��J��0 0�8,�� ,� �9V# 08,�� 9�`��T?H��,

�

<��,��;��,��,��6��,��

��&�%�� M��6��. ��

#�-��#�/��+�� =��8+�� CC9&��

+�� CC�N�0� ��;��,��,��6��,�� 6��E��,��&�

0��, ��)�� /�� :��

�� /? ��/@� ��/B ��/E

.��L=�� 0 0 ?@E±?F ?@G±@? ?BE±@? ?BE±@? cI�I? I�@B��

8�� &��?9# 0 ?@B±?H ??F±@B ?@E±?D ?B?±@?

DD

#�� 0 0 AI�F±??�F AB�D±?E�A GD�F±?@�I AI�@±?@�F I�I@ I�AD�� /;��8��"�9

# 0 A?�B±H�F AI�F±H�E GA�@±D�B A?�G±?I�D

+�� 0 0 BA�B±I�A BA�G±I�A BA�F±?�I BA�F±I�F cI�I? I�HI�� # 0 BA�@±I�D BA�G±I�D BA�H±I�H BA�H±I�H

8i 9

K� ��;��±�-�,��#��0 0�8,�� ,� �9��T?HV# 08,�� 9��T?F/?��!�� G�?I��7�� ,�� 7�� V�/@��GI��7��V�/B�N��7��V�/E�N�� 0�� /�6��&

%��,��,;��,��

��,� ��1�� 0+��/?�S??F�� /E��

S??B�� 8�TI�IB9& � �:�� =��

��8�TI�G@9�8+�� CCC9&

+�� CCC�N�0� ��;��,�� ,�� 6��E��,��&�

0��, ��)�� /�� 0��0��

/? /@ /B /E

0+�8� 9 0 0 ??F�A±BE�? ??G�A±BA�B ??A�@±BG�H ??B�F±BA�A��I�IB��I�G@# 0 ?@?�@±BD�@ ??H�@±BE�F ??B�@±BF�@ ??B�E±BF�G

.��L=�� 0 0�� @I�D±B�@�� @I�F±@�H� @?�@±@�H @?�I±@�D��I�GE��I�D?K��7�� # 0 @I�F±@�A @I�D±?�F @?�I±@�I @I�F±?�F

%@ 0 0�� GB�I±E�? G@�?±?�D EF�B?±?@�B GI�B±?B�I��I�@D��I�GA��7�� # 0 GE�F±G�F G@�@±?�D� G@�@G±?�F E�I±H�?

5�%@ 0 0 FH�E±I�D FH�B±I�FF FH�E±I�F FH�E±I�H��I�@?��I�BG# 0 FH�A±I�A FH�E±?�I@ FH�D±I�H FH�A±I�D

DH

#+ %@ 0 0�� B@�G±E�F B@�A±A�?� BG�@±G�H B@�@±F�F��I�ID��I�DB# 0 BI�F±G�D B?�H±E�F BG�G±G�I B@�E±F�F

K� ��;��±�-�,��#��0 0�8,�� ,� �9�T?HV�# 08,�� 9�T?F/?��!�� G�?I��7�� ,�� 7�� V�/@��GI��7��V�/B�N��7��V�/E�N�� 0�� /�6��&/��+�3��3�� <�33�&�K ��C�:��1��?�@�0+�,� ��1��V �5�%@� �� 1�� V �#+ %@��#�� %@�1��7��

%��,� ��7�� 8##C9��;��8#�;��9�

��&��;��##C��

@E�@��"@%�� 0 0� � � � �@@�B � ��"@%�� # 0� 8�TI�?@G9& ��#�;��

��,� �� 0 0�?I�E��"@%��# 0�?I�E��"@%�8�T�

I�HF@9�8+�� C09&

�� =�� 6�� 8 �*�9 � � �� =�� 8 ��9�

��, ��&�<��0 0��,� ��

�� *��,��A�A�[�@�G�� &��"@%�?��/E��A�D�[�@�H�

� &��"@%�?V��# 0�� H�I�[�E�?�� &��"@%E?��

��D�A�[�B�H�� &��"@%E?�� &�:�� =��

��6�� 8�TI�BE9��

��8�TI�B9&��=��7��1��;��

�� 8+�� C09&

DF

+�� C0�N�0� ��;��,��,��6��7��E��,��&�

0��, ��)��/�� 0��0��

��

/? /@ /B /E

##C 0 0 @B�H±E�G @E�B±B�D @E�B±B�H @E�A±E�E I�F@ I�??# 0 @@�F±E�A @@�@±B�A @@�?±B�A @@�I±B�A

#�;�� 0 0 ?I�H±@�@ F�F±?�G�� ?I�B±?�I ?I�A±?�D I�AB I�BB# 0 ?I�A±?�H ?I�A±?�A ?I�A±?�D ?I�?±?�E

#� �� 0 0 @I�D±B�G @?�B±B�B @I�F±G�@ @?�E±G�B I�HA I�BH# 0 @I��D±E��D @I�B±B�E ?F�F±B�E ?F�F±B�B

#::# 0 0�� G�D±I�G�� G�H±I�E�� G�F�±I�@�� G�D±?�I�� I�EH I�G?# 0 G�H±I�E�� G�F±I�B�� G�H±I�B�� G�F±I�@

K�4 0 0 EB�D±?B�D EE�@±?G�@ EB�G±?@�G EF�?±@B�D I�EE I�@D# 0 EI�H±??�D BF�H±F�G BF�E±H�F BF�?±F�H

HI

�� 0 0 D�F±@�A�� D��H±B�G� D�D±B�@�� D�F±B�E�� I�?D I�BF# 0 F�B±E�F�� H�D±B�H�� H�A±E�I�� H�G±E�@

�*� 0 0 A�A±@�G� �A�A±@�A�� A�H±@�H�� A�D±@�H�� I�BE I�BI# 0 H�I±E�?�� D�GH±B�F� D�H±B�H�� D�A±B�H

K� ��;��±�-#&�0 0�8,�� ,� �9��T?HV�# 0�8,�� 9��T?F�/?��!�� G�?I��7�� ,�� 7�� V�/@��GI��7��V�/B�N��7��V�/E�N�� 0�� /�6��&/��+�3��3�� <�33�&�K ��C�:��1��?�@##C��7��8��"@%9V�#�;��;��,��;��8��"@%9V� #� �� (� 8��"@%9V� #::#�� ,�� 1��8��"@%9V�K�4��=��7��8��"@%YJY�9V �� =��7��8� &��?"@%9V� �*�� =��6��7��8� &��?"@%9

:�� $��#�%@ ��;��

)��0 0�#�%@�S�@BE��"��)��# 0��#�%@�S�@@?��"��8�TI�E@F9�8+�� 09&

��;��#� %@��0 0��BA�A��"�V��# 0��BD�G�

��"�� ,��,��8�TI�GH@9��

��8+�� 09&�

+�� 0�N�0� ��;��,��,��;��B��,��&/��

0��, � )�� :��/?��/@ ��/E ��

�" 0 0 D�BD±I�IA D�BA±I�IE D�BB±I�IE cI�II? I�HD# 0 D�BF±I�IA D�BA±I�IA D�BB±I�IA

#� %@ 0 0 BA�B±G�D BA�I±B�A BD�A±E�B I�?I I�F@# 0 BA�E±A�B BD�?±E�F BH�G±G�G

" %B 0 0 @I�D±@�E ?F�H±@�? ?F�B±?�H cI�II? I�FG# 0 @?�B±@�@ @I�E±@�@ ?F�D±@�?

H?

%@+ 0 0 @?�H±@�G @I�F±@�? @I�E±?�F cI�II? I�FG# 0 @@�G±@�B @?�G±@�@ @I�F±@�@

:� 0 0 �B�DH±@�AE �E�FB±@�BI �A�IB±@�IE I�I@ I�BF# 0 �@�FF±@�@? �E�EE±@�GH �E�@H±E�DB

#�%@ 0 0 @@E�G±AI�? @@A�H±GF�E @BF�?±EE�G I�DD I�HA# 0 @?D�G±AG�D @@?�A±GG�A @@?�B±EI�F

K� ��;��±�-#&�0 0�8,�� ,� �9�T?HV�# 08,�� 9�T?F��"��:�Y V�� %@��"�V�" %B��:�Y V� %@+��:�Y V�:&�&81��9�:�Y V�#%@��"�/?��!�� G�?I��7�� ,�� 7�� V�/@��GI��7��V�/E�N�� 0�� /�6��&

%��"�� 0 0��D�BD��:�Y � ��# 0��D�BF�

�:�Y ��/E��,��,��8�cI�I?9��"��

D�BB��:�Y ��&��,��

" %B��;��8+�� 09&

%��2�� ,�� 6��

�� & �<��0 0�� ,��

� ��2��,�� # 0��

8�TI�F?G9&��<��,��1��7��

��.�%@�

��

H@

1 �� C�� G9��<��D��

��;��!�� G�� G��

��A��9��<��H�� <��D�� H��

�� AU�� 9�� =��<��D��

�� O��@��A� �� < �� >�� 9 ��9��

�� <��A� ��J�� W�:@�!�� <��D�� ?1�(� ��

(�(@�� =��D�� 9�� =��

��H�� 9 �� W��@�!�

�Q�� <��AU��?�T��HIC��D��@�� 9 ��

��Q�9�� 9��

!�� <��D�� (�(� ��1�(��

��<�� < �� ;>��

��D��<��G��0�� :�� H��

�� 9��< ��H�� :�G�� < ��

� � �� A� ��O�� H�� E ��G��

��> ��9��

� �� A� �� ( �� 9 ��9��

9�� (�(�� H�� <��E��> ��

��9��<��A� �� ( � ��X�0P�� )��_9E)�

#��D��A� �� < �� <�� 9��

<�J�� H�� : �� H�� :��

<�J��J�� <��> �8�< ��

��J��: �� :�� G� ��6E)Z�� 9��

HB

<��9��C�� )-�)/��#�� <�J��

��9;��G�� A� �� : ��:�� HI��

��A� �� A� ��:��Q�9�� < ��

�� ))�),�� H�� : �� :�� E��H��

��11��`�,Z��F,!��>��9��<�J�� H�� :��

�� T��>��<�� H�� 11�� 9�� <��

��,,E,/�3��F,!��G�� H�� <�J�� : ��

�� A�� < ��>�� H�� )3�

��>��C��<�J�� D�� ?� ��>U��@��:G��

��H�� >��

��J�� AU��

��<��A� �� <��<�J��

!��H�� G��A� �[��

�� A� �� :�E��H��

� ��<�� :�� <��:�� )5�

�� >�� 1�!,�� 9 ��9��

� �� <��D�� <��A� �?1� �!,��1��!,@��:G��

��<�� H��<��AU��T��

��1� �!,�� <��:��

�� A� ��;�� :��<��A� ��

��9�� A��

�� < ��<�� A�� 1��1�?0��F,!@�� :��< �E

HE

��H��HIC�� H�� 9�� :��

�� <� ��: �� 9�� Y��

�� =�� >�9��A� ��

� ��C�� "�:�E��H�� G��J��

� ��C�� :G�� >�9��A� �� 9��Y�)6E,)��

� �� :�� <��A� �� 9��;��

�� A� ��1�!,��: ��

�� C�� C��

��D��

'��9��G��3Z�� (�(��1�(��

��A�� > ��:��A� �� 9��9I��

�� ;<��9�� :��

��A� �� 5�6�� :�� '��9��

�� H�� A��;<��

��:��[��9�� G�� <��D��

��H��<��A� �� C��<��D��5�6�

�� (�(��1�(� ��J��

�� 9��,,E,/�� H��<�� )Z��

� ��C��D��9��<�� D��

<��D�� (�(��1�(� �� 9��<��

1�!,�� !,�� !,�� D��

��9�� H��1�(�� :��G�� C��

��D��9��<��,,��!�� < ��<�� /Z��

HG

9��<�� (��1�(� ��J�� G9��

� � �H�� <�� 9��?(�(@��

��:�� <�� 9��9I�� H��

� �� <��D�� A��;�� E

�� A� ��C��,0�� :�� A��<�J��

��G�� A�� 9�� 9��<��

�� <��A� � ��A��

�� <��

��A� �[��AU�� F��'� � ��9��

:��<�� H�� 9��

� �� D��:��A� �� 9�� ;�� 1� <�<��

��:D�� :�� H�� :� �

�� H�C�� ,3��A��9��<��

��9�� <��

/�� GIU�� 2��

,�� 6�� 7�� @�B��

�� & �X�� ,� � � � �� M� � ��

,�� , ��

�� @D& �:�� 2��

��L=�� 7�� ,��

��;��,�� 7�� ,��#+ %@ ��#� %@+&�.��

��2��,��1��&�

HA

X�� ;��

�� @H& � :�� 1�� ,��

��;�� M��

��0+� ?I� &3��?� ��#::#&�:�� 7��,��

��0+� ?I� &3��?� � ��,�� 1�� M��

�� 7�� & � >� �� ;�� ,�� 7�� ,�

� ��M� �@F& � :�� ;��

,�� 7�� M��;� ��,��

�� M��&�

�� , �$��1��

�� ,�� 7��& � +��;� � ��

�� 1�� ,��

�� @F�BB&�-��

�� ; � �� ,�& � %� � ��

�� M��6�� L��

��!��

?F�BG�BA&�<��,��

��

��M� � �� 1��& � +��;� � ��

��

�� ,��&

X��$��;��;��

�� 1�� ,��

HD

�� ;��&�<��

�� ;��

��&�X�� $�

�� 1��A��G��&�:��,��

�� DIU��,��?@��BA��&

/��,��=�� ,��

��6��&��;�� ,��

�� & �<�� ,�� ,��

�� ,� ��,��

��6��7��,;��2��1��

,�� ,��

��&

��<��A� ��=��;��

<�� D9�� Q�9�� 9��

�� E��<�J�� <��:�� D��J�� H��9��

�� H�� <��A� �� H��

��J�� H��<��# �1��F T�� A� ��

�� J�A� �� :��T�� A��

��=��D�� <��<�� < ��

��U��J��<��AU�� 9��

��&��4��"�'�'%�!��$&��"

?&�Z��"/-��e�� >/ ��5��./-��0��5�� +��"��K��Z��,�5��"��

� &�:�� 0�� K �� *��3��5*�� #��

4��<�� # ��*�.��&�� *��@IIIVFB�?E?B�?D&�

@& �+��3�� C� � C�� / )&�"��7�� 0�� # �� 6��& �C<� � ��,� �� Z��

"��/K�8��9& �0�� 6�� &�@�� &�5��

#� ��V�@IIE&��&B��??&

B&�.��:�.��.��<�/&�%�� &�C��/��>�� 8��9&��

#��+;��&�@��&�#�� /;��V�?FFD&��?IG�?E?&

E&�.��:�.��.��<�/&�0�� &�C��/��>�� 8��9&��

#��+;��&�@��&�#�� /;��V�?FFD&��?E@�?AD&

G& �#��0��"� ��/��J�� /��%5&�0� �� ,��

�� ,�� 2�*��

��*��1*��&�>�#�� @II@V@H�?G�@@&�

A&�#�,��>��)��# K��;��5&�0�� /�6��#��C<��#�,��>#��)��

# �8��9&�/��C��,��#��&�K��>��K,��V�@IIG&��EHD�GIH&��

D& � �� <�& �:�� ,��*�� * � 4�,�� 1��&�

�� *�?FAHV@H�BFI�BFG&�

H&��<�&�:��,��*��*�4�,��

�� ,�� &��K,�K��-��?FAFV?II�G?H�G@G&�

F&�+��)��5"��+��/ 3��.��)��:��+��:��/ ��#)"��

J��&��:��K��/��,��

�� *�@IIB��FH�?E�@@&

?I& �#�� K & �#��C��,� ��& � C��)��*�)��8�9�#��&�

E��&�5��.�� J�,��V�@II@&�HB?�HFH&��

??&�W��"�� *�-"��W��3�K)&� ��

��*�� &�� *�?FFDVHA�A@D�B?&

FI

?@&��3��/��/*��KK&�/��4�*��4��, ��&�

��%��@II@��?G�B@F�BBD&

?B&�:�3�>��-��)J��#�� )��)��&�#��

�� ,�� &�#��@II@V?@�EFG�EFH&

?E&�Z��/��- �3��)*�� -� �� /��)�� &�5��

�� &��>��@IIEV�F@�DH�HE&

?G&�/��+��)��)��<��)��K&�K��

�� C�� 4��&� �� /��@IIIV@HV?GFG�

?GFH&

?A& � . �� /J� � 5�� :� � �� 5-& � :�� +� � �� #�� C��

X�� &�� @IIBV�FD�?A?@�A&

?D&��W&+�3��3�&�/�� %��& 7��/�� @IEI?IIHE&��@II?

?H&��1��""��"� *�Z��>� � �� &� C��1*��

��4�� ,�� &�<�:�� >�/��?FABV�@AF�FF?�A&

F?

?F&�<��"��"��>��<4�� >J&�:��

�� *� � �1*�� * � �� ,��

�� &��>�K�� /��?FFDV?GA�??HG�??HF&

@I&�)��J��.��:��"��*�� *�,�� &�

��%��@II@V?G�?A?�?AA&

@?&�+��Z��%��>��3�� &��4�*��

�� *��,� ��&�� 5��@II@V�

EA�G@F�GBA&

@@& � �� :� � d�� )KK& � ��* � ��

,�� ,��*�� &� ��?FFIVFH�?EEG�?EEF&

@B&�-�*��-��5��)&�:,��*� ��4��

�K-5��f�C��,� ��/��@II@V�@@@H�@BI�@BE&

@E& ��) ��"��<� ��5��J �� 5��/�&�� %��0� ��

�� #��K� �� 0� � � �� 0�� C� � �� K��* � .�� &�

�� ?FFDV@�DG�DD&�

F@

� @G&� 5��3 ��#� �-��* �W/� �/�� WJ� �J��J�� /� �"�� 5/& �#��

�� ,�� 4��,��&�#�� /��

@IIEVG�?BB�?BH&

@A& �)�� /��5�� <"��%��>�& �#�*��

�� &� ��>��?FFEVE?��GI�GE&

@D&��4��W�� 5��3��>&�� K��*��:��5,��

��"� ��C��*�� &�� *&�?FFHV�HF��HA�F@&

@H&��.��- ��5�� W�2�3�4��%��/��+�� +�� &�

/�� 0�� J�� .�� *��3�� #��

:1�� /�� :��1��&�� *�@IIGV�?I@�BB?�BF&

@F&�"��3 ��W&�)&VZ� ��>&V"��5&�>&j>��3��K&�J�4�/�� *�K��C��

K��*�-��5*��X��J�4��0� ��#��J��0�� Z��

#��,�"��,�5��*&� �� /��?FFEV@@�?GAH�?GDH&

BI&�� 5��*��<&j�� :&�-&�-� ��C��*�. �4�Z�,�.��C��,��

J�� /�� )�� :1�� C� � #�� %� � C�� #��, � #��

0�� &�C��,� ��/��?FHGV�??�AH�DG&�

FB

B?& � >�� "& �:�� %� �� /�� )�� :1�� %� �-��

C��* � )�� . �4 � #�� C� � #�� X�� K��* � +��& � ��

�� 5��?FDGV�?F�?F�@D&�

B@& �/g��>&V)�� >& �:&V:�� >& �J&V#� ��K&j-�J�� &�#��

�� 0�� /�� 0�� Z��-� ��C��*�. �4&�

��&� ��/��?FFBV�@?�??EB�??EH&

BB&��/�%�� 50��/��-/��/�� K��5��)##��.� ��) �� &�

:��,��,�� *�� *��*��

�*��&�<�:�� >�/�&�?FFHV�BBH�BED�BGE&

BE&�#� ��)��J� ��#��J��/� ��:�� >��K��&�:��%��C��*�. �4�

Z�,�� %� �� )�� K��* �/�� %�� "��

5��*&�:��K��>�?FFDV?I�@H@I�@H@E&�

BG&�+�� /�� :��W��"��5��3�/&& � ��%��0� ��

Z�� #�� 0�� -�� %��J�� & �� >�

��&?FFDVDFF��BIA�B?I&�

BA&� �1�K��:4��&�+��4��*�

�*�� *��&�#��@II?V??�@F?�@FA&

%< JX5k:5

%�� M��

:��;�� 5��C��CC��

,�� 6�� ,� � ��

��M��6��&�

��, ��

��,�� 6�� ,� ��

�� &

0� �� ;��

��

�� &

�� =��7��

��,�� 7�� !��&� ��

FG

�� 6�� V � ��;��6��

,�� 7��$��6��&�

FA

��8!"

��

��

��

�,�

,

,�

,��

��8!�� !��!�"�� !��&!��!

�# !��a�*+!�1��1�� 1��!��1�!.� !��1�"]#�"�

�!��!��" #�!O��AU��=��D�� 9�� 9�� G�� A� �� <��D��O�<��A� �� < ��?(�(@��<��A� �� ?1�(@�� <�"&��%��%=��'�"��%�>��1��H�� ;<��O�1� ��.�� 1�<��&�:��T�� & ��O�YY335Y03��-��&�'��???????????????????????????????????????

�& K�&",/��"%!'L%/��64�"& 6/��,��!�3��'�M�#'3��/��&+' %!�/"��!9,'3%&�/"�� !"#�" � @��A � +%�% � %7%!'%� � % � ?6,-.� �+6!4�,%� � /" � 3�'%,-%& � /6�%, " �% �%,"& "&'%:� 6%!4", "$�4'!<%�"&�3�'%,-%&�&.��&684" '/%&�N�3'�6�#'%�3�4�%,"& "&'%�#"�%!�,"3"&&' %,/��6&��/"�7", '!%-.��4"3O,'3%: � �& ��"1'#"�F6"��%,"& "&'& % �&%'8%�F6%! �% �4"!<�� ?��4%�/"�7", '!%��&�+6!4P"&�+%�%�F6"�,.��', "�?'�%�3�4�%�&6%�&%M/"�"�/"&",7�!7'4", ��/"�&"6�?'!<�:

��&",<��@%A� "�K�F6"��"&+�,/"��%!#64%&�+"�#6, %&�%, "&�/%�3'�6�#'%$�/", ��/%& �+�'4"'�%& �(E �<��%&$ �%+5& �% ��+"�%-.� �" � � ,% �%! % �<�&+' %!%�: �Q � &"#6'�$ � �"3"8"�K �64� "!"?�,"4%�+��&"4%,%�/6�%, "�E�&"4%,%&:��%,"& "&'%�/"�&"6�?'!<��&"�K��"%!'L%/%�3�,?��4"�%�� ',%�/��:$��%3�"&3", %,/��&"�%�3�!" %�/"�D�%4�& �%&�/"�&%,#6"�%� "�'%!:��%,"& "&'%�3�,&'& '�K�/"�%,"& "&'%�#"�%!�%&&�3'%/%��%�64�8!�F6"'��%,"& 0&'3��+%�%�/'4',6'-.��/%�/��,��+5&��+"�% 5�'�:�%�%�F6"�&"6&�+6!4P"&�?6,3'�,"4�,��4%!4", "�0�,"3"&&K�'��F6"�"!"&�&";%4�R7", '!%/�&S:��"&&"�"& 6/��7%4�&�6 '!'L%��/6%&�?��4%&�/"�7", '!%-.�� 64%�/",�4',%/%�RN�7�!64"S�"�64%� �?��4%�%! "�,% '7%�/",�4',%/%�RN�+�"&&.�S:��"�K��"%!'L%/��64�&�� "'��+%�%�&%8"��F6%!�/�&�/�'&� '+�&�/"�7", '!%-.��&"�K�6 '!'L%/��/6�%, "�%��3'�6�#'%:��/�&��&�+%3'", "&$�/6�%, "�%��"%!'L%-.��/%�3'�6�#'%�3�4�%,"& "&'%�#"�%!� 24�&"6&�+6!4P"&�7", '!%/�&�6 '!'L%,/��64%�/"&&%&�/6%&�?��4%&�/"�7", '!%�:��F6"�,5&�F6"�"4�&�3�4+%�%��0�&"��3��%-.��"��&�+6!4P"&�&"�3�4+�� %4�/%�4"&4%�4%,"'�%��6�,.��,�&�/�'&� '+�&�/"�7", '!%-.�:�!04�/'&&�$�F6"�"4�&�&%8"��&"�4%'&� %�/"$�% 0�64�42&�%+5&�%�3'�6�#'%�<%7"�K�%!#64%�/'?"�",-%�,%�&6%��"36+"�%-.��/"�&"6�?'!<�:�

�3�!" % �/" �%4�& �%& �/" �&%,#6"� %� "�'%!� ,� � �%,&�+"�% 5�'� �&"�K � �"%!'L%/%�",F6%, ��&"6�?'!<��"& '7"��%,"& "&'%/�$�/"&&%�?��4%�"!"�,.��&", '�K�/��:� �� F6%, '/%/"�/"�&%,#6" � 3�!" %/% � &"�K �+"F6",%� ,�� 4K1'4� � � C4! �+�� 3�!" %:� ��/"4��3��"�$ � �%�%4", "$�%!#64%&�3�4+!'3%-P"&� ,%�3�!" % �/" �&%,#6"$ � %'& �3�4� �?��4%-.��/" �3�K#6!�&$ �!"&.��/"�,"�7�&�+�51'4�&$��"/6-.��/��?!61��&%,#G',"�$�<"4% �4%��6�',?"3-.��,��!�3%!�/%�+6,-.�:�

�%6 ��'L%-.��+%�%�+%� '3'+%-.��/"�&"6�?'!<��0�7�!6, K�'%: ��"�3�,3��/%�$�&"�K�6 '!'L%/%�64%�/%&�/6%&� ?��4%&�/"�7", '!%� �"4�&6%�3'�6�#'%: ��32��"3"8"�K�7'&' %&�/"�64�40/'3��/%�"F6'+"�/%�,"& "&'%$�/%�&6%�', "�,%-.��% 0�%��%! %�/��<�&+' %!:�"+�'&$�"4�3%&%$�7�32��"3"8"�K� "!"?�,"4%&$�64�%�3%/%�&"4%,%$� � %!'L%,/��F6% �� "!"?�,"4%&:��& "�"& 6/��&"#6"�%&�3�,/'-P"&�"& %8"!"3'/%&�,%��"&�!6-.��C=*J=*�/��,&"!<��%3'�,%!�/"��%M/":

D:��T��/"3'&.��/"�7", '!%� ��&�+6!4P"&�N�7�!64"� ��6�N�+�"&&.��,.��4�/'?'3%�%�

',3'/2,3'%�/"�3�4+!'3%-P"&�/��+��3"/'4", ��3'�M�#'3�:��&��'&3�&�&.��&�4"&4�&�,�&�/�'&� '+�& � /" � 7", '!%-.� � " � "& .� � �"!%3'�,%/�& � 3�4 � % � %+!'3%-.� � /� � %+%�"!<� � 7", '!%/�� +%�%�&68& ' 6'��%�&6%��"&+'�%-.�:� ��"&+"�%/%$�+"!��6&��/"�7", '!%/��$�%! "�%-.��,��"F6'!98�'��/�&�7%&�&�&%,#69,"�&�"�/��3��%-.�$�3�4�F6"/%�/%�+�"&&.��"�4�/'?'3%-P"&�/��3�4+�� %4", ��/��+6!4.�:��6&��/"�4�,' ��"&�"�%�+�"&",-%�3�,& %, "�/��,"& "&'& %�+"�4' "4�F6"�"& %&�&";%4� �% %/%&��%+'/%4", "�"�,.��', "�?'�%4�3�4�%�&6%�&%M/":��",�&�?�"FG", "4", "�+�/"��3��"� � �6+ 6�% �/% �4"48�%,%�/��+6!4.�� @+,"64� 5�%1A ��6 �%� �/", �� /% �+"!" � @",?'&"4%�&6836 O,"�A:�

� ��& %�9%4�& � /" � "&3!%�"3"� � F6" � "7", 6%!4", " � +�/"�K � &6�#'� � %!#64%�3�4+!'3%-.��%',/%�,.��/"&3�' %:

��',7"& '#%/��"&+�,&K7"!�+�/"�K�"13!69�!��/��"& 6/�$�',/"+",/", "�/"�&"6�3�,&", '4", �$�3%&��;6!#6"�,"3"&&K�'��+%�%��4"!<��",3%4',<%4", ��/��&"6�3%&��6�&"��&",<��@%A�,.��364+�'��+��#�%4%�"& %8"!"3'/�:

*��&",<��@%A�+�/"�/"&'& '��/"�+%� '3'+%��%�F6%!F6"��4�4", �:��6%&�/"3'&P"&�/"�,.��+%� '3'+%� ��6�/"�/"'1%��%�+"&F6'&%�/"+�'& �/"� ','3'%/%�,.��%?" %�K��% ",/'4", ��40/'3��+�& "�'��:

U��/%&�%&�',?��4%-P"&��8 '/%&�,"& "�"& 6/��8"4�3�4��+��, 6K�'��<�&+' %!%��+�/"�K�&"��+68!'3%/��3�4�?',%!'/%/"�3'", 9?'3%$�4%, ",/��&"��&'#'!��F6%, ��N�'/", '?'3%-.��/�&�+%3'", "&:

"3!%�� "��!'/��6�4"�?��%4�!'/%&��%&�',?��4%-P"&�%3'4%�%, "&�/"�%&&',%��"& " � ?��46!K�'�: � ��'�4" � /%/% � %4+!% � �+�� 6,'/%/" � /" � ?%L"� � +"�#6, %&$ � "&3!%�"3",/��+!",%4", "�4',<%&�/M7'/%&:�6 ��'L��%�+%� '3'+%-.��/"�VVVVVVVVVVVVVVVVVVVVVVVVVVVVVV�,��+�"&", "�+��;" ��/"�+"&F6'&%:��

�� !"#�"$VVVVVV�/"�VVVVVVVVVVVVV/"�())E:

��4"�/��"&+�,&K7"!��9,36!��6�+%�", "&3�

��&&',% 6�%��"&+�,&K7"!�+"!��+%3'", "�







Doc 6 –

Effect of Synchronized Intermittent Mandatory

Ventilation

Anesthesiology 2001; 95:881–8 © 2001 American Society of Anesthesiologists, Inc. Lippincott Williams & Wilkins, Inc.

Effect of Synchronized Intermittent MandatoryVentilation on Respiratory Workload in Infants afterCardiac SurgeryHideaki Imanaka, M.D.,* Masaji Nishimura, M.D.,† Hiroshi Miyano, M.D.,* Hideki Uemura, M.D.,‡Toshikatsu Yagihara, M.D.§

Background: Synchronized intermittent mandatory ventilation(SIMV) is commonly used in infants and adults. However, fewinvestigations have examined how SIMV reduces respiratoryworkload in infants. The authors evaluated how infants’ changingrespiratory patterns when reducing SIMV rate increased respira-tory load. The authors also investigated whether SIMV reducesinfant respiratory workload in proportion to the rate of manda-tory breaths and which rate of SIMV provides respiratory work-loads similar to those after tracheal extubation.

Methods: When 11 post–cardiac surgery infants aged 2–11months were to be weaned with SIMV, the authors randomlyapplied five levels of mandatory breathing: 0, 5, 10, 15, and20 breaths/min. All patients underwent ventilation with SIMVmode: pressure control ventilation, 16 cm H2O; inspiratorytime, 0.8 s; triggering sensitivity, 0.6 l/min; and positive end-expiratory pressure, 3 cm H2O. After establishing steady-stateconditions at each SIMV rate, arterial blood gases were ana-lyzed, and esophageal pressure, airway pressure, and airflowwere measured. Inspiratory work of breathing, pressure–timeproducts, and the negative deflection of esophageal pressurewere calculated separately for assisted breaths, for spontaneousbreaths, and for total breaths per minute. Measurements wererepeated after extubation.

Results: As the SIMV rate decreased, although minute ventila-tion and arterial carbon dioxide tension were maintained atconstant values, spontaneous breathing rate and tidal volumeincreased. Work of breathing, pressure–time products, and neg-ative deflection of esophageal pressure increased as the SIMVrate decreased. Work of breathing and pressure–time productsafter extubation were intermediate between those at a SIMV rateof 5 breaths/min and those at 0 breaths/min.

Conclusion: When the load to breathing was increased pro-gressively by decreasing the SIMV rate in post–cardiac surgeryinfants, tidal volume and spontaneous respiratory rate bothincreased. In addition, work of breathing and pressure–timeproducts were increased depending on the SIMV rate.

PATIENT-TRIGGERED ventilation (PTV), which includessynchronized intermittent mandatory ventilation (SIMV),assist control ventilation, and pressure support ventila-tion, is commonly used in adults because patient–venti-lator synchrony is thought to enhance patient accep-tance of mechanical ventilation and decrease the work

of breathing (WOB).1,2 SIMV assists the spontaneousbreathing of the patient with a preset number of venti-lator-delivered breaths each minute. Because it can flex-ibly provide ventilatory support over a range of levels,SIMV has two main indications: as a primary means ofventilatory support and as a weaning tool.1 Weaninginvolves a gradual decrease in the number of mandatorybreaths and an increase in the proportion of the venti-latory requirement assumed by the patient. Recently,SIMV using continuous flow, time- and patient-cycled,pressure-limited ventilation has been applied to infantsand children.3–5 SIMV is superior to conventional inter-mittent mandatory ventilation because it improves pa-tient breathing patterns and oxygenation.4–7 However,reports about the effects of SIMV on the respiratoryworkloads of infants are few.8 It remains to be clarifiedwhether infants respond by increasing tidal volume aswell as the frequency of spontaneous breaths when weprogressively increase the load to breathing, in this caseby decreasing SIMV rates. Given the difference in respi-ratory control mechanisms between adults and infants,9

this answer is not obvious.In adults, WOB decreases as the SIMV rate increas-

es.1,10–12 When SIMV is used to wean adult patients frommechanical ventilation, the rate of SIMV is usually de-creased gradually, depending on the patient’s tolerance.Extubation is performed when the SIMV rate is success-fully reduced to less than 5 breaths/min.2 However, nostudy has shown that this protocol is similarly validwhen weaning infants from the ventilator. We tested thehypothesis that pressure control SIMV reduces the respi-ratory workloads of infants in proportion to the SIMVrate and examined which SIMV rate in infants providesrespiratory workloads most similar to those afterextubation.

Subjects and Methods

The study was approved by the ethics committee ofthe National Cardiovascular Center (Osaka, Japan), andwritten informed consent was obtained from the parentsof each patient.

PatientsEleven infants who had undergone cardiac surgery to

repair congenital heart disease were included in thisstudy (table 1). Enrollment criteria were: (1) correctivesurgery for cardiac anomalies; (2) stable hemodynamics;

* Staff Physician, Surgical Intensive Care Unit, ‡ Staff Surgeon, Department ofCardiovascular Surgery, National Cardiovascular Center. † Associate Professor,§ Director, Intensive Care Unit, Osaka University Hospital, Osaka, Japan.

Received from the Surgical Intensive Care Unit, National Cardiovascular Cen-ter, Osaka, Japan. Submitted for publication January 19, 2001. Accepted forpublication May 22, 2001. Support was provided solely from institutional and/ordepartmental sources. Presented in part at the International Conference ofAmerican Thoracic Society and American Lung Association, Toronto, Canada,May 8, 2000.

Address reprint requests to Dr. Imanaka: Surgical Intensive Care Unit, NationalCardiovascular Center, 5-7-1 Fujishiro-dai, Suita, Osaka, Japan 565-8565. Addresselectronic mail to: [email protected]. Individual article reprints may bepurchased through the Journal Web site, www.anesthesiology.org.

Anesthesiology, V 95, No 4, Oct 2001 881

and (3) leakage around the uncuffed endotracheal tube(3.5–4.5 mm ID) of less than 5% of the inspired tidalvolume (VT). We excluded candidates if they hadchronic lung disease, central nervous system disorders,postoperative phrenic nerve damage, or any metabolicdisorder. We diagnosed phrenic nerve damage if (1) theattempt to wean infants from mechanical ventilation wasunsuccessful, (2) abnormal elevation of the unilateraldiaphragm was noted on the chest radiograph duringcontinuous positive airway pressure or after extubation,and (3) paradoxical movement of the affected hemidia-phragm was confirmed by fluoroscopic imaging. All pa-tients were kept in the supine position during the mea-surements. Arterial blood pressure, heart rate, centralvenous pressure, and pulse oximeter signal (PM-1000;Nellcor Inc., Hayward, CA) were monitored continuouslyin all patients. No sedatives or opioids were administeredduring the measurement, although fentanyl (23–47 �g/kgtotal) and midazolam (0.36–1.61 mg/kg) had been admin-istered during the surgery (145–375 min). We did notuse neuromuscular blocking agents or any reverse.

MeasurementsFlow, volume, and airway pressure (Pao) were mea-

sured at the airway opening. A heated pneumotachom-eter (range, 0–35 l/min; model 3500; Hans-Rudolph Inc.,Kansas City, MO) was placed at the proximal end of theendotracheal tube. The pressure difference across thepneumotachometer was measured with a differentialpressure transducer (TP-602T, �5 cm H2O; Nihon Ko-hden, Tokyo, Japan), amplified (AR-601G; Nihon Ko-hden), and converted to flow values. Volume was calcu-lated from digital integration of flow using dataacquisition software (Windaq; Dataq Instruments Inc.,Akron, OH). Intrapleural pressure was estimated fromesophageal pressure (Pes). An esophageal balloon(6 French; Bicore, Irvine, CA) was introduced transna-

sally and positioned in the lower third of the esophagus.The balloon was inflated with 0.2 ml air at the start ofeach measurement. The position of the esophageal bal-loon was adjusted using an occlusion technique whenthe patients regained spontaneous breathing.13,14 Wecompared the maximal deflection in Pes with the maxi-mal deflection in Pao while the infants made respiratoryeffort against occlusion of the airway opening. When theratio of Pes to airway pressure was maximal (� 0.95), wesecured the position of the balloon. Pes and Pao at theproximal end of the endotracheal tube were measuredusing differential pressure transducers (TP-603T, �50 cmH2O; Nihon Kohden) and amplified (AR-601G). Respira-tory inductive plethysmography (RIP; SY07 RespitracePlus; NIMS, Miami Beach, FL) was used to estimate in-spiratory time (TI), VT, and asynchrony between the ribcage and the abdomen. A rib cage band was positionedat the nipple line, and an abdomen band was positioned0.5 cm below the umbilicus. Baseline calibrations for RIPwere made using the qualitative diagnostic calibrationprocedure.15 Maximum compartment amplitude (MCA)was calculated as the sum of the absolute value fromtrough to peak of the rib cage and abdominal compart-ments, regardless of their timing in relation to the sumsignal.15 When the motions of the rib cage and theabdomen are in phase, the ratio of MCA/VT is equivalentto 1.0, where VT is calculated from the summed signal ofthe rib cage and the abdomen. When the motions are outof phase, the ratio of MCA/VT exceeds 1.0. The airwayand esophageal pressure transducers were simulta-neously calibrated at 20 cm H2O using a water manom-eter. Flow was calibrated at 10 l/min with a calibratedflowmeter (P/N 9220; Bird Corp., Palm Springs, CA)using a gas mixture with exactly the same oxygen con-centration for each patient. Volume was calibrated witha 50-ml calibration syringe.

Table 1. Patient Profile

No.Age

(months)BW(kg)

Height(cm) Gender Diagnosis Operation

CPB(min)

ETT Size(mm ID)

CRS(ml � cm H2O�1 � kg�1)

CCW(ml � cm H2O�1 � kg�1) FIO2

Length of MV(h)

1 10 8.10 67 F VSD VSD closure 91 4.0 1.00 5.67 0.4 32 11 10.6 81 M VSD VSD closure 80 4.5 1.26 3.65 0.35 43 2 4.46 57 M VSD VSD closure 80 4.0 0.64 2.94 0.4 54 7 4.78 64 F ASD ASD closure 37 3.5 1.26 7.39 0.4 65 3 3.37 59 M VSD,

ASDVSD and ASD closure 82 3.5 1.13 7.52 0.4 7

6 7 4.32 61 F VSD, MR VSD closure, MVP 91 4.0 0.67 2.53 0.4 47 11 8.45 74 F VSD VSD closure 75 4.5 0.98 3.45 0.5 78 5 4.46 59 F VSD,

ASDVSD and ASD closure 56 4.0 1.36 4.00 0.4 5

9 6 3.42 57 F VSD VSD closure 137 3.5 0.99 4.90 0.4 510 11 9.00 71 F VSD VSD closure 73 4.5 0.83 4.75 0.4 511 7 6.14 68 F VSD VSD closure 80 3.5 0.78 2.68 0.5 5Mean 7.3 6.10 65 80.2 0.99 4.50 0.41 5.1

BW � body weight; CPB � duration of cardiopulmonary bypass; ETT � endotracheal tube; CRS � compliance of the respiratory system; CCW � compliance ofthe chest wall; FIO2 � inspired oxygen fraction; MV � mechanical ventilation; VSD � ventricular septal defect; ASD � atrial septal defect; MR � mitralregurgitation; MVP � mitral valve plasty.

882 IMANAKA ET AL.

Anesthesiology, V 95, No 4, Oct 2001

Study ProtocolWe used V.I.P. Bird ventilators (Bird Corp.) with con-

tinuous flow, time- and patient-cycled, pressure-limitedventilation. Initial ventilatory settings were as follows:assist control mode; positive end-expiratory pressure,3 cm H2O; pressure control ventilation, 16 cm H2O; TI,0.8 s; continuous flow, 20 l/min; and triggering sensitiv-ity, 0.6 l/min. The inspired oxygen fraction (FIO2) wasadjusted by attending physicians to maintain an arterialoxygen pressure (PaO2) greater than 100 mmHg.

We started taking measurements when the patientshad recovered spontaneous breathing in the surgicalintensive care unit and had satisfied our weaning criteria:ratio of PaO2 to FIO2 greater than 200; pH greater than7.30; VT greater than 5 ml/kg; and respiratory rate lessthan 50 breaths/min at a backup ventilatory rate of6 breaths/min and a pressure control of 7 cm H2O.14

Next, we measured compliance of the respiratory system(CRS) and chest wall (CCW). After hyperventilating the pa-tients for 2 or 3 min to lessen their inspiratory efforts, weswitched ventilatory settings to TI of 1.5–2 s, respiratoryrate of 10 breaths/min, and pressure control of 16 cm H2O.At the end-inspiratory phase, conditions of zero gas flow topermit the measurement of quasi-static compliance wereconfirmed on the computer display that we used to mon-itor data acquisition (fig. 1). Compliance was calculatedusing the following formulas16:

CRS � VT/�end-inspiratory Pao � end-expiratory Pao�

CCW � VT/�end-inspiratory Pes � end-expiratory Pes�

We repeated the measurements five times and averagedthem.

Then, we switched the ventilatory mode to SIMV.Five levels of mandatory breathing (0, 5, 10, 15, and20 breaths/min) were applied in random order, withpressure control ventilation of 16 cm H2O, TI of 0.8 s,positive end-expiratory pressure of 3 cm H2O, continu-ous flow of 20 l/min, triggering sensitivity of 0.6 l/min,and termination sensitivity of 5% of the peak inspiratoryflow. Randomization was performed using computer-generated numbers. A setting of zero-rate SIMV is equiv-alent to a continuous positive airway pressure of 3 cmH2O. After establishing steady-state conditions (approx-imately 15 min), airflow, Pao, Pes, rib cage signals, andabdominal signals of RIP were recorded. All these signalswere digitally recorded at a sampling rate of 100 Hz foreach parameter (Windaq) during the last 5 min at eachsetting. Arterial blood samples were analyzed with acalibrated blood gas analyzer (ABL 505; Radiometer,Copenhagen, Denmark).

All subjects underwent successful extubation 90 minafter completion of all measurements. After extubation,we waited at least 60 min and repeated the measurementof Pes, rib cage signals, and abdominal signals of RIP andarterial blood gas analysis during quiet breathing. We did

not measure the flow directly after extubation because itwas likely that the stimuli of face masks would alter thepatients’ inspiratory patterns. Instead, we computed thevolume using RIP signals.14,15

Data AnalysisThe respiratory workload was assessed at mandatory

rates of 0, 5, 10, 15, and 20 breaths/min and duringunassisted breathing after extubation. The onset of in-spiration was defined as the point at which Pes started todecrease. The end of inspiration was determined in twoways: (1) as the zero crossing of inspiratory flow duringmechanical ventilation (fig. 2) or (2) as the peak of theRIP value after extubation. We confirmed that the valuesof each definition of TI were equivalent during mechan-ical ventilation (precision and bias, 0.01 � 0.04 s). TI, theratio of inspiratory time to total respiratory cycle time(TI/TT), and respiratory rate were calculated using theflow or RIP signal. VT and minute ventilation were ob-tained from the expiratory flow.

Inspiratory WOB done by the patient was computed aspreviously described.7,11,17 First, we established a Camp-bell diagram, which consisted of the inspiratory Pes/VT

Fig. 1. Quasi-static measurement of compliance. Flow, airwaypressure (Pao), and volume (Vol) tracings in patient 5 afterhyperventilation. At the end-inspiratory phase, conditions ofzero gas flow were observed. See text for details. Ventilatorysettings: inspiratory time of 2 s, respiratory rate of 10 breaths/min, and pressure control of 16 cm H2O.

883EFFECTS OF SIMV ON WOB IN INFANTS


curve and chest wall compliance curve (fig. 2, bottom).Then we evaluated WOB per breath during each respi-ratory cycle by computing the area bound by the twocurves. WOB per liter of ventilation (WOB/l; J/l) wascomputed as WOB per breath divided by the breath’stidal volume. WOB/l was expressed separately for as-sisted breaths and for spontaneous breaths. WOB perminute (WOB/min; J · min�1 · kg�1) was calculated asthe total inspiratory work done by the patient duringboth assisted and spontaneous cycles in 1 min and wasnormalized by body weight.

The pressure–time product (PTP) is regarded as anindex of oxygen cost of breathing of the respiratorymuscles as well as WOB17–19: here, we used the PTP of

Pes to estimate the inspiratory muscle load. The PTP foreach respiratory cycle was calculated as the area sub-tended by the Pes tracing and the chest wall static recoilpressure for inspiratory time (fig. 2). The chest wallstatic recoil pressure curve was obtained from valuesfor CCW and volume. The PTP per breath (PTPb; cmH2O · s) was calculated both for assisted and forspontaneous breaths. The PTP per minute (PTP/min;cm H2O · s · min�1) was obtained in the same manneras was WOB/min. Negative deflection of esophagealpressure (�Pes) was also measured as the maximal neg-ative excursion from the baseline over breath. Afterextubation, values for VT, minute ventilation, WOB, PTP,and MCA/VT were calculated from the volume obtainedby the RIP.

All recorded breaths for 5 min were analyzed at eachSIMV rate. The values of respiratory rate, VT, �Pes, WOB,and PTP were averaged separately for assisted breaths,for spontaneous breaths, and for the total of all breaths.

Statistical AnalysisData are presented as mean � SD. Using repeated-

measures analysis of variance, mean values were com-pared across different levels of ventilatory support (SIMVrate of 0, 5, 10, 15, and 20 breaths/min, and after extu-bation). When significance was observed, multiple com-parison testing of means was performed using the pairedStudent t test with Bonferroni correction. Comparisonsbetween data for the spontaneous and assisted cycles ateach SIMV rate were made by the two-tailed Student ttest. Statistical significance was set at P � 0.05.

Results

The infants ranged in age from 2 to 11 months (median,7 months), and body weight ranged from 3.37 to 10.6 kg(table 1). Mean duration of cardiopulmonary bypass was80 min (table 1). In six patients with a body weight lessthan 5 kg, blood priming and modified ultrafiltration(150–1,170 ml) were applied during cardiopulmonary by-pass. All patients underwent successful extubation within7 h after the study, and no side effects were notedthrough this study. Table 2 shows respiratory param-eters under each ventilatory setting. As the SIMV ratewas reduced, the frequency of spontaneous breathsand total breaths increased without significant changein TI/TT. The minute ventilation was kept stablewithin the range of 221–254 ml · min�1 · kg�1 at allSIMV rates. At each SIMV rate, VT was greater inassisted breaths than in spontaneous breaths (P �0.01). As the SIMV rate decreased, VT increased signif-icantly both during spontaneous breaths and during as-sisted breaths (P � 0.01). pH, arterial carbon dioxidepressure (PaCO2), PaO2, heart rate, arterial blood pres-sure, and central venous pressure were not affectedsignificantly by the SIMV setting.

Fig. 2. Airway pressure (Pao), flow (inspiration upward), andesophageal pressure (Pes) tracings in patient 10. Pressure–timeproduct was calculated using the integral of the difference be-tween Pes and the chest wall recoil pressure. Work of breathingwas calculated using the Campbell diagram. See text for details.The first and second vertical broken lines show the start andend of inspiration. The hatched area represents integration ofthe Pes versus either time or volume. The dotted area showscontribution of the chest wall recoil pressure to pressure–timeproduct or work of breathing.

884 IMANAKA ET AL.


Figure 3 is a representative tracing of the Pes for manda-tory breaths and for spontaneous breaths at five levels ofSIMV and breathing after extubation. Reducing the SIMVrates resulted in greater negative deflection in Pes both formandatory breaths and for spontaneous breaths. After ex-tubation, the negative deflection in Pes was smaller than atthe SIMV rate of 0 breaths/min (table 3).

Work of Breathing and Pressure–Time ProductsAs the SIMV rate was decreased, WOB increased in

proportion on both a per-liter basis (fig. 4) and a per-minute basis (table 3). After extubation, WOB was largerthan at SIMV rates of 15 and 20 breaths/min (P � 0.05);WOB after extubation was equivalent to a value inter-mediate between that at 0 breaths/min SIMV and that at5 breaths/min SIMV. Similarly, the values of PTP/minincreased in accordance with withdrawal of SIMV rates(fig. 5 and table 3). After extubation, values of PTP/minwere equivalent to values intermediate between those at0 and 5 breaths/min SIMV and were significantly largerthan at SIMV rates of 15 and 20 breaths/min (P � 0.05).The values of MCA/VT observed in RIP were approxi-mately equal to 1.0 at high rates of SIMV, whereas theytended to increase when the SIMV rate was decreased(table 3).

Discussion

The main findings of this study are: (1) when the rateof assisted breaths during SIMV was decreased, VT dur-ing spontaneous breaths increased, the respiratory rateincreased, and minute ventilation and PaCO2 remainedconstant; (2) in proportion to the rate of assisted breaths,

SIMV reduced WOB, PTP, and �Pes; and (3) WOB andPTP values after extubation were intermediate to thosefound between 5 and 0 breaths/min SIMV.

Clinical ImplicationsIn adults, increasing the SIMV rate decreases respira-

tory work.1,10–12 This finding has been empirically ex-trapolated to infants, and applying SIMV and then grad-ually decreasing the SIMV rate has been used as a meansof weaning infants from mechanical ventilation.3 Weundertook this study because in the absence of an ex-tensive body of experimental evidence, it was difficult tobe confident about how effective the SIMV weaningstrategy is for infants. Our findings show that the respi-ratory workload of infants decreases directly in propor-tion to the SIMV rate. WOB and PTP values increased ina linear manner as the rate of SIMV was decreased from20 to 0 breaths/min. When the SIMV rate was reduced,VT increased, respiratory rate increased, and the sameminute ventilation was maintained. These results suggestthat SIMV may be effective as a weaning strategy forinfants.

Patient-triggered Ventilation and SIMVIn adults, SIMV, pressure support ventilation, assist

control ventilation, and other types of PTV have beenwidely used. These types of PTV are reported to providegood patient–ventilator interaction and to give goodresults when weaning from mechanical ventilation.2 Al-though technological innovation, in particular of thesensors and microprocessors that control ventilators, hasmade it possible to extend PTV to pediatric patients,there are few reports of experimental investigations into

Table 2. Parameters at Each Ventilatory Setting

SIMV Rate(breaths/min)

After Extubation20 15 10 5 0

Respiratory rate (breaths/min) 22.8 � 4.2 24.4 � 5.6 26.2 � 5.3 28.3 � 4.9* 28.4 � 6.0* 31.1 � 6.7*†‡SB rate (breaths/min) 2.8 � 4.3 9.2 � 5.6*† 16.2 � 5.4*† 23.3 � 4.8*†‡ 28.4 � 6.0*†‡§ 31.1 � 6.7*†‡§SIMV rate (breaths/min) 20.0 � 0.2 15.1 � 0.3 10.1 � 0.2 5.0 � 0.1 ND ND

Inspiratory time (s) 0.89 � 0.07 0.88 � 0.07 0.84 � 0.09 0.80 � 0.10 0.80 � 0.14 0.66 � 0.11*†‡§TI/TT 0.34 � 0.07 0.35 � 0.08 0.36 � 0.05 0.37 � 0.04 0.37 � 0.04 0.33 � 0.05Minute ventilation (ml � min�1 � kg�1) 254 � 37 239 � 36 233 � 38 225 � 40 221 � 43 229 � 54Tidal volume

SB (ml/kg) 4.4 � 0.9 5.2 � 0.8 6.0 � 1.1 6.7 � 1.2*† 7.8 � 0.9*†‡§ 7.4 � 1.0*†SIMV (ml/kg) 12.1 � 2.0 12.6 � 1.9 13.4 � 2.2 13.9 � 2.2*† ND ND

pH 7.43 � 0.04 7.42 � 0.04 7.41 � 0.03 7.41 � 0.03 7.41 � 0.04 7.41 � 0.04PaCO2 (mmHg) 40.1 � 4.3 42.2 � 3.9 42.7 � 4.8 43.2 � 5.0 43.2 � 4.8 41.7 � 4.4PaO2 (mmHg) 172 � 24 162 � 29 167 � 27 160 � 24 159 � 29 189 � 85Heart rate (beats/min) 146 � 13 147 � 15 147 � 15 146 � 17 146 � 17 143 � 13Systolic BP (mmHg) 101 � 13 99 � 12 101 � 12 101 � 13 102 � 15 100 � 8Mean BP (mmHg) 74 � 10 74 � 9 74 � 8 75 � 7 75 � 9 75 � 6CVP (mmHg) 7.8 � 2.6 7.8 � 2.4 8.0 � 2.1 7.7 � 2.1 8.2 � 2.4 7.5 � 1.9

After extubation, volume was measured by respiratory inductive plethysmography.

* P � 0.05 versus SIMV 20. † P � 0.05 versus SIMV 15. ‡ P � 0.05 versus SIMV 10. § P � 0.05 versus SIMV 5.

SIMV � synchronized intermittent mandatory ventilation; SB � spontaneous breath; ND � not detected; TI/TT � a ratio of inspiratory time to total respiratorycycle time; PaCO2 � arterial carbon dioxide tension; PaO2 � arterial oxygen tension; BP � blood pressure; CVP � central venous pressure.



the application of PTV to small children. Greenough etal.20 evaluated the triggering function of PTV machinesin neonates, while Bernstein and Cleary4–6 evaluatedpatient–ventilator synchrony during PTV. Jarreau et al.7

demonstrated that PTV with peak inspiratory pressuresof 10 and 15 cm H2O reduces WOB in infants more than

conventional intermittent mandatory ventilation does.Dimitriou et al.8 suggested that assist control ventilationprovides faster weaning compared with SIMV, althoughthey did not evaluate WOB. SIMV has proved superior toconventional intermittent mandatory ventilation becauseof its delivery of a larger and more consistent tidalvolume,4 improved oxygenation in neonates with respi-ratory distress syndrome,5 and reduction of mean airwaypressure at similar oxygenation index values.6 In thesestudies, the focus has been on neonate–ventilator syn-chrony and on gas exchange. For infants, we could findonly limited information regarding (1) how infantschange respiratory pattern when increasing respiratoryload during SIMV, (2) whether infants’ WOB is reducedin proportion to the SIMV rate, and (3) the SIMV rates atwhich infants should undergo extubation. To our knowl-edge, the current study is the first to evaluate respiratoryworkloads in infants from near-full SIMV support to post-extubation spontaneous breathing.

We found that respiratory workload was reduced inproportion to the level of SIMV, which is not the case inadults.10,11 In these post–cardiac surgery infants, wedetected small amounts of respiratory work at a highlevel of SIMV, whereas in ventilator-dependent adults,WOB was found to be high at all levels of SIMV.10,11

Leung et al.12 recently found that increasing levels ofSIMV and pressure support ventilation cause progressiveand proportional decreases in the PTP/min values ofadults. Our protocol involved infants who had under-gone cardiac surgery. When the lungs of such patientsare inflated with relatively high peak inspiratory pres-sure (19 cm H2O) during SIMV, the Hering-Breuerreflex may suppress their inspiratory efforts more ef-ficiently than it does in adults. In fact, we found thatinspiratory effort was less at high SIMV rates and thatthis resulted in greater triggering delay and longerinspiratory time (table 2). Second, the infants inour study showed near-normal lung mechanics21

(respiratory system compliance of healthy infants at

Fig. 3. Representative tracings of esophageal pressure. (Left)Mandatory breath. (Right) Spontaneous breath. (Top to bottom)Synchronized intermittent mandatory ventilation (IMV) rates of20, 15, 10, 5, and 0 breaths/min, and after extubation.

Fig. 4. Work of breathing per liter (WOB/l)at synchronized intermittent mandatoryventilation (SIMV) rates of 20, 15, 10, 5, and0 breaths/min, and after extubation. (Left)Mean WOB/l for all breaths. (Right) WOB/lpresented separately for spontaneous (SB)and assisted breaths (SIMV). *P < 0.05versus SIMV rates of 20 breaths/min;†P < 0.05 versus 15 breaths/min; ‡P <0.05 versus 10 breaths/min; §P < 0.05 ver-sus 5 breaths/min.

886 IMANAKA ET AL.


1–12 months, 1.3–2.1 ml · cm H2O�1 · kg�1) andgas exchange status (table 1), whereas in previousreports on adults,10,11 the patients were dependent onmechanical ventilation because of acute lung disease inthe medical intensive care unit10 or acute exacerbationof chronic obstructive pulmonary disease.11 The normallung mechanics and lower respiratory drives of the in-fants in our study may have reduced the SIMV require-ment. Infants and children with primary lung diseaseswould respond differently to decreases in SIMV rates.

Third, our protocol was conducted with continuousflow and flow-triggered SIMV, whereas previous reportsused pressure triggering without continuous flow. Dur-ing pressure-triggered SIMV, insufficient flow deliveryduring early inspiration may increase WOB during spon-taneous breathing cycles because the demand valve cir-cuitry is not responsive enough for efficient synchroni-zation.1 In contrast, continuous flow and flow triggeringmay require less absolute work during spontaneous cy-cles and may provide more efficient support at low SIMVrates than demand valve systems.22–24 Fourth, we ap-plied a mode of pressure control ventilation, whereasprevious reports on adults used volume-controlled ven-tilation with a fixed flow supply profile. Pressure controlventilation reduces breathing effort more effectivelythan volume-controlled ventilation does when used inconjunction with flow triggering.23,24 The flow profileduring pressure target ventilation may provide bettersynchrony from a low level to a high level of SIMVsupport.

ExtubationIn the infants we studied, who had relatively normal

lung mechanics and gas exchange status, WOB, PTP, and�Pes values after extubation were intermediate betweenthose at SIMV of 5 and 0 breaths/min. During continuouspositive airway pressure, although values for WOB andPTP tended to be higher than at SIMV of 5 breaths/minand after extubation, the differences did not reach sig-nificance. Provided that clinical and gas exchange levelsare satisfactory, low levels of SIMV may indicate thefeasibility of extubation.

Fig. 5. Pressure–time products per minute (PTP/min) for eachpatient. The values are presented after combining spontaneousand assisted breaths. SIMV � synchronized intermittent man-datory ventilation.

Table 3. Respiratory Workloads at Each Ventilatory Setting

SIMV Rate(breaths/min)

After Extubation20 15 10 5 0

WOB/l (J/l) 0.09 � 0.08 0.17 � 0.09* 0.30 � 0.11*† 0.38 � 0.14*†‡ 0.50 � 0.18*†‡§ 0.46 � 0.14*†‡SB (J/l) 0.25 � 0.08 0.30 � 0.10 0.37 � 0.13† 0.42 � 0.15† 0.50 � 0.18*†‡§ 0.46 � 0.14†SIMV (J/l) 0.08 � 0.07 0.14 � 0.07 0.23 � 0.08*† 0.28 � 0.12*† ND ND

WOB/min (J � min�1 � kg�1) 0.023 � 0.022 0.042 � 0.023 0.070 � 0.028* 0.088 � 0.036*† 0.112 � 0.041*† 0.103 � 0.031*†SB (J � min�1 � kg�1) 0.007 � 0.009 0.017 � 0.012 0.038 � 0.021 0.068 � 0.030*† 0.112 � 0.041*†‡§ 0.103 � 0.030*†‡SIMV (J � min�1 � kg�1) 0.020 � 0.016 0.027 � 0.013 0.031 � 0.011 0.020 � 0.008 ND ND

PTP/min (cm H2O � s � min�1) 14.1 � 19.2 32.2 � 20.1* 55.8 � 21.1*† 74.3 � 25.5*†‡ 93.8 � 27.4*†‡§ 81.8 � 25.2*†‡�

PTPb (cm H2O � s)SB (cm H2O � s � breaths�1) 1.6 � 0.5 2.0 � 0.7 2.6 � 0.9 2.8 � 1.0† 3.4 � 1.2*†‡ 2.7 � 1.1�

SIMV (cmH2O � s � breaths�1)

0.4 � 0.5 0.9 � 0.5 1.5 � 0.6*† 1.9 � 0.9*† ND ND

�Pes (cm H2O) 0.9 � 1.1 2.0 � 1.2* 3.5 � 1.5*† 4.6 � 1.9*†‡ 6.0 � 2.4*†‡§ 5.1 � 2.1*†�

SB (cm H2O) 3.1 � 1.3 3.8 � 1.5 4.7 � 1.8 5.1 � 1.9† 6.0 � 2.4†‡ 5.1 � 2.1�

SIMV (cm H2O) 0.6 � 0.8 1.1 � 0.8 1.8 � 1.1*† 2.5 � 1.8*† ND NDMCA/VT 1.05 � 0.05 1.08 � 0.06 1.08 � 0.05 1.11 � 0.09 1.13 � 0.12 1.13 � 0.15

After extubation, volume was measured by respiratory inductive plethysmography.

* P � 0.05 versus SIMV 20. † P � 0.05 versus SIMV 15. ‡ P � 0.05 versus SIMV 10. § P � 0.05 versus SIMV 5. � P � 0.05 versus SIMV 0.

SIMV � synchronized intermittent mandatory ventilation; WOB � work of breathing; SB � spontaneous breath; ND � not detected; PTP � pressure–timeproduct; PTPb � pressure–time product per breath; �Pes � negative deflection of esophageal pressure; MCA/VT � maximum compartment amplitude/tidalvolume.



LimitationsThe current study has several limitations. First, after

tracheal extubation, we used RIP to calculate tidal vol-ume, PTP, and WOB. The accuracy of RIP calibrationmay have been less precise because the chests of post–cardiac surgery patients are typically covered with chesttubes and bandages. Second, the patients in our studyhad relatively normal lung mechanics after correctivesurgery for congenital heart diseases. Several patientshad retarded gain of body weight, probably due to pro-ceeding heart failure, although lung mechanics seemednormal for the body weight. The response of moreseriously compromised patients to machine support maybe quite different. Further studies are needed to corrob-orate the relevance of our findings for acutely ill andventilator-dependent infants. Third, the number of pa-tients was small, and the values of WOB and PTP in-cluded wide variations (table 3). However, in each pa-tient, these parameters showed consistent changes inresponse to reducing the SIMV rate (fig. 5). Finally,because we did not measure the electrical activity ofrespiratory muscles, this report says nothing about therole of inspiratory neuromuscular output relating to theexternal inspiratory force.11

In conclusion, after cardiac surgery, for infants withhealthy lungs, SIMV reduces WOB and PTP in proportionto the level of assisted breathing. The analysis of WOBand PTP shows that low levels of SIMV may indicate thefeasibility of tracheal extubation.

References

1. Sassoon CSH: Intermittent mandatory ventilation, Principles and Practice ofMechanical Ventilation. Edited by Tobin MJ. New York, McGraw-Hill, 1994, pp221–37

2. Esteban A, Frutos F, Tobin MJ, Alía I, Solsona JF, Valverdú I, Fernández R, DeLa Cal MA, Benito S, Tomás R, Garriedo D, Macías S, Blanco J: A comparison offour methods of weaning patients from mechanical ventilation. N Engl J Med1995; 332:345–50

3. Donn SM, Nicks JJ: Special ventilatory techniques and modalities, I: Patient-triggered ventilation, Assisted Ventilation of the Neonate, 3rd edition. Edited byGoldsmith JP, Karotkin EH. Philadelphia, WB Saunders, 1996, pp 215–28

4. Bernstein G, Heldt GP, Mannino FL: Increased and more consistent tidalvolumes during synchronized intermittent mandatory ventilation in newborninfants. Am J Respir Crit Care Med 1994; 150:1444–8

5. Cleary JP, Bernstein G, Mannino FL, Heldt GP: Improved oxygenation

during synchronized intermittent mandatory ventilation in neonates with respi-ratory distress syndrome: A randomized, crossover study. J Pediatr 1995; 126:407–11

6. Bernstein G, Mannino FL, Heldt GP, Callahan JD, Bull DH, Sola A, AriagnoRL, Hoffman GL, Frantz ID III, Troche BI, Roberts JL, Dela Cruz TV, Costa E:Randomized multicenter trial comparing synchronized and conventional inter-mittent mandatory ventilation in neonates. J Pediatr 1996; 128:453–63

7. Jarreau PH, Moriette G, Mussat P, Mariette C, Mohanna A, Harf A, Lorino H:Patient-triggered ventilation decreases the work of breathing in neonates. Am JRespir Crit Care Med 1996; 153:1176–81

8. Dimitriou G, Greenough A, Giffin F, Chan V: Synchronous intermittentmandatory ventilation modes compared with patient triggered ventilation duringweaning. Arch Dis Child 1995; 72:188–90

9. Greenough A, Milner AD: Control of the respiratory system, NeonatalRespiratory Disorders. Edited by Greenough A, Roberton NRC, Milner AD. Lon-don, Arnold, 1996, pp 27–47

10. Marini JJ, Smith TC, Lamb VJ: External work output and force generationduring synchronized intermittent mechanical ventilation. Am Rev Respir Dis1988; 138:1169–79

11. Imsand C, Feihl F, Perret C, Fitting JW: Regulation of inspiratory neuro-muscular output during synchronized intermittent mechanical ventilation. ANES-THESIOLOGY 1994; 80:13–22

12. Leung P, Jubran A, Tobin MJ: Comparison of assisted ventilator modes ontriggering, patient effort, and dyspnea. Am J Respir Crit Care Med 1997; 155:1940–48

13. Coates A, Stocks J, Gerhardt T: Esophageal manometry, Infant RespiratoryFunction Testing. Edited by Stocks J, Sly PD, Tepper RS, Morgan WJ. New York,Wiley-Liss, 1996, pp 241–58

14. Takeuchi M, Imanaka H, Miyano H, Kumon K, Nishimura M: Effect ofpatient-triggered ventilation on respiratory workload in infants after cardiacsurgery. ANESTHESIOLOGY 2000; 93:1238–44

15. Adams JA: Respiratory inductive plethysmography, Infant RespiratoryFunction Testing. Edited by Stocks J, Sly PD, Tepper RS, Morgan WJ. New York,Wiley-Liss, 1996, pp 139–64

16. Nunn JF: Elastic forces and lung volumes, Nunn’s Applied RespiratoryPhysiology, 4th edition. Edited by Nunn JF. Oxford, Butterworth-Heinemann,1993, pp 36–60

17. Sassoon CSH, Mahutte CK: Work of breathing during mechanical ventila-tion, Physiological Basis of Ventilatory Support. Edited by Marini JJ, Slutsky AS.New York, Marcel Dekker, 1998, pp 261–310

18. Sassoon CSH, Light RW, Lodia R, Sieck GC, Mahutte CK: Pressure-timeproduct during continuous positive airway pressure, pressure support ventila-tion, and T-piece during weaning from mechanical ventilation. Am Rev Respir Dis1991; 143:469–75

19. McGregor M, Becklake MR: The relationship of oxygen cost of breathingto respiratory mechanical work and respiratory force. J Clin Invest 1967; 40:971–80

20. Greenough A, Pool J: Neonatal patient triggered ventilation. Arch Dis Child1988; 63:394–7

21. Fletcher ME, Baraldi E, Steinbrugger B: Passive respiratory mechanics,Infant Respiratory Function Testing. Edited by Stocks J, Sly PD, Tepper RS,Morgan WJ. New York, Wiley-Liss, 1996, pp 283–327

22. Nishimura M, Imanaka H, Yoshiya I, Kacmarek RM: Comparison of inspira-tory work of breathing between flow-triggered and pressure-triggered demandflow systems in rabbits. Crit Care Med 1994; 22:1002–9

23. Aslanian F, El Atrous S, Isabey D, Valente E, Corsi D, Harf A, Lemaire F,Brochard L: Effects of flow triggering on breathing effort during partial ventila-tory support. Am J Respir Crit Care Med 1998; 157:135–43

24. Giuliani R, Mascia L, Recchia F, Caracciolo A, Fiore T, Ranieri M: Patient-ventilator interaction during synchronized intermittent mandatory ventilation:Effect of flow triggering. Am J Respir Crit Care Med 1995; 151:1–9

888 IMANAKA ET AL.








Doc 7 –

Pressure Support

Conclusion

The PSV mode is an invaluable addition tothe practice of anesthesia. The use of PSVallows patients to breathe spontaneouslywhile reducing the patient’s work of breathing.This can be a clinical benefit in bothoutpatient and same day surgicalanesthesia.

The increased use of LMAs means morespontaneous breathing is permitted duringanesthesia. PSV offers significant benefitsin patients breathing with LMAs becauselower airway pressures are required, thereby decreasing leaks around the LMA seal.

PSV provides a new and clinically usefulventilation strategy that was only commonin the intensive care units and for theextremely ill pulmonary patient. With PSV inanesthesia, a larger patient population canbe served.

From the Ventilation Series

Clinical Focusby Datex-Ohmeda

Pressure SupportVentilation: Impact on

Anesthesia PracticeAssisting the spontaneously

breathing patient

ED41

37-B

/12

02 1

© 2

002

Date

x-Oh

med

a, In

c. A

ll rig

hts

rese

rved

. Sub

ject

to c

hang

e wi

thou

t not

ice.

Prin

ted

in U

SA.

Date

x®, O

hmed

a® a

nd o

ther

trad

emar

ks a

re p

rope

rty o

f Ins

trum

enta

rium

Cor

p. o

r its

sub

sidi

arie

s. A

ll ot

her p

rodu

ct a

nd c

ompa

ny n

ames

are

pro

perty

of t

heir

resp

ectiv

e ow

ners

.

Datex-Ohmeda, Inc.P.O. Box 7550, Madison, WI 53707-7550, USATel. 800 345 2700 • Fax 608 221 [email protected]

Please visit our websites for additional educational materialwww.datex-ohmeda.com • www.us.datex-ohmeda.com

Additional reading:1. Brimacombe J, Keller C, Hörmann C. Pressure

Support Ventilation versus Continuous PositiveAirway Pressure with the Laryngeal Mask Airway.Anesthesiology 2000;92:1621-1623

2. Rathegeber J. Grundlagen der maschinellenBeatmung: Handbuch für Ärzte und Pflegepersonal.Aktiv Druck & Verlag. Göttingen 1999

3. Peter N, Göran H. Ventilatory Support by ContinuousPositive Airway Pressure Breathing Improves GasExchange as Compared with Partial VentilatorySupport with Airway Pressure Release Ventilation.Anesth Analg 2001; 92:950-958

4. Hiroaki T, Toshiaki T, Tomoko I, Tomihiro F, Toshio I,Yuko N, Yoshinori K. The Effect of Breath TerminationCriterion on Breathing Patterns and the Work ofBreathing During Pressure Support Ventilation.Anesth Analg 2001; 92:161-165

Guest Editors

George Arndt, MDProfessor of Anesthesiology

Department of AnesthesiologyUniversity of Wisconsin at Madison, Madison, WI

Dr. Eric PetersResident in Anesthesiology

Department of AnesthesiologyUniversity of Wisconsin at Madison, Madison, WI

Pressure Support Ventilation:Impact on Anesthesia Practice

Though Pressure Support Ventilation (PSV) has beenavailable in the intensive care setting since 1981, ithas only recently become available for use duringgeneral anesthesia. Aside from technical issuesrelating to the basic differences between ICU andOR ventilation, there have been few opportunities toemploy spontaneous breathing during anesthesiauntil the early 1990s and the introduction of theLaryngeal Mask Airway (LMA). The LMA, coupledwith newer inhalation anesthetics, has encouragedclinicians to allow patients to breathe spontaneouslythrough much, or all, of the anesthetic. PSV can beused to assist those patients in whom spontaneousbreathing is elected. The following Clinical Focus,produced by the Department of Clinical Affairs, willdiscuss PSV for anesthesia.

What is PSV?While many other names have been used, the basicidea behind PSV is to support spontaneous breathingby applying pressure to the airway in response to patientinitiated breaths. PSV is patient triggered and eitherflow or time cycled. For PSV to be of value duringclinical anesthesia the patient must be breathingspontaneously. Other ventilation modes such asSynchronous Intermittent Mandatory Ventilation (SIMV),either alone or in combination with PSV are available forpatients who require a mandatory minute volumeprovided by a mechanical ventilator.

During PSV, once a breath is initiated the ventilatorpressurizes the airway to a given inspiratory support

pressure (Psupport). This pressure is usually from 5 to 10cm H2O pressure and provides the additionalventilatory support required to offset the effects ofgeneral anesthesia. Each PSV assisted breath isterminated according to a preset decrease in flow orafter a specific duration, as a backup.

By applying pressure to the airway immediately uponsensing a patient breathe, PSV enhances inspiratoryflow and provides improved gas distribution within thelungs. This enhanced gas distribution results in a lowerpeak airway pressures which is quite advantageouswhen LMAs are used; lower pressure results in less gasleakage around an LMA seal. If LMA seal leaks arepresent, PSV is able to better compensate for theseleaks since the airway pressure is maintained irrespectiveof the volume, accounting for the delivered tidal volumeand leak volume.

The advantage of PSV is its ability to assume some ofthe patient’s increased work of breathing imposed bythe patient breathing system used during anesthesia.PSV can also counter the reduction in functionalresidual capacity as well as the decrease in musclecontraction produced when modern inhalationanesthetics are used. In supporting a patient’sspontaneous breathing, PSV provides for sustainedor enhanced tidal volumes, maintains normalend-tidal CO2 concentrations, and provides forventilator assistance even when using airwaydevices that may introduce leaks such as the LMA.

Inhalation Agents and PSVWhile PSV can be used anytime in a patient that hasthe ability to initiate a spontaneous breath, it is bestsuited to anesthetics where a normal, or near normal,respiratory rate is expected. Such cases may includeagents like sevoflurane or desflurane. These twoagents are well suited to permitting spontaneousbreathing and, as a consequence, for the applicationof PSV. Sevoflurane is becoming the standard for usein children. Desflurane is increasingly common forrapid recovery in adults.

How to implement PSVWhile some parameters used during PSV are patientcontrolled, a pressure support level (Psupport) must beadjusted on the ventilator. Since the volume, rate,and timing of each breath are patient controlled thereis no adjustment for these during PSV. If clinicalconditions require, positive end-expiratory pressuremay be added.

The initial level of Psupport will vary from patient topatient depending on the patient’s pulmonary physiology,compliance and other clinical issues. Since thepatient’s tidal volume is determined by individual lungcharacteristics and breathing efforts, the effect of theadded support will be ventilator augmented tidalvolumes. Clinically, it is easiest to start with lowerlevels of pressure support, in the 5 - 10 cm H2Orange, gradually increasing the support pressure to alevel where an adequate tidal volume is maintained.







Doc 8 –

Performance Characteristics of Five New Anesthesia

Ventilators

� LABORATORY INVESTIGATIONSAnesthesiology 2006; 105:944–52 Copyright © 2006, the American Society of Anesthesiologists, Inc. Lippincott Williams & Wilkins, Inc.

Performance Characteristics of Five New AnesthesiaVentilators and Four Intensive Care Ventilators inPressure-support Mode

A Comparative Bench StudySamir Jaber, M.D., Ph.D.,* Didier Tassaux, M.D.,† Mustapha Sebbane, M.D.,‡ Yvan Pouzeratte, M.D.,‡ Anne Battisti,§Xavier Capdevila, M.D., Ph.D.,� Jean-Jacques Eledjam, M.D., Ph.D.,# Philippe Jolliet, M.D.**

Background: During the past few years, many manufactur-ers have introduced new modes of ventilation in anesthesiaventilators, especially partial-pressure modalities. The currentbench test study was designed to compare triggering and pres-surization of five new anesthesia ventilators with four intensivecare unit ventilators.

Methods: Ventilators were connected to a two-compartmentlung model. One compartment was driven by an intensive careunit ventilator to mimic “patient” inspiratory effort, whereasthe other was connected to the tested ventilator. The settings ofventilators were positive end-expiratory pressures of 0 and 5cm H2O, and pressure-support ventilation levels of 10, 15, and20 cm H2O with normal and high “patient” inspiratory effort.For the anesthesia ventilators, all the measurements were ob-tained for a low (1 l/min) and a high (10 l/min) fresh gas flow.Triggering delay, triggering workload, and pressurization at 300and 500 ms were analyzed.

Results: For the five tested anesthesia ventilators, the pres-sure-support ventilation modality functioned correctly. For in-spiratory triggering, the three most recent anesthesia machines(Fabius, Dragerwerk AG, Lubeck, Germany; Primus, DragerwerkAG; and Avance, GE-Datex-Ohemda, Munchen, Germany) had atriggering delay of less than 100 ms, which is considered clin-ically satisfactory and is comparable to intensive care unit ma-chines. The use of positive end-expiratory pressure modifiedthe quality of delivered pressure support for two anesthesiaventilators (Kion, Siemens AG, Munich, Germany; and Felix,Taema, Antony, France). Three of the five anesthesia ventilatorsexhibited pressure-support ventilation performance character-

istics comparable to those of the intensive care unit machines.Increasing fresh gas flow (1 to 10 l/min) in the internal circuitdid not influence the pressure-support ventilation performanceof the anesthesia ventilators.

Conclusion: Regarding trigger sensitivity and the system’sability to meet inspiratory flow during pressure-supportedbreaths, the most recent anesthesia ventilators have compara-ble performances of recent-generation intensive care unit ven-tilators.

THE new-generation anesthesia ventilators tend to bemore innovative and sophisticated than their predeces-sors to allow a better adaptation of the machines topatients’ ventilatory needs. During the past few years,many manufacturers have introduced new modes of ven-tilation in anesthesia ventilators, especially partial-pres-sure modalities.1–5

Pressure-support ventilation (PSV) is a ventilatorymode in which the patient’s spontaneous inspiratoryeffort triggers the ventilator to provide a variable flow ofgas that increases until airway pressure reaches a se-lected level. Thus, during each spontaneous inspiration,the patient receives pressure-limited assisted ventilation.PSV is used in the intensive care setting to improvepatient–ventilator synchrony and facilitate weaning.6–9

A few studies have suggested that the use of PSV duringgeneral anesthesia could provide some advantages (re-duction of atelectasis, improved gas exchange, de-creased level of sedation).6,10–13 More often, the use ofPSV in the operating room was performed in anesthe-tized patients with a laryngeal mask airway. Therefore,PSV use progressively increased in the operating room,because spontaneous breathing alone or with ventilatoryassistance is recommended with laryngeal mask airwaybecause of leaks.14 Moreover, studies reported that PSVimproves gas exchange and reduces work of breathingin anesthetized adults and children with an endotrachealtube6,12 or laryngeal mask airway.11,13

Several lung model studies, however, demonstratedthat technical differences among intensive care unit(ICU),4,15–17 transport,18,19 and home ventilators17,20

may markedly affect their performance, especially re-garding the trigger function and the pressurization pro-cess. Overall, these studies showed that considerableprogress has been made in the performance and func-tionality of these devices. However, although today

This article is accompanied by an Editorial View. Please see:Tantawy H, Ehrenwerth J: Pressure-support ventilation in theoperating room: Do we need it? ANESTHESIOLOGY 2006;105:872–3.

�

* Assistant Professor in Anesthesiology and Critical Care, ‡ Assistant in Anes-thesiology and Critical Care, # Professor of Anesthesiology and Critical Care,Department of Anesthesia and Critical Care B (DAR B), Hopital Saint-Eloi, CentreHospitalier Universitaire Montpellier, Universite Montpellier 1. † Assistant inAnesthesiology and Critical Care, § Chest Physiotherapist, ** Assistant in CriticalCare, Division of Medical Intensive Care, Cantonal University Hospital, Geneva,Switzerland. � Professor of Anesthesiology and Critical Care, Department ofAnesthesia and Critical Care A (DAR A), Hopital Lapeyronie, Centre HospitalierUniversitaire Montpellier, Universite Montpellier 1.

Received from the Department of Anesthesia and Critical Care B (DAR B),Hopital Saint-Eloi, CHU Montpellier, Universite Montpellier 1, Montpellier,France. Submitted for publication September 14, 2005. Accepted for publicationJune 7, 2006. Support was provided solely from institutional and/or departmentalsources. Equipment used in the study was provided by the manufacturers, eachof which is listed within this article.

Address correspondence to Dr. Jaber: Department of Anesthesia and CriticalCare B (DAR B), Hopital Saint Eloi, 80 avenue Augustin Fliche, 34295 MontpellierCedex 5, France. [email protected]. Individual article reprints may bepurchased through the Journal Web site, www.anesthesiology.org.

Anesthesiology, V 105, No 5, Nov 2006 944

there are numerous anesthesia ventilators available pro-viding PSV with a standard anesthesia circle system, nostudies have evaluated their technical performance. Theaim of the current study was to evaluate in a bench studythe performance of the new generation of anesthesiaventilators for delivering PSV and to assess how theycompare with ICU ventilators.

Materials and Methods

Ventilators TestedThe five anesthesia ventilators evaluated were the Felix

(Taema, Antony France), Kion (Siemens AG, Munich,Germany), Fabius GS (Dragerwerk AG, Lubeck, Ger-many), Primus (Dragerwerk AG), and Avance worksta-tion (GE-Datex-Ohmeda, Munchen, Germany) equippedwith the model 7900 ventilator. This last ventilator canalso be found in the GE-Datex-Ohmeda Aestiva, Aisys,and Aespire anesthesia workstations. The four ICU ven-tilators tested were the Servo 900C (Siemens), Servo 300(Siemens), Horus (Taema), and Evita 4 (Dragerwerk AG).The main characteristics of anesthesia and ICU ventila-tors tested are presented in table 1.

The machines were provided by the manufacturers

after a full revision had been made just before our inves-tigation. All machines were stock, no modification wasperformed, and all were tested in operating conditionsconforming to the manufacturer’s specifications.

Test Lung ModelAll ventilators were connected to a classic, validated two-

compartment lung model (Pneu View AI 2601I TTL; MichiganInstruments, Grand Rapids, MI) which has been described indetail in previous studies.17,20 Briefly, the model consists oftwo separate chambers linked by a rigid metal strip. Onechamber is connected to an ICU ventilator (Evita 4; Drager-werk AG), which is set in volume control mode to mimicpatient inspiratory effort (fig. 1). The magnitude and durationof the latter can thus be adjusted by changing the settings onthis “driving” ventilator. The two chambers being linked, in-flation of the first necessarily inflates the second, which isconnected to the ventilator being tested. The onset of passiveinflation is therefore detected as an “inspiratory” effort by thetested device, which triggers a pressure-support response.The elastance (E) and airway resistance (R) of each compart-ment can be adjusted separately.

Thus, the model allows simulation of various magni-tudes of inspiratory effort, types of respiratory mechan-

Table 1. Main Characteristics of Anesthesia and ICU Ventilators Tested

Inspiratory Trigger UnitsPressurization Phasein Pressure Support

MainInspiratory:Expiratory

Cycling Criteria

Anesthesia ventilatorsFelix (Taema,

Antony, France)Flow trigger Flow: 1 to 10 l/min Fixed Fixed, 25% of peak

inspiratory flowKion (Siemens AG,

Munich,Germany)

Flow or pressuretrigger

1 to 9 arbitrary units Fixed Fixed, 5% of peakinspiratory flow

Fabius (Drager AG,Lubeck,Germany)

Flow trigger andpressure trigger

Flow: 2 to 15 l/min Adjustable with themaximum flow

Fixed, 25% of peakinspiratory flow foradults and 5% forchildren

Primus (Drager) Flow trigger andpressure trigger

Flow: 0.3 to 15 l/min Slope adjustable from0 to 2 s

Fixed, 25% of peakinspiratory flow foradults and 5% forchildren

Avance (GE-Datex-Ohmeda,Munchen,Germany)

Flow trigger Flow: 1 to 10 l/min Fixed Fixed, 25% of peakinspiratory flow

ICU ventilatorsServo 900

(Siemens)Pressure trigger Pressure: 0 to �20

cm H2OFixed Fixed, 25% of peak

inspiratory flowServo 300

(Siemens)Adjustable flow or

pressure triggerFlow: 2 l/min;

pressure: 0 to �17cm H2O

Adjustable pressureramp slope (0 to10% of maximuminspiratory time)

Fixed, 5% of peakinspiratory flow

Horus (Taema) Adjustable flow andpressure trigger

Flow: 0.1 to 5 l/min;pressure: �0.5 to�5 cm H2O

Adjustable pressureramp slope (50 to150 cm H2O/s)

Adjustable 0 to 30l/min

Evita 4 (Drager) Adjustable flow andpressure trigger

Flow: 0.3 to 15 l/min Duration adjustablefrom 0 to 2 s

Fixed, 25% of peakinspiratory flow

ICU � intensive care unit.

945PRESSURE-SUPPORT VENTILATION AND ANESTHESIA VENTILATORS

Anesthesiology, V 105, No 5, Nov 2006

ics, and tested ventilator settings. The ventilator circuitsconnected to each chamber were equipped with a pneu-motachograph and pressure transducer (Biopac Systems,Goleta, CA). Data were acquired online via an analog–digital converter (MP100; Biopac Systems), sampled at500 Hz, and stored in a laptop computer for subsequentanalysis (Acqknowledge software; Biopac Systems).

All measurements were performed in ambient temper-ature and pressure-saturated conditions. Automatic bodytemperature and pressure-saturated compensation wasdisabled on the Evita 4, and all other devices werecalibrated in ambient temperature and pressure-satu-rated conditions. Gas compressibility was not accountedfor, given its negligible quantitative contribution in theconditions of the current tests.21

Measured VariablesInspiratory trigger and pressurization ramp were eval-

uated as previously reported.16,17,20 Figure 2 shows themethod used to calculate the trigger characteristics andthe pressurization phase during PSV based on the airwaypressure–time curve.

Inspiratory Trigger. At each sensitivity conditiontested, triggering performance was assessed accordingto three criteria: the time delay, the pressure fall, and theairway pressure–time product per cycle.

● Triggering delay (DT): time between the onset of in-spiratory effort and that of detectable pressurization.

● Pressure fall (DP): the maximal decrease in airwaypressure measured from its baseline value. DP reflectsin such way the inspiratory work required to triggerthe ventilator; therefore, the lower its value, thesmaller the work required of inspiratory muscles.22

● Airway pressure–time product per cycle (PTP, cm H2O� ms) during the trigger phase, defined as the areaunder the Paw signal during the DT interval (computedas DP � DT).

Pressurization. The pressure–time products at 300and 500 ms for each respiratory cycle (PTP300 and

PTP500) are computed as the area under the time–pres-sure curve 300 and 500 ms after the onset of inspiratoryeffort. These two parameters reflect the speed of pres-surization and the device’s capacity to maintain the setpressure during inspiratory effort. They depend both onthe ventilator’s performance and the magnitude of in-spiratory effort, the former being determined by thepressurization ramp and the flow generated by the de-vice’s bellows or piston. PTP300 and PTP500 are ex-pressed in cm H2O � s.

Experimental ProtocolDT, DP, PTP, PTP300, and PTP500 were measured as

described above and in figure 2 at three successive levelsof PSV: 10, 15, and 20 cm H2O.

To mimic normal and strong inspiratory efforts bypatients, the tidal volumes of the driving ventilator wereset at 220 and 440 ml, respectively. These efforts wereactually associated with pressures 100 ms after occlusion(P0.1) of 2 cm H2O (normal effort) and 4 cm H2O (strongeffort), respectively, as measured on the bench.16,19 Theduration of inspiratory effort on the driving ventilatorwas set at 1 s for all tests. Inspiratory trigger was set atthe maximum sensitivity without the presence of auto-triggering. The pressurization slope was set to its steep-est value. When the inspiratory:expiratory cycling crite-ria was adjustable, it was maintained at its default value.

During the tests, E and R of the “driving” chamberwere set to normal (E � 20 cm H2O � l�1, R � 5.6 cmH2O � l�1 � s).

Fig. 1. Schematic representation of the experimental setup.

Fig. 2. Schematic drawing of the assessment of the performanceof the ventilator triggering systems. DP (cm H2O) and DT (ms)are the changes in pressure and time delay, respectively, re-quired to open the inspiratory valve. PTP is the pressure–timeproduct, expressed as DP � DT (cm H2O � ms). The triggeringsystems were adjusted to their maximal sensitivity. PEEP �positive end-expiratory pressure.

946 JABER ET AL.


For all ICU and anesthesia ventilators, the measure-ments were performed at positive end-expiratory pres-sure (PEEP) of 0 and 5 cm H2O for each of the twodifferent efforts (normal and strong) and for the threePSV levels (10, 15, and 20 cm H2O).

For the five anesthesia ventilators, two levels offresh gas flow were tested: 1 and 10 l/min. Thus, 12conditions were evaluated for each ICU ventilator and24 conditions were tested for each anesthesia ventila-tor.

Statistical AnalysisAll parameter values represent the average of three to

five breaths obtained during steady state. All results areexpressed as mean � SD or median with 95% confidenceinterval, depending on the normal or nonnormal distri-bution of the variables. Comparative statistics relied onthe Kruskal-Wallis one-way analysis of variance on ranks.Post hoc analysis was performed with the Scheffe test ifanalysis of variance reached significance. Significancewas set at P � 0.05.

Fig. 3. Performance of the triggering sys-tems, assessed by pressure drop (DP),trigger delay (DT), and pressure–timeproduct (PTP) of the five anesthesia ven-tilators and the four intensive care unitventilators assessed with a normal levelof inspiratory effort (P0.1 � 2 cm H2O)with positive end-expiratory pressure �0 (ZEEP; Z) and positive end-expiratorypressure � 5 cm H2O (PEEP; P) for thethree pressure-support ventilation (PSV)levels: 10, 15, and 20 cm H2O. * P < 0.05comparisons between machines. # P <0.05 comparisons between ZEEP andPEEP.



Results

One hundred sixty-eight conditions were evaluated,120 for the anesthesia ventilators and 48 for the ICUventilators. None of the ventilators mistriggered, nor didany ventilator prematurely cycle to expiration.

Specific Triggering System EvaluationDT, DP, and PTP values measured in zero end-expira-

tory pressure (ZEEP) and with PEEP for all studied ven-tilators at a level of P0.1 � 2 and P0.1 � 4 cm H2O arepresented in figures 3 and 4, respectively.

On ZEEP, the inspiratory trigger time delay was signif-

icantly shorter with the ICU ventilators compared withall of the anesthesia ventilators except for the Primusand the Avance. PEEP had no impact on the inspiratorytime delay in all ICU ventilators, whereas for anesthesiaventilators, it influenced the performance of the triggersystem for two of the ventilators (Felix and Kion) what-ever the level of PSV studied (10, 15, or 20 cm H2O). Forthe Kion, changes in pressure and trigger time delayrequired to open the inspiratory valve significantly in-creased with PEEP compared with ZEEP. The oppositewas observed with the Felix, in which DT was signifi-cantly shorter in PEEP than in ZEEP (figs. 3 and 4).

Fig. 4. Performance of the triggering sys-tems, assessed by pressure drop (DP),trigger delay (DT), and pressure–timeproduct (PTP) of the five anesthesia ven-tilators and the four intensive care unitventilators assessed with a high level ofinspiratory effort (P0.1 � 4 cm H2O) withpositive end-expiratory pressure � 0(ZEEP; Z) and positive end-expiratorypressure � 5 cm H2O (PEEP; P) for thethree pressure-support ventilation (PSV)levels: 10, 15, and 20 cm H2O. * P < 0.05comparisons between machines. # P <0.05 comparisons between ZEEP andPEEP.

948 JABER ET AL.


For all machines, DT was not affected by the magni-tude of inspiratory effort, except for the Kion, whose TDsignificantly increased with PEEP (figs. 3 and 4) as in-spiratory effort increased.

For all anesthesia ventilators tested, increased fresh gasflow from 1 to 10 l/min did not significantly modifytriggering performance on ZEEP or PEEP.

Dynamic Evaluation of PSVPTP300 and PTP500 values measured on both ZEEP and

PEEP for all studied ventilators according to a level ofP0.1 � 2 cm H2O (normal effort) are presented in figure5 and to a level of P0.1 � 4 cm H2O (strong effort) arepresented in figure 6.

At all levels of PSV studied (10, 15, and 20 cm H2O),the pressurization capacity of all ICU ventilators wascomparable, whereas it varied among the anesthesiaventilators, the difference being more marked withPEEP. At 300 ms in PEEP, the values obtained with theFelix and the Kion were half those obtained with theFabius, Primus, and Avance.

For all ICU ventilators except the Servo 900, PTP300

was not affected by the magnitude of inspiratory effort.However, with the anesthesia ventilators, PTP300 tendedto decrease as inspiratory effort increased (figs. 5 and 6).

Discussion

The current study is the first to provide a strictlyprotocoled bench test evaluation of the performance indelivering pressure support of five new-generation anes-thesia ventilators. The major findings of this trial can besummarized as follows: (1) For the five tested anesthesiaventilators, the PSV modality functions correctly; (2)performance was more homogeneous among the mod-ern ICU ventilators than among the anesthesia ventila-tors; (3) the use of PEEP modified the quality of deliveredpressure support in two anesthesia ventilators (Kion andFelix) but not in the three others (Fabius, Primus, andAvance); and (4) increasing fresh gas flow (1 to 10 l/min)in the internal circuit did not influence the PSV perfor-mance of the anesthesia ventilators.

This bench test study also showed that triggering delayis less than 100 ms for all ICU ventilators except in theolder Servo 900C as reported by previous studies 15–17,19

and is less than 100 ms only for only two anesthesiaventilators (Primus and Avance) (figs. 3 and 4).

The inspiratory workload required to trigger the ven-tilators (i.e., PTP) is very low for the modern ICU venti-lators, in line with previous studies of these ma-chines.17,20 The most recently developed anesthesia

Fig. 5. Pressure–time products at 300 and500 ms for each respiratory cycle (PTP300

and PTP500), computed as the area underthe time–pressure curve 300 ms (top) and500 ms (bottom) after the onset of in-spiratory effort, with normal inspiratoryeffort (P0.1 � 2 cm H2O) with positiveend-expiratory pressure � 0 (ZEEP; Z)and positive end-expiratory pressure � 5cm H2O (PEEP; P) for the three pressure-support ventilation (PSV) levels: 10, 15,and 20 cm H2O. * P < 0.05 comparisonsbetween machines. # P < 0.05 compari-sons between ZEEP and PEEP.



ventilators exhibited comparable performance, the high-est PTP being measured in the Kion, which is the firstanesthesia ventilator with PSV mode and the oldest ofthe anesthesia ventilators tested. The best characteristicsof the pressurization phase for the anesthesia ventilatorswere obtained with the Fabius, Primus, and Avanceunder all tested conditions and were comparable withthose of obtained with the ICU ventilators. The Fabius,Primus, and Avance are “piston ventilators,” which usean electric motor to compress gas in the breathing cir-cuit, creating the driving force for mechanical insuffla-tion to proceed. Therefore, they use no driving gas andmay be used without depleting the oxygen cylinder incase of oxygen pipeline failure. These features may ex-plain in part that these more recent anesthesia ventila-tors have comparable performance to modern ICU ven-tilators.

It is important for the users to know that the Fabius GSand the Fabius Tiro have the same ventilator (electricalpiston) and use the same software management thatregulates the ventilator. Similarly, the Primus (world-wide outside of the United States) and the Apollo(United States) anesthesia stations have the same venti-lator (electrical piston). But the implemented softwarethat regulates the ventilator is different between the

Fabius GS/Tiro and Primus/Apollo. This difference mayexplain in part the differences in performances obtainedwith the two ventilators (Fabius vs. Primus).

The newer technologies used by manufacturers, i.e.,microprocessors, servo valves, and fast and potent tur-bines, have substantially improved both modern anesthe-sia and ICU ventilators regarding global trigger response.

It seems that the industry has so far chosen not toinvest heavily in the development of PSV on anesthesiaventilators. This might seem surprising, because PSV hasbeen available on ICU ventilators for more than 20 yr.Two main factors probably account for this. The first isof a technical nature. Indeed, an anesthesia ventilator iscomposed of two circuits, one for driving gas, the otherfor the patient circuit with the anesthetic gases, withindependent bellows, and the resultant large internalvolume makes it more difficult to implement fast-re-sponding and efficient triggering mechanisms. The sec-ond factor is mainly clinical, i.e., that whereas ICU ven-tilators need to provide a mode of partial ventilatorysupport tailored to the patient’s breathing pattern duringweaning, the need for such a mode in anesthesia hasonly become apparent in recent years. The need is prob-ably linked to the use of laryngeal masks, which is aspontaneous-assisted mode with leaks resembling nonin-

Fig. 6. Pressure–time products at 300 and500 ms for each respiratory cycle (PTP300

and PTP500), computed as the area underthe time–pressure curve 300 ms (top) and500 ms (bottom) after the onset of in-spiratory effort, with a high level of in-spiratory effort (P0.1 � 4 cm H2O) withpositive end-expiratory pressure � 0(ZEEP; Z) and positive end-expiratorypressure � 5 cm H2O (PEEP; P) for thethree pressure-support ventilation (PSV)levels: 10, 15, and 20 cm H2O. * P < 0.05comparisons between machines. # P <0.05 comparisons between ZEEP andPEEP.

950 JABER ET AL.


vasive ventilation in some aspects, and the increasinguse of local–regional anesthesia combined with lightsedation during which spontaneous breathing is main-tained.

Pressure-support ventilatory modes have recently beenintroduced as readily available options on newer anes-thesia ventilators. Unlike ICU ventilators, which ventexhaled gases to the atmosphere and directly release gasfrom the wall outlet into the circuit, anesthesia ventila-tors recirculate exhaled gases into the inspiratory limb. Ithas been suggested but never evaluated that in pressure-support mode, this feature mandates a large internalvolume, which may in turn alter the performance oftriggering and pressurization systems.5,23 In the currentstudy, we did not find a significant difference for theperformance of triggering and pressurization systemsbetween a low (1 l/min) and a high (10 l/min) fresh gasflow for all anesthesia ventilators and all tested dynamicconditions.

We tested the influence of PEEP on the quality of thetrigger and pressure delivering because recent studiessuggested that use of PEEP may have some benefits onpulmonary function.24–28

LimitationsThe most important limitation of this study is the fact

that it was performed on a lung model instead of inpatients. It is possible that performance in patients maydiffer greatly from the performance demonstrated here.The advantage of the model is that mechanical charac-teristics can be standardized and reproduced. In addi-tion, the test lung was modified to simulate spontaneousbreathing. Hence, the different machines were testedunder similar conditions during dynamic experiments.However, it is clear that these laboratory conditions arenot real life and, therefore, that the results of thesebench studies should be extrapolated to patients withcaution. Therefore, a clinical study evaluating these char-acteristics and other aspects of the performance of an-esthesia ventilators, e.g., on gas exchange and comfort inthe operating room, should be performed. In addition,and perhaps more importantly, little is known about thevalidated indications of PSV in anesthetized patients.Further exploration of this topic is clearly warranted.

Conclusion

For the five tested anesthesia ventilators, PSV func-tioned correctly. The efficiency of delivering PSV for theanesthesia ventilators is acceptable, comparable to older-generation ICU ventilators (i.e., Servo 900C); however, itdid not reach the level of performance of the new-generation ICU ventilators for three of the five testedanesthesia ventilators.

The use of PEEP modified the quality of delivered

pressure support for two anesthesia ventilators (Kionand Felix). Increasing fresh gas flow (1 to 10 l/min) inthe internal circuit did not influence PSV performance ofthe tested anesthesia ventilators. Regarding trigger sen-sitivity and the system’s ability to meet inspiratory flowduring pressure-supported breaths, the most recent an-esthesia ventilators have performances comparable tothose of the modern ICU ventilators. Further clinicalstudies should now be conducted to better define theindications of PSV during anesthesia.

The authors thank Jerome Pigeot (Biomedical Engineer, Fisher-Paykell, Courta-boeuf, France) for his technical assistance during the bench study.

References

1. Jaber S, Langlais N, Fumagalli B, Cornec S, Beydon L, Harf A, Brochard L:Performance studies of 6 new anesthesia ventilators: Bench tests. Ann Fr AnesthReanim 2000; 19:16–22

2. Stayer SA, Bent ST, Campos CJ, Skjonsby BS, Andropoulos DB: Comparisonof NAD 6000 and servo 900C ventilators in an infant lung model. Anesth Analg2000; 90:315–21

3. Stayer SA, Andropoulos DB, Bent ST, McKenzie ED, Fraser CD: Volumeventilation of infants with congenital heart disease: A comparison of Dragger,NAD 6000 and Siemens, Servo 900C ventilators. Anesth Analg 2001; 92:76–9

4. Takeuchi M, Williams P, Hess D, Kacmarek R: Continuous positive airwaypressure in new-generation mechanical ventilators: A lung model study. ANESTHE-SIOLOGY 2002; 96:162–72

5. Tung A, Drum M, Morgan S: Effect of inspiratory time on tidal volumedelivery in anesthesia and intensive care unit ventilators operating in pressurecontrol mode. J Clin Anesth 2005; 17:8–15

6. Christie JM, Smith RA: Pressure support ventilation decreases inspiratorywork of breathing during general anesthesia and spontaneous ventilation. AnesthAnalg 1992; 75:167–71

7. Kuhlen R, Putensen C: Maintaining spontaneous breathing efforts duringmechanical ventilatory support. Intensive Care Med 1999; 25:1203–5

8. Pearl RG, Rosenthal MH: Pressure support ventilation: Technology transferfrom the intensive care unit to the operating room. Anesth Analg 1992; 75:161–3

9. Putensen C, Zech S, Wrigge H, Zinserling J, Stuber F, Von-Spiegel T, MutzN: Long-term effects of spontaneous breathing during ventilatory support inpatients with acute lung injury. Am J Respir Crit Care Med 2001; 164:43–9

10. Hedenstierna G, Tokics L, Lundquist H, Andersson T, Strandberg A,Brismar B: Phrenic nerve stimulation during halothane anesthesia: Effects ofatelectasis. ANESTHESIOLOGY 1994; 80:751–60

11. Brimacombe J, Keller C, Hormann C: Pressure support ventilation versuscontinuous positive airway pressure with the laryngeal mask airway: A random-ized crossover study of anesthetized adult patients. ANESTHESIOLOGY 2000; 92:1621–3

12. Bosek V, Roy L, Smith RA: Pressure support improves efficiency of spon-taneous breathing during inhalation anesthesia. J Clin Anesth 1996; 8:9–12

13. von Goedecke A, Brimacombe J, Hormann C, Jeske H, Kleinsasser A, KellerC: Pressure support ventilation versus continuous positive airway pressure ven-tilation with the ProSeal laryngeal mask airway: A randomized crossover study ofanesthetized pediatric patients. Anesth Analg 2005; 100:357–60

14. Devitt J, Wenstone R, Noel A, O’Donnell M: The laryngeal mask airway andpositive-pressure ventilation. ANESTHESIOLOGY 1994; 80:550–5

15. Bunburaphong T, Imanaka H, Nishimura M, Hess D, Kacmarek R: Perfor-mance characteristics of bilevel pressure ventilators: A lung model study. Chest1997; 111:1050–60

16. Richard JC, Carlucci A, Breton L, Langlais N, Jaber S, Maggiore S, FougereS, Harf A, Brochard L: Bench testing of pressure support ventilation with threedifferent generations of ventilators. Intensive Care Med 2002; 28:1049–57

17. Tassaux D, Strasser S, Fonseca S, Dalmas E, Jolliet P: Comparative benchstudy of triggering, pressurization, and cycling between the home ventilatorVPAP II and three ICU ventilators. Intensive Care Med 2002; 28:1254–61

18. Miyoshi E, Fujino Y, Mashimo T, Nishimura M: Performance of transportventilator with patient-triggered ventilation. Chest 2000; 118:1109–15

19. Zanetta G, Robert D, Guerin C: Evaluation of ventilators used duringtransport of ICU patients: A bench study. Intensive Care Med 2002; 28:443–51

20. Battisti A, Tassaux D, Janssens J, Michotte J, Jaber S, Jolliet P: Performancecharacteristics of ten mechanical ventilators in pressure support: A comparativebench study. Chest 2005; 127:1784–92.

21. Lofaso F, Brochard L, Hang T, Lorino H, Harf A, Isabey D: Home versusintensive care pressure support devices: Experimental and clinical comparison.Am J Respir Crit Care Med 1996; 153:1591–9



22. Aslanian P, El Atrous S, Isabey D, Valente E, Corsi D, Harf A, Lemaire F,Brochard L: Effects of flow triggering on breathing effort during partial ventila-tory support. Am J Respir Crit Care Med 1998; 157:135–43

23. Marks J, Schapera A, Kraemer R, Katz J: Pressure and flow limitations ofanesthesia ventilators. ANESTHESIOLOGY 1989; 71:403–8

24. Pelosi P, Ravagnan I, Giuretta G, Panigada M, Bottino N, Tredici S, EccherG, Gattinoni L: Positive end-expiratory pressure improves respiratory function inobese but not in normal subjects during anesthesia and paralysis. ANESTHESIOLOGY

1999; 91:1221–3125. Pontoppidan H: From continuous positive-pressure breathing to ventilator-

induced lung injury. ANESTHESIOLOGY 2004; 101:1015–7

26. Bregeon F, Delpierre S, Chetaille B, Kajikawa O, Martin T, Autillo-Touati A,Jammes Y, Pugin J: Mechanical ventilation affects lung function and cytokineproduction in an experimental model of endotoxemia. ANESTHESIOLOGY 2005;102:331–9

27. Maeda Y, Fujino Y, Uchiyama A, Matsuura N, Mashimo T, Nishimura M:Effects of peak inspiratory flow on development of ventilator-induced lung injuryin rabbits. ANESTHESIOLOGY 2004; 101:722–8

28. Boker A, Haberman C, Girling L, Guzman R, Louridas G, Tanner J, CheangM, Maycher B, Bell D, Doak G: Variable ventilation improves perioperative lungfunction in patients undergoing abdominal aortic aneurysmectomy. ANESTHESIOL-OGY 2004; 100:608–16

952 JABER ET AL.








Doc 9 –

CPAP Principles

40 THE BUYERS’ GUIDE TO RESPIRATORY CARE PRODUCTS

PRINCIPLES OF CPAP AND AUTO-ADJUSTING CPAP DEVICES05

Main image:Royce Degrieistockphoto

BG-05 CPAP:BG-05 CPAP 27/7/09 21:34 Page 40

THE BUYERS’ GUIDE TO RESPIRATORY CARE PRODUCTS 41

PRINCIPLES OF CPAP AND AUTO-ADJUSTINGCPAP DEVICESR. Farré1,2, J.M. Montserrat2,3

1 Biophysics and Bioengineering Unit, Faculty of Medicine, 3 Pneumology Dept, Hospital Clinic, Faculty of Medicine, Universitat de Barcelona (IDIBAPS), Barcelona, and2 CIBER Enfermedades Respiratorias, Bunyola, Spain.

This article was originally printedin Breathe, the journal of the ERSSchool. Breathe 2008; 5: 42–50.

SUMMARY

Obstructive apnoea–hypopnoeasyndrome (OSAHS) is veryprevalent. It causes a considerablereduction in patients’ quality of lifeand induces important short- andlong-term consequences, such astraffic accidents and cardiovasculardiseases. The application ofcontinuous positive airway pressure(CPAP) by means of a nasal mask iscurrently the most widespread andeffective treatment for OSAHS. Thepresent review article will addressthe following questions. What is thephysiological rationale of CPAP?What are the principles of CPAPequipment? How can we optimiseits use? What are auto-adjustingCPAP devices and how do theyoperate? To what extent are theyuseful in the treatment of OSAHS?

INTRODUCTION

OSAHS is the most prevalent of allsleep breathing disorders. This

syndrome is currently a publichealth problem because, accordingto several studies, up to 5% and 2%of the adult male and femalepopulation, respectively, aresuffering from OSAHS [1, 2]. Giventhat this sleep disorder is directlyassociated being overweight [3, 4],it is expected that the prevalence ofOSAHS will increase in parallelwith the growing epidemics ofobesity in Western and developingcountries [5].

OSAHS is characterised byrecurrent obstructions during sleepcaused by an abnormal increase inthe collapsibility of the upperairway, which is triggered byseveral factors, includinganatomical alterations and obesity[6, 7]. Figure 1a illustrates the caseof a normal subject during sleep insupine position. During inspirationthere is a negative (lower thanatmospheric) pressure in the lumenof the upper airway and,consequently, its soft wall wouldtend to collapse. However, in anormal upper airway, thesurrounding muscles are able toexert sufficient force to maintainthe airway open, regardless ofnegative intraluminal pressure

during inspiration, allowingnormal ventilation during sleep(figure 1a). In contrast, in anOSAHS patient, the upper airwaymuscles are unable to withstandthe collapsing force due to negativeintraluminal pressure, so the upperairway tends to collapse.Depending on the degree ofabnormal increase in upper airwaycollapsibility, the OSAHS patientcan experience partial upperairway obstruction (figure 1b) ortotal collapse (figure 1c). In theformer case, a hypopnoea appearsbecause the reduction in airwaylumen results in an increasedresistance high enough to reduceventilation, even though theinspiratory effort is increased.When the upper airway iscompletely collapsed (figure 1c),the patient is no longer able toinspire and experiences anobstructive apnoea. In the mostsevere cases of OSAHS thecollapsibility of the upper airwayduring sleep increases considerablyand collapse is induced even incases where the intraluminalpressure is zero (atmospheric level)or slightly positive. In these severepatients, therefore, the upperairway is collapsed not only

Correspondence

R. FarréUnitat de Biofísica i BioenginyeriaFacultat MedicinaUniversitat de Barcelona-IDIBAPS and CIBER deEnfermedades RespiratoriasCasanova 14308036 BarcelonaSpain

E-mail: [email protected]

Potential conflicts of interest

None declared.

PRINCIPLES OF CPAP AND AUTO-ADJUSTING CPAP DEVICES 05




during inspiration but also duringexpiration.

Figure 2 shows some of the signalsrecorded during a polysomno-graphic study and illustrates thesleep events experienced by apatient with severe OSAHS. Thebreathing flow signal shows threeapnoeas (identified by zero flow)lasting ~20 s each. These apnoeaswere obstructive because thepatient was exerting breathingefforts, as indicated by the thoraco-abdominal movement signals. Eachobstructive event finished with ashort arousal, as evidenced by theelectroencephalogram (EEG)signals. Since the patient wastemporarily awake during thearousal (although not conscious ofthe short awakening), the upperairway muscles were activated(indicated by the genioglossuselectromyogram (EMG) in figure 2)and the airway was open; thepatient was, therefore, able toventilate. However, as the patientfell asleep again immediately afterthe arousal, airway obstructionresumed: after a few breathingcycles with snoring, a new apnoeaensued (figure 2). The arterialoxygen saturation measured bypulse oximetry (SpO2) shows that,as a consequence of the recurrentapnoeas, this patient experiencedintermittent hypoxaemia with arepetition period of ~40 s (figure 2).

The short-term symptoms describedby OSAHS patients are related toalterations in normal ventilation(choking, gasping or dry mouth)and disruption of sleep architecturecaused by recurrent arousals(excessive sleepiness, lack ofattention and irritability). Patientswith OSAHS have an increased riskof traffic accidents, probably as aresult of somnolence [8]. Moreover,the nocturnal events chronicallyexperienced by OSAHS patientscontribute to the development oflong-term comorbidities, such ascardiovascular and cerebrovasculardiseases and inflammatory,metabolic, cognitive and moodalterations [9-14].

��

��

��

Figure 1. Diagram showing the upper airway patency during inspiration: a) in a normalsubject; b) in a patient experiencing an obstructive hypopnoea with snoring (represented bythe sound produced); c) in a patient experiencing an obstructive apnoea; and d) in a patientsubjected to nasal CPAP. The red arrows in b) and c) indicate net collapsing force on theupper airway wall. The red arrows in d) indicate that application of nasal CPAP results in anet force opening the upper airway.

��

��

��

��

��

��

��

��

��

��

Figure 2. Physiological signals recorded during a nocturnal polysomnography in a patientwith OSAHS. EEG1 and EEG2 correspond to the C4/A1 and C3/A2 EEG channels. EMGrefers to EMG of the genioglossus. Snoring was monitored by a sound recording. Flow refersto the breathing flow. Red rectangles in EEG1 indicate arousals. Vertical white lines indicate10-s periods. Tho: Thoracic breathing effort; Abd: Abdominal breathing effort; SpO2: arterialoxygen saturation measured by pulse oximetry (ranging from 93–76% in this example).




CPAP

Several approaches can be used totreat OSAHS. The first option is torecommend that the patient losesweight, avoids sleeping in a supineposition and avoids theconsumption of alcohol andsedative drugs. However, in mostpatients, these behaviouralmeasures are not effective fornormalising sleep, and more activetreatments are required. It has beenshown that in some patients thenocturnal use of mandibularadvancement devices aimed atprotruding the mandible could beeffective for increasing thedimensions of the upper airwayand maintaining its patency duringsleep [15]. Other patients couldbenefit from surgical treatment toreduce anatomical upper airwayobstruction in the nose, oropharynx

and hypopharynx [16]. However,for the vast majority of OSAHSpatients the most effectivetreatment is the nocturnalapplication of nasal CPAP [17].

Nasal CPAP does not eliminate theprimary causes that increase upperairway collapsibility in OSAHS. Infact, CPAP is a palliative treatmentfor mechanically preventing upperairway obstruction. NocturnalCPAP, applied by means of a nasalmask (figure 1d), imposes a positiveintraluminal pressure on the upperairway that plays a role similar tothat of normal upper airwaymuscles. As illustrated in figure 1d,CPAP opens the upper airway andprevents its partial or totalobstruction. The effectiveness ofCPAP in preventing upper airwaycollapse in OSAHS is illustrated bythe computed tomography (CT)

scan of a patient’s pharyngeal areaduring sleep (figure 3). The upperimages (figure 3a and b) show twosections of the upper airwayobtained when the patient wassleeping under normal conditions(no CPAP). The right scan section(figure 3b) shows that the lumen ofthe upper airway was extremelyreduced, indicating a virtuallyclosed airway. When the patientwas subjected to CPAP, the upperairway lumen increasedconsiderably at this point ofobstruction. The other upper airwaysections also increased their lumenwhen CPAP was applied, indicatingthat nasal pressure preventedobstruction along the wholecollapsible airway (figure 3c and d).

The value of nasal pressure thatnormalises breathing during sleepdoes not depend on the severity of aparticular patient’s OSAHS, asmeasured by the number ofnocturnal respiratory events(apnoeas and hypopnoeas per h)but, instead, depends on the degreeof collapsibility of the patient’supper airway. Accordingly, eachpatient should be subjected to anindividual CPAP titration procedureduring sleep, in order to determinethe optimal nasal pressure fortreatment. Figure 4 shows the datacorresponding to a 1-night CPAPtitration in a patient with OSAHS.At the beginning of the night, whenawake, the patient was subjected toa minimal CPAP of 4 cmH2O (0.4kPa). When the patient started tosleep, respiratory events (mainlyobstructive apnoeas) and markedoxygen desaturations appeared. Thesleep technician then graduallyintensified the application of nasalpressure. As CPAP increased, thenumber of apnoeas decreased andthe number of hypopneas increased,indicating that the upper airwayobstruction was being progressivelyreduced and breathing was beingnormalised (figure 4). Similarly, themagnitude of oxygen desaturationswas also progressively decreasing.When CPAP was equal to 9 cmH2O(0.9 kPa), there were no longer anyobstructions or desaturations.

��

� ��

Figure 3. Axial CT scans of the upper airway obtained during sleep from a patient withOSAHS. a) and b) Two head sections of the untreated patient. c) and d) The same twosections during application of CPAP. The yellow arrows indicate the upper airway lumen.


PRINCIPLES OF CPAP AND AUTO-ADJUSTING CPAP DEVICES 05


The normalisation of sleep wasreflected by the fact that the patientachieved rapid eye movement(REM) sleep. Subsequently, thetechnician maintained CPAP at 9-10cmH2O (0.9-1.0 kPa) for 3-4 h,observing no clear improvementbetween 9 and 10 cmH2O. Thequality of sleep was good, as thepatient experienced two more REMsleep periods. To test whether CPAPcould be reduced while maintainingnormal sleep, the technician reducednasal pressure to 8 cmH2O (0.8 kPa),with the result that hypopnoeas andoxygen desaturations appearedagain, indicating that 9-10 cmH2O(0.9-1.0 kPa) was the optimal nasalpressure for treating this OSAHSpatient.

PRACTICAL ISSUESREGARDING CPAPEQUIPMENT

Although in a patient treated withCPAP the primary causes of OSAHSremain present, from the functionalviewpoint, his/her sleep resemblesthat of a healthy subject. However,CPAP is effective only as long as thepatient is subjected to the treatment.In this regard, it has been shown

that there is a linear dose-responserelationship between the number ofhours of CPAP use per night andthe attainment of normal levels ofobjective and subjective daytimesleepiness [18]. Accordingly, anyeffort made to improve the patient’sacceptance of CPAP treatment willenhance the effectiveness of thetherapy [19]. To this end, it isimportant to select high-qualityCPAP equipment, use it inaccordance with the manufacturer’sspecifications and train patients onCPAP therapy.

The CPAP systems used for OSAHStreatment are usually based on ablower and an exhalation port(intended leak orifice), as shown infigure 5. The blower takes room airand generates a constant airflowthrough a flexible tubing (~1.5 mlength, ~2 cm internal diameter).When the patient is not breathingand the nasal mask is adequatelyfitted on the patient’s face to avoidleaks between the mask and theskin, all the airflow generated by theblower reaches the atmosphereagain through the exhalation port.Accordingly, the pressure (CPAP) atthe nasal mask is the product of theairflow and the resistance of the

exhalation port. For a givenexhalation port, the value of CPAPcan be increased or decreased bymodulating the magnitude of theflow generated by the blower. Inaddition to being the nasal pressuresource, the airflow generated by theblower plays also the important roleof avoiding rebreathing. To this end,a minimum airflow through theexhalation port is required toadequately renew the air inhaled bythe patient. In commerciallyavailable CPAP devices, thepressure ensuring sufficient airrenewal, and, therefore, theminimum selectable CPAP value, isgenerally ~4 cmH2O (0.4 kPa). Inmost devices, the exhalation port isan orifice characterised by nonlinearresistance. This type of resistor hasthe advantage of a range of blowerairflow (and therefore machinenoise) covering the full range oftherapeutic CPAP values (4-16cmH2O; 0.4-1.6 kPa) that is lowerthan that of an exhalation port withlinear resistance. The exhalationport can be either an orifice in themask wall (as in figure 5) or aspecial device connecting the tubingoutlet and the nasal mask. In thelatter case, the air volume in thenasal mask is an additional smalldead space for breathing.

As indicated in figure 5, for a givenairflow generated by the blower thevalue of nasal pressure is constant,as long as the patient is notbreathing. When the patientinspires, however, an air fractionfrom the blower flow enters thelungs and hence the airflowthrough the exhalation port isreduced. Therefore, nasal pressure,which depends on this flowmagnitude, is decreased and theequipment represents a load to thepatient’s breathing. Thisconventional design of CPAPequipment (figure 5) poses twomain technical problems withregard to optimising the system forpatient comfort. First, given that the effective resistance of theexhalation port is considerable, thepatient’s breathing flow mainlycirculates through the tubing and

Figure 4. Data from a 1-night CPAP titration in a patient with OSAHS. Staging: sleep status(W: wake; R: REM; 1-4: non-REM sleep stages 1-4). Body pos: body posture. CPAP: nasalpressure applied. Desat: arterial oxygen desaturation, as indicated by SpO2. The plot at thebottom indicates the number of different respiratory events detected: obstructive apnoeas,mixed apnoeas, central apnoeas and hypopnoeas. The numbers in the boxes shows the totalnumber of events with a fall in SpO2 >4%. The time scale indicates time from titration start.

!��

�� "

#��$%�&'#�&�

�()��

�*

*

��

�**

+,

��

$$

-��

!

.(��! ��/��0��/��!/��

�*��,1

*� $* *� *$ *% *, *2 *+$* $* $* $* $*

��3�)�


PRINCIPLES OF CPAP AND AUTO-ADJUSTING CPAP DEVICES


05

blower [20]. Consequently, theblower should be designed togenerate high pressure whilepresenting a low load (resistance) tobreathing. Secondly, in order tokeep the nasal pressure (i.e. CPAP)constant, the blower should be ableto automatically modify thegenerated airflow with the aim ofkeeping the airflow constantthrough the exhalation port,regardless of the patient’s breathing.One common method for this typeof regulation is the measurement ofthe flow and pressure generated atthe CPAP device and the calculationof the pressure at the nasal maskfrom the known airflow resistanceof the tubing and exhalation port.This procedure requires the tubingand the exhalation port connectedto the CPAP machine to bematched, otherwise the calculationof nasal pressure would beincorrect. In order to circumventthis potential problem, some CPAPdevices measure mask pressuredirectly by means of a thin catheterplaced along the CPAP tubing. As itis important to maintain a fairlyconstant nasal pressure, the mainCPAP device quality index is givenby the magnitude of the “swings”in nasal pressure during breathing:the smaller the swings, the betterthe CPAP equipment.

An adequate selection of CPAPequipment (i.e. compatible CPAPmachine, tubing and exhalationmask) does not ensure correcttreatment application, as two types

of unintended leaks could reducethe performance of the CPAPsetting. An air leak between thenasal mask and the skin as theresult of an unsuitable mask fittingcould affect the therapy: the flowgenerated by the blower wouldincrease (and, hence, the noise) andthe nasal pressure could be lowerthan expected. Such a leak couldalso cause patient discomfort,particularly if the leak airflow isdirected toward the eyes. The needto reduce mask leaks as much aspossible highlights the importanceof adequately choosing the nasalmask type that best fits the patient.Good mask fitting should beachieved without any excessivecompression, as this would damagethe patient’s skin and, therefore,compromise tolerance of CPAP.Another type of leak that could

negatively affect CPAP treatmentoccurs when the patient’s mouth ispartially open. In this case, there isa constant airflow through theupper airway, from the nostrils atpositive pressure to the mouth atzero (atmospheric) pressure, withthe result that the effective pressureat the upper airway lumen is lowerthan expected. The use of achinstrap could help to preventmouth leaks in some patients. Animportant additional problemrelated to mouth opening duringCPAP is the presence of acontinuous flow of dry and coldroom air, which could result innasal and throat mucosa drynessand irritation, thereby causingdiscomfort, and even rhiniticsymptoms, to the patient. Apossible way of reducing the risk ofthis nasal drying is to use a heatedhumidifier (figure 5) [21]. Thepotential advantage of a humidifieris counterbalanced by somepotential drawbacks: the need toclean the water chamber to avoidcontamination; more expensiveequipment; and increased breathingroute resistance. Althoughhumidifiers are useful for somepatients, there is no clear evidenceto recommend their systematic usefor CPAP therapy in OSAHSpatients.

Prescribing updated and high-quality CPAP equipment isobviously important for patient

Figure 5. Diagram of a conventional CPAP system.

Figure 6. Training session on the use of CPAP.

&�� )�� 4��

'��5)�� 4��

6�� !� ()�� 4��

&�� )4��!�

��)��7� � .�3��4� ��)!��

��5

'��5)� �4� �




compliance. However, it should bementioned that patient adherence toCPAP is considerably improved byimplementing some routineprotocols to the start and follow-upof the treatment [19, 22]. On the onehand, initial educational andtraining sessions before CPAPtitration allow the patient to betterunderstand the treatment andimprove adaptation to theequipment (figure 6). On the otherhand, periodic follow-up sessionsare useful for answering anyquestions posed by the patientabout the treatment, and also forthe early detection and solution ofproblems, such as discomfort withthe mask, air leaks or rhinitic side-effects, that could reduce adherenceto the treatment [19, 22].

AUTO-ADJUSTING CPAPDEVICES

Upper airway collapsibility inOSAHS patients depends on severalfactors; therefore, it may vary in theshort and long term. Different bodypostures during sleep (supineposture promotes airway collapse,compared with lateral decubitus)could cause changes within a singlenight. This well-known fact is takeninto account during routine CPAPtitration, when at least supine sleepposture is studied. Moreover,changes in upper airwaycollapsibility within consecutivenights could be the result of alcoholingestion or drug treatment,particularly with drug affectingmuscle tone. Furthermore, within alonger time period (over a period ofweeks or months), upper airwaycollapsibility could change as thepatient’s body weight varies. Giventhat the CPAP required to avoidobstructive events is directlydetermined by upper airwaycollapsibility, the optimal CPAPwould be not the same over time.Consequently, a conventional CPAPdevice would apply a fixed nasalpressure that could be higher orlower than required, depending onthe patient’s current situation. Auto-adjusting CPAP devices are designed

to solve this problem. These“intelligent” devices are intended todetect a patient’s respiratory eventsand modify the applied CPAP tonormalise patient’s breathing.

In addition to conventional CPAPequipment elements (figure 5),auto-adjusting CPAP devicesincorporate a complex algorithm(figure 7). The sensors in the deviceestimate the patient’s breathing byassessing snoring, flow pattern and,in some devices, airwayobstruction. The first step in thealgorithm of an auto-adjustingCPAP device is to correctly detectand classify the different breathingevents (normal breathing, apnoea,hypopnoea, snoring and flowlimitation) from the availablesignals. The device must be able todistinguish true obstructive eventsfrom typical artefacts, such as thosecaused by awakening of the patient,cough, sighs or mouth breathing.The second step in the auto-adjusting CPAP device algorithm isto modify the nasal pressureapplied in response to the breathing

events that are detected. Figure 8 isan example of the functioning of anauto-adjusting CPAP device. Thedevice was subjected to a bench testby connecting it to a simulatedOSAHS patient who, depending onthe applied pressure, exhibitedapnoeas, hypopnoeas, flowlimitation events or normalbreathing. Initially, the simulated

�� 84��)�� 999�

��!�/��:��!��)��)��7� �

�7��!)��)� !�4� !��!� !��

-� ��)��) ��

��

��!��!�� )�� (

Figure 7. Diagram showing the rationale ofautomatic CPAP devices.

Figure 8. Nasal pressure applied by a commercially available auto-adjusting CPAP devicewhen subjected, on the bench, to a simulated patient with OSAHS. As indicated by the timescale, the bottom plot is the continuation of the plot on top. Details of the flow pattern areshown at different relevant times (positive flow corresponds to inspiration). The green linesindicate 10 cmH2O (1 kPa) of nasal pressure (Press).

� ��

� 3

.��

�,�*,*

*9,*

/*9,/�

�9*

��

)';�

% 2 1 �* �� % �1�2��3�)3��

� ��

� 3

.��

�,�*,*

*9,*

/*9,/�

�9*

��

)';�

�* �� % �2 �1 $* $��3�)3��




05

patient was breathing normally andthe applied CPAP was 4 cmH2O (0.4kPa). Subsequently, as thesimulated patient fell asleep,apnoeas ensued. The devicedetected the apnoeas and increasedthe CPAP. As the CPAP becameprogressively higher, the simulatedpatient then exhibited hypopneasand flow limitation; finally, thebreathing pattern was normalisedwhen CPAP reached 12 cmH2O (1.2kPa). From then on, the auto-settingdevice slightly decreased orincreased CPAP to detect theappearance and disappearance ofabnormal breathing. This processmaintained the optimal CPAP(minimum value avoidingbreathing events) for the simulatedpatient.

Given that auto-adjusting CPAP isa relatively new technology, someissues affecting its potential clinicaluse are still open to debate. In

contrast to the detection andclassification of events [23], thereare no generally accepted criteriafor defining the optimum methodof modifying nasal pressure inresponse to breathing events. Forinstance, after how manyapnoeas/hypopnoeas/snoringevents should pressure beincreased? What should the stepfor increasing pressure be? Whatshould the rate for modifyingpressure be? If no events aredetected, how long should thedevice wait before reducingpressure? Given the number ofopen points, each manufacturer ofan auto-adjusting CPAP device usesa proprietary algorithm that isusually undisclosed. Consequently,devices provide different resultswhen subjected to the samebreathing pattern [24, 25]. As anexample, figure 9 shows theresponse of three currentlyavailable auto-adjusting CPAP

devices when subjected to a normalbreathing pattern, followed by apersistent period of flow limitationduring a well-controlled bench test.Two devices responded byincreasing nasal pressure, but thepressure increase rate was clearlydifferent. The third device did notmodify pressure when subjected toan abnormal breathing pattern(figure 9). This lack of responsecould be caused by the device’sinability to detect the event when itoccurred, or it could mean that thedevice algorithm did not considerthis well-detected and classifiedevent as a reason for modifyingpressure. Such differences betweendevices, which have also beendocumented in patient studies,make it difficult to assess the cost-effectiveness of auto-adjustingCPAP and to compare differentclinical studies, as the results arealways dependent on the deviceused in each test [25].




1. Young T, Palta M, Dempsey J,Skatrud J, Weber S, Badr S. Theoccurrence of sleep-disorderedbreathing among middle-aged adults.N Engl J Med 1993; 328:1230–1235.

2. Duran J, Esnaola S, Rubio R, IztuetaA: Obstructive sleep apnea-hypopnea and related clinicalfeatures in a population-basedsample of subjects aged 30 to 70 yr.Am J Respir Crit Care Med 2001;163: 685-689.

3. Schwartz AR, Patil SP, Laffan AM,Polotsky V, Schneider H, Smith PL.Obesity and obstructive sleep apnea:pathogenic mechanisms and

therapeutic approaches. Proc AmThorac Soc 2008; 5: 185-192.

4 Young T, Peppard PE, Taheri S.Excess weight and sleep-disorderedbreathing. J Appl Physiol 2005; 99:1592-1599.

5. Prentice AM. The emerging epidemicof obesity in developing countries. IntJ Epidemio. 2006; 35: 93-99.

6. Patil SP, Schneider H, Schwartz AR,Smith PL. Adult obstructive sleepapnea: pathophysiology anddiagnosis. Chest 2007;132: 325-337.

7. Eckert DJ, Malhorta A.Pathophysiology of adult obstructivesleep apnea. Proc Am Thorac Soc

REFERENCES

To date, several clinical studieshave been carried out to test thetherapeutic application of auto-adjusting CPAP for treatingOSAHS. Although these deviceshave been shown to apply a meannasal pressure lower thanconventional fixed CPAP devices,their effectiveness in reducing thenumber of sleep breathing eventsis similar with both nasal pressuremodalities [26-28]. Accordingly,the currently available data do not

make it possible to recommendsystematic application of auto-adjusting CPAP to the generalspectrum of OSAHS patients,particularly taking into account itsgreat cost when compared withconventional CPAP. This CPAPmodality could be better suited forselected subpopulations ofOSAHS patients, for instancethose exhibiting a clear number ofrespiratory events when changingbody posture or those treated with

a high level of CPAP. However,the cost-effectiveness of auto-adjusting CPAP for OSAHStreatment needs to be bettersubstantiated in future studies[17].

Interestingly, auto-adjusting CPAPdevices can be also used for anapplication different from theoriginal intention (continuoustailoring of a patient’s treatment).In fact, these devices are able tocarry out simplified CPAP titrationeither in the sleep laboratory or ina patient’s home. Instead ofmanually modifying nasalpμressure to determine the optimalCPAP, auto-adjusting devices canautomatically determine theoptimal pressure, thereby reducingthe workload in sleep laboratories(figure 8). CPAP titration at homehas the advantage that the patientis sleeping in his/her actualenvironment and that the titrationprocess can be extended to severalnights at an affordable cost (whencompared with titration in thesleep laboratory). Simplifiedtitration with auto-adjusting CPAPdevices has proven useful whenapplied to selected subpopulationsof patients [29, 30]. However, thegeneralised use of this titrationmodality should be cautious [17,31], as a number of patientsrequire full polysomnographicCPAP titration in the sleeplaboratory. ■

Figure 9. Nasal pressure applied by three commercially available auto-adjusting CPAP deviceswhen subjected, on the bench, to an initial pattern of normal breathing (up to minute 5)followed by a pattern of persistent flow limitation. The flow signal is shown on the top; adetail of the flow patterns in minute 5 shows the transition from the normal to the flow-limited breathing pattern (positive flow corresponds to inspiration). The green lines indicate10 cmH2O (1.0 kPa) of nasal pressure (Press).

� ��

� 3

.��

�,�*,*

� ��

� 3

.��

�,�*,*

� ��

� 3

.��

�,�*,*

*9,*

/*9,/�

�9*

��

)';�

% 2 1 �* �� % �1�2��3�)3��




05

2008; 5: 144–153.

8. Mulgrew AT, Nasvadi G, Butt A, et al.Risk and severity of motor vehiclecrashes in patients with obstructivesleep apnoea/hypopnoea. Thorax2008; 63: 536-541.

9. Sateia MJ. Neuropsychologicalimpairment and quality of life inobstructive sleep apnea. Clin ChestMed 2003; 24: 249–259.

10. Arzt M, Young T, Finn L, et al.Association of sleep-disorderedbreathing and the occurrence ofstroke. Am J Respir Crit Care Med2005; 172: 1447–1451.

11. Wolk R, Somers VK. Sleep and themetabolic syndrome. Exp Physiol2007; 92: 67-78.

12. Caples SM, Garcia-Touchard A,Somers VK. Sleep-disorderedbreathing and cardiovascular risk.Sleep 2007;30: 291-303.

13. Golbin JM, Somers VK, Caples SM.Obstructive sleep apnea,cardiovascular disease, andpulmonary hypertension. Proc AmThorac Soc 2008; 5: 200-206.

14. Tasali E, Ip MSM. Obstructive sleepapnea and metabolic syndrome:alterations in glucose metabolismand inflammation. Proc Am ThoracSoc 2008; 5: 207-217.

15. Chan ASL, Lee RWW, Cistulli PA.Non-positive airway pressuremodalities: mandibularadvancement devices/positionaltherapy. Proc Am Thorac Soc 2008;5: 179-184.

16. Won CHJ, Li KK, Guilleminault C.Surgical treatment of obstructivesleep apnea: upper airway andmaxillomandibulr surgery. Proc AmThorac Soc 2008; 5: 193-199.

17. Sanders MH, Montserrat JM, FarréR, Givelber RJ. Positive pressuretherapy: a perspective on evidence-based outcomes and methods ofapplication. Proc Am Thorac Soc2008; 5: 161-172.

18. Weaver TE, Maislin G, Dinges DF,et al. Relationship between hoursof CPAP use and achieving normallevels of sleepiness and dailyfunctioning. Sleep 2007; 30: 711-719.

19. Weaver TE, Grunstein RR.Adherence to continuous positiveairway pressure therapy: thechallenge to effective treatment.Proc Am Thorac Soc 2008; 5: 173-178.

20. Farré R, Montserrat JM, BallesterE, Navajas D. Potential rebreathingafter continuous positive airway

pressure failure during sleep. Chest2002; 121: 196-200.

21. Mador MJ, Krauza M, Pervez A,Pierce D, Braun M. Effect of heatedhumidification on compliance andquality of life in patients with sleepapnea using nasal continuouspositive airway pressure. Chest2005;128: 2151-2158.

22. Santamaria J, Iranzo A, MontserratJM, de Pablo J. Persistentsleepiness in CPAP treatedobstructive sleep apnea patients:evaluation and treatment. SleepMed Rev 2007; 11: 195-207.

23. Redline S, Budhiraja R, Kapur V, etal. The scoring of respiratoryevents in sleep: reliability andvalidity. J Clin Sleep Med 2007; 3:169-200.

24. Rigau J, Montserrat JM, Wohrle H,et al. Bench model to simulateupper airway obstruction foranalyzing automatic continuouspositive airway pressure devices.Chest 2006; 130: 350-361.

25. Brown LK. Autotitrating CPAP: howshall we judge safety and efficacyof a “black box“? Chest 2006; 130:312-314.

26. Ayas NT, Patel SR, Malhotra A, etal. Auto-titrating versus standardcontinuous positive airway pressurefor the treatment of obstructivesleep apnea: results of a meta-analysis. Sleep 2004; 27: 249-253.

27. Nolan GM, Ryan S, O’Connor TM,McNicholas WT. 2006. Comparisonof three auto-adjusting positivepressure devices in patients withsleep apnoea. Eur Respir J 2006;28: 159-164.

28. Meurice JC, Cornette A, Philip-JoetF, et al. Evaluation of autoCPAPdevices in home treatment of sleepapnea/hypopnea syndrome. SleepMed 2007; 8: 695-703.

29. Masa JF, Jimenez A, Duran J, et al.Alternative methods of titratingcontinuous positive airwaypressure: a large multicenter study.Am J Respir Crit Care Med 2004;170: 1218-1224.

30. Mulgrew AT, Fox N, Ayas NT, RyanCF. Diagnosis and initialmanagement of obstructive sleepapnea without polysomnography.Ann Int Med 2007; 146: 157-166.

31. Rodenstein D. Determination oftherapeutic continuous positiveairway pressure for obstructivesleep apnea using automatictitration: promises not fulfilled.Chest 2008; 133: 595-597.

REFERENCES continued








Doc 10 –

CPAP Benefits

DOI 10.1378/chest.109.6.1470 1996;109;1470-1476Chest

Crichton F. Ramsay, Ian J. Deary and Neil J. DouglasHeather M. Engleman, Nima Asgari-Jirhandeh, Andrew L. McLeod, of CPAP Therapy : A Patient SurveySelf-Reported Use of CPAP and Benefits

http://chestjournal.chestpubs.org/content/109/6/1470

and services can be found online on the World Wide Web at: The online version of this article, along with updated information

ISSN:0012-3692)http://chestjournal.chestpubs.org/site/misc/reprints.xhtml(

the prior written permission of the copyright holder.of this article or PDF may be reproduced or distributed without Dundee Road, Northbrook, IL 60062. All rights reserved. No part1996 by the American College of Chest Physicians, 3300

CopyrightPhysicians. It has been published monthly since 1935. is the official journal of the American College of ChestCHEST

1996 BY THE AMERICAN COLLEGE OF CHEST PHYSICIANS by guest on July 5, 2010chestjournal.chestpubs.orgDownloaded from


http://chestjournal.chestpubs.org/site/misc/reprints.xhtml

http://chestjournal.chestpubs.org/

Self-Reported Use of CPAP and Benefitsof CPAP Therapy*A Patient Survey1Heather M. Engleman, BSc; Nima Asgari-Jirhandeh, BSc;Andrew L. McLeod, MBChB; Crichton F. Ramsay, MRCP;Ian /. Deary, PhD; and Neil J. Douglas, MD

The benefits of continuous positive airway pressure (CPAP) therapy in patients with the sleep ap¬nea/hypopnea syndrome (SAHS) are poorly documented and patients use CPAP less than physiciansrecommend. To establish patients' perceptions of benefit from CPAP and to identify determinantsof CPAP use, 204 CPAP users completed a questionnaire relating to use of CPAP therapy, sleepi¬ness, and road traffic incident rate before and after CPAP, perceived change in daytime functionand nocturnal symptoms with treatment, and problems with CPAP. Variables from these domainsof interest were examined, reduced through principal components analysis and correlated to assess

associations between these and polysomnographic measures of illness severity. Self-reported CPAPuse averaged 5.8±SD 2 h a night. Subjective sleepiness rated by the Epworth sleepiness scale androad traffic incident rate were significantly reduced by CPAP (p<0.0001). A broad range of functionand symptom items were highly significantly improved with CPAP (p<0.0001), corroborating the costto community and industry from SAHS and the preventive value ofCPAP. Road traffic incident ratebefore treatment was correlated with pre-CPAP sleepiness and SAHS severity. Subjective CPAP use

correlated with sleepiness before treatment but not with SAHS severity. CPAP mask problems andside effects were not associated with reduced CPAP use, but "nuisance" complaints of awakenings,noise, and sore eyes from CPAP correlated negatively with reported use. Greater reported CPAPuse was associated with better resolution of sleepiness and greater improvement in daytime func¬tion and nocturnalsymptoms. (CHEST 1996; 109:1470-76)

Key words: automobile accidents; compliance; CPAP; sleep apnea/hypopnea syndromeAbbreviations: AHI=apnea+hypopnea index; CPAP=continuous positive airway pressure; MSLT=multiple sleep latencytest; SAHS=sleep apnea/hypopnea syndrome; SNSL=Scottish National Sleep Laboratory

T^Tasal-contmuous positive airway pressure (CPAP)-*- ^ therapy is the treatment of choice for the sleepapnea/hypopnea syndrome (SAHS) and related disor¬ders. CPAP is effective in reducing nocturnal events of

For editorial comment see page 1416

SAHS and may improve objective daytime sleepi¬ness,1"4 cognitive function,1"3 and well-being.1,2 YetCPAP is frequently rejected by patients,5,6 at leastpartly because of the unwieldy and inconvenientnature of the treatment. Patients' use ofCPAP is likelyto be determined by perceived benefits and drawbacks

*From the Department of Respiratory Medicine (Ms. Englemanand Mrsr. Asgari-Jirhandeh, McLeod, Ramsey, and Douglas) andDepartment of Psychology (Dr. Deary), University of Edinburgh,Edinburgh, UK.'Copies oi the survey are available from the authors by contactingHeather Engleman, Scottish National Sleep Laboratory, ScottishInfirmary ofEdinburgh, Lauriston Place, Edinburgh EH3 9YW,UK; phone (+44)131 536-2355; fax (+44)131 536-3255.Supported by a grant from the British Lung Foundation (H.M.Engleman).Manuscript received July 5, 1995; revision accepted November 29.

oftreatment, but the composition ofthese factors is notwell understood.

Studies of CPAP use, its determinants, and effectsare highly variable in terms of patient selection andoutcome measures employed, and thus in resultsreported.1"20 Studies based primarily on objectivemeasures of function may neglect patient assessmentof benefit.1'4,7 Those studies using self-reported mea-

sures5,6,18"20 usually examine limited areas of function,often sleepiness alone, while open-ended question¬naire formats12 restrict comparability between studies.With two exceptions,15,20 studies ofthe effects ofCPAPhave been performed in small patient samples of lessthan 100 patients.

Average CPAP use rate varies from 3.2 to 6.7 h a

night, depending on whether new CPAP users,1'8,916cross-sectional CPAP clinic populations,15'18,20 or se¬

lected long-term acceptors of CPAP14,17,19 are studied.The literature on the determinants of CPAP compli¬ance and acceptance is contradictory, with CPAP use

predicted by polysomnographic severity in some6'18,19

1470 Clinical Investigations



and not in other studies,5,8,9,12,1416 by prior sleepinessin some5,6 but not in others,8,9,12,18 and complianceadversely affected by side effects in some8,12,18 but notin other studies.5,9,11,14,20A particular area of interest is the effect ofCPAP on

driving competence. The road traffic accident rate inSAHS is increased by a factor oftwo to seven times thatof the normal population,21,22 and a report has relateddriving accident rate to SAHS severity.23 A proportionof accidents in SAHS patients will be sleep related,causing more fatalities than other accidents.24 Labo¬ratory-based studies, whether using monotonous driv¬ing-based vigilance tasks1,13 or more realistic simula¬tors,25 suggest improved driving performance afterCPAP1,13 and uvulopalatopharyngoplasty.25 Previoussmall studies, conducted with 222^ and 1427 SAHSpatients, respectively, have indicated fewer reportedroad traffic incidents following CPAP.

This questionnaire-based study therefore assessedreported use of CPAP and a wide range of perceivedbenefits and drawbacks ofCPAP therapy in our clinicpopulation, so that these factors could be describedand relationships between them could be examined.

Materials and MethodsStudy DesignA questionnaire was sent in June 1994 to all patients issued CPAP

units by the Scottish National Sleep Laboratory (SNSL) for 2 weeksor longer. Questionnaire data were supplemented with information,obtained from SNSL records, on age, sex, polysomnographic SAHSseverity, objective CPAP use from run-time clock readings, andobjective daytime sleepiness on the multiple sleep latency test(MSLT).28 Posttreatment MSLTs were conducted after at least 4weeks of receiving CPAP. Information from the questionnaire andSNSL sources was grouped into domains of illness severity, CPAPcompliance, road traffic incidents and sleepiness before and afterCPAP, perceived change in function and symptoms, problems withCPAP use, and weight change.CPAP users underwent polysomnography before commencing

treatment. The criteria for prescribing CPAP were reported symp¬toms ofSAHS in association with an apnea + hypopnea index (AHI)of greater than 5/h sleep, or in association with snoring and recur¬

rent microarousals. Patients received practical demonstration andexperience during the daytime in the mechanisms and use ofCPAPand underwent a mask fitting before a night of CPAP titration.Telephone advice and appointments with nursing staff were avail¬able during office hours for patients experiencing problems.Patients were reviewed in an outpatient clinic 4 weeks after com¬mencement of therapy, when problems with CPAP were sought.Subsequent follow-up interval varied between 2 and 6 months de¬pending on whether problems were present.

QuestionnaireAll 253 patients issued a CPAP unit by the SNSL, and their

partners, were sent a four-page questionnaire inquiring about useof CPAP, sleepiness and road traffic incidents before and afterCPAP, changes in nocturnal and daytime function, problems withCPAP therapy, and weight change.

Self-Reported CPAP Use: Patients were asked how many nightsper week and for how long each night CPAP was used.

Epworth Sleepiness Score: Patients' subjective sleepiness afterand, retrospectively, before CPAP was rated by patients and theirpartners using the Epworth sleepiness scale.29,30Road Traffic Incidents: Drivers were asked their yearly mileage

and the frequency of road traffic incidents in the 5 years beforestarting CPAP and in the time since CPAP was commenced. Self-reported incidents were divided into near-misses, casualty-freecollisions ("minor" collisions), and accidents causing injury ("major"collisions) and further subdivided for those believed to be sleeprelated or not. The rates of road traffic incidents per 10,000 mileswere calculated for each class of event.

Function and Symptoms: Patients were asked to rate changes infunction and symptoms on a bipolar five-point scale with options ofmuch worse, worse, no change, better, and much better, coded-2,-l,0,+l, and +2, respectively. Items rated by patients were

snoring, breathing pauses, daytime sleepiness, sleep quality, tired¬ness, concentration ability, ability to drive long distances safely,work efficiency, time taken offwork, sex drive, and general health.Partners were asked to rate change in patients' snoring, breathingpauses, daytime sleepiness, and temper.

Problems With CPAP Use: Patients were presented with a

12-item list of side effects and problems with CPAP use, and askedto indicate on a four point-scale whether each problem was absent,a minor problem, a significant problem but not interfering withCPAP use, or a significant problem interfering with CPAP use. Theitems comprised nasal stuffiness, dry throat, red/sore eyes, leakingmask, cold airstream, nosebleeds, mask rubbing, difficulty exhaling,more frequent awakenings, excessive noisefrom CPAP unit, stom¬ach bloating/'flatulence, and chest wheeze.

Change in Weight: Patients were asked to report any weight gainor loss since the commencement of CPAP treatment.

Items not completed by or inapplicable to individuals were ex¬

cluded from relevant item analyses.Statistics

The significance ofinterindividual differences was assessed usingWilcoxon tests. Principal components analysis31 was conducted toreduce the number ofvariables for a rank correlation analysis, whichexamined associations between domains. All analyses were per¬formed using specific software (SPSS-PC+).32

Results

Questionnaire ResponseOf 253 patients (26 female) issued CPAP units, 215

(85%) returned questionnaires. Nonresponders were

significantly younger (mean [±SD] age, 46±9 years)than responders (53±10 years; p<0.0001), but wereotherwise no different from the responders, whohad a mean AHI33 of 47±38 per hour slept, 47±40microarousals34 per hour slept, average minimumoxygen saturation of 74±18%, and mean durationof CPAP treatment of 632 days (range, 16 to 2,921days).

Eleven patients (5% of responders) stated that theyno longer used CPAP. Three patients cited mask dis¬comfort as a factor, three cited lack ofbenefit, and one

each cited frequent awakenings, excessive CPAP pres¬sure, and throat dryness. One patient's nasal stuffiness,following nasal surgery, precluded CPAP use. Threepatients gave no reason for discontinuing treatment. OfCPAP users, 21 patients (10%) had an AHI less than15. The responses of the 204 patients (17 female) who

CHEST /109 / 6 / JUNE, 1996 1471



Table 1.Sleepiness Before and After CPAPPre-CPAP, Mean±SD Post-CPAP, Mean±SD p Value

Epworth sleepiness score (patient)Epworth sleepiness score (partner)MSLT, min

15±614±64.6±3.4

7±58±5

6.6±3.2

<0.0001<0.0001<0.001

Table 2.Mileage-Adjusted Road Traffic Incident Rates Before and After CPAPPre-CPAP, Mean±SD Post-CPAP, Mean±SD p Value

Incident rate (per 10,000 miles)Near-miss 0.92±2.96

Minor 0.09±0.44Major 0.005±0.027

Total incidents 1.02±3.17Sleep-related incident rate (per 10,000 miles)

Near-miss 0.86±2.94Minor 0.07±0.43

Major 0.003±0.021Total incidents 0.93±3.15

0.32±1.530.09±0.52

0.001±0.0150.41±1.63

0.11 ±0.630.03±0.20

00.14±0.68

<0.0001>0.3>0.20.0001

<0.0001>0.2>0.1<0.0001

indicated that they were continuing with CPAP ther¬apy were analyzed.

Self-Reported and Objective CPAP Use

Self-reported compliance in 204 CPAP users aver¬

aged 5.8±2.0 h per night, ranging from 0.1 to 9.5 h pernight. Synchronous CPAP run-time clock readings,available in 62 patients, yielded an average objectiveCPAP use of 5.1 ±2.5 h per night, significantly lowerthan that reported by the same patients (6.0± 1.9 h pernight; p=0.0003). Subjective and objective compliancedata were significantly correlated (r=0.68; p<0.0001).

¦ Sleep-related incidents

0 All incidents

Before AfterNEAR-MISSES

Before AfterMINOR

COLLISIONS

Before AfterMAJOR

COLLISIONS

Figure 1. Prevalence of road traffic incidents before and afterCPAP.

Change in Subjective and Objective Sleepiness WithCPAP

Patients' sleepiness, whether subjectively rated bypatient or partner, was significantly improved withCPAP, as was objective daytime sleepiness assessed byMSLT (Table 1). Pre-CPAP MSLT datawere availablein 52 patients and post-MSLT data were available in 41patients. Pre-CPAP scores on Epworth scale andMSLT correlated significantly (r=-0.38; p=0.01), butpost-CPAP scores for the two measures of sleepinessdid not (r=0.06; p>0.3).Changes in Road Traffic Incidents With CPAP

Information on road traffic incidents was obtainedfrom 147 driving patients. Sleep-related near-miss in¬cidents, unadjusted for time receiving CPAP therapy,were reported by39% and 5% ofpatients, respectively,before and after therapy (Fig 1). No sleep-relatedmajor collisions were reported after the commence¬

ment ofCPAP treatment. Mileage- and time-adjustedroad traffic incident rates showed a significant reduc¬tion in the rate of near-miss incidents after CPAPtherapy (Table 2).Change in Function and Symptoms With CPAP

All items relating to function and symptoms, ratedby patients and partners (Table 3), showed highly sig¬nificant improvements with CPAP, except sex drive.

Problems With CPAPPatients' reports of problems with CPAP use are

shown in Table 4. No life-threatening complications,such as meningitis, pneumoencephaly, or pneumo¬thorax, were seen.




Change in WeightReported weight rose significantly but trivially from

treatment commencement (mean gain, 1±8 kg;p=0.005). Most patients (55%) reported no change inweight.

Principal Components AnalysisPrincipal components analysis was conducted to

examine the structure of intercorrelations betweenresponses within the function/symptom items andCPAP-related problem items, respectively. By speci¬fying patterns of similarities in responses on the mul¬tiple items within these two domains, this techniqueallows cognate variables, all associating significantlywith an underlying component, to be identified. Itemsloading significantly on a component can then beconsolidated into a summary score, thus allowing thenumber of variables for subsequent correlation to berationally reduced.

Items with excessively skewed distributions (nose¬bleeds, wheezing, sore eyes, difficulty exhaling) or withreduced sample sizes (sex drive, bloating, work effi¬ciency, days taken off work, ability to drive longdistances safely) were excluded from the principalcomponents analyses. Components with eigenvaluesgreater than 1 were extracted and rotated, to increaseinterpretability, using the varimax method. Signifi¬cance for variable factor loadings was set at 0.30.The items relating to change in function and symp¬

toms with CPAP reduced to two rotated components,the first having significant loadings on tired/sleepquality/general health/concentration ability/excessivedaytime sleepiness (called "daytime function") and thesecond on snoring/breathing pauses (called "nocturnalsymptoms"). All seven items in this analysis also hadhigh loadings on the first unrotated principal compo¬nent. This indicated that, in addition to the two clearlyseparable components, daytime function and noctur¬nal symptoms, the total score from the seven itemscould be used as a "general function" measure.

Principal components analysis revealed that prob¬lems with CPAP use formed three rotated compo¬nents, with significant loadings on frequent awaken¬ings/noise/sore eyes (called "nuisance"), leaking mask/mask rubbing (called "mask problems"), and drythroat/nasal stuffiness (called "side effects").

Scores for created variables named daytime func¬tion, nocturnal symptoms, general function, nuisance,mask problems, and side effects were constructed bysumming item scores loading on each of these com¬

ponents. General function, nocturnal symptoms, anddaytime function were all highly significantly improvedwith CPAP (p<0.0001), with 95%, 95%, and 91% ofpatients, respectively, reporting improvement on eachof these summary scores. Nuisance, mask problems,

Table 3.Change in Function and SymptomsWith CPAP

Measure

PercentageReporting

ImprovementChange in Score,

Mean±SD

Patient ratingBreathing pausesSnoringDaytime sleepinessSleep qualityTirednessAbility to drive long

distances safelyConcentrationWork efficiencyGeneral healthTime taken off workSex drive

Partner ratingSnoringBreathing pausesDaytime sleepinessTemper

949284817977

6866613222

95907949

1.6±0.6*1.6±0.7*1.3±0.8*1.2±0.9*1.0±0.8*1.3±0.9*

0.9±0.9*0.9±0.9*0.8±0.9*0.5±0.8*0.1±0.9

1.6±0.7*1.4±0.8*1.1±0.9*0.6±1.0*

*p<0.0001.

and side effects were rated as present in some degreeby 66%, 72%, and 73% of patients, respectively.Rank Correlation

Putative predictive associations among domains ofillness severity, sleepiness, road traffic incident rates,change in symptoms and function, and problems withCPAP use were assessed with rank correlation (Table5). Subjective CPAP use was not significantly corre¬

lated with any objective index of severity of SAHS, butwas positively correlated with pre-treatment Epworthsleepiness score and negatively correlated with thedegree of nuisance of CPAP therapy reported. CPAPnuisance was negatively correlated with SAHS sever¬

ity. Improvements in daytime function and nocturnalsymptoms correlated with baseline Epworth sleepinessTable 4.Percentage of Patients Reporting Problems

With CPAP Use

Percentage Reporting Percentage ReportingProblem Severe Problem

Nasal stuffiness 64Mask leak 63Dry throat 62Cold airstream45Noise from CPAP unit 41Mask rubbing 41Bloating/flatulence 37More frequent 32

awakeningsRed/sore eyes 31Chest wheeze 21Difficulty exhaling 18Nosebleeds 10

4<11221

<12

1<110

CHEST / 109 / 6 / JUNE, 1996 1473



Table 5.Rank Correlation Among Polysomnography, CPAP Use, Sleepiness, Changes in Function and Symptoms,Road Traffic Incidents, and CPAP Programs*

AHIAROUS- SUBJ- PRE- POST¬ALS MIN02 USE ESS ESS

NUIS- MASK SIDEANCE PROB EFF BNMSLP BNMNON BMINSLP BMINNON

AHIAROUSALSMINO2SUBJUSEPBE-ESSPOST-ESSNUISANCEMASK PROBSIDE EFFdDAYFUNCdNOCTSYMPdGENFUNC

X0.61f

-047f0.040.11

-0.11-0.13*-0.030.070.030.130.09

XX

-0.27f0.000.13

-0.03-0.12-0.060.20§

-0.020.110.04

xxX

-0.08-0.29*0.20$

-0.040.000.01

-0.21*-0.05

xXxX0.22$

-0.18$-0.15*0.17$

-0.070.4410.35f

XXXXX0.11

-0.030.090.060.20$0.13*

XXXXXX0.11

-0.070.11

-0.43*-0.20$

XXXXXXX0.28*0.19$

-0.17*-0.22$

XXXXXXXX

0.050.100.12

-0.19* 0.46* 0.1 -0.44* -0.20$ 0.09

XXXXXXXXX

-0.10-0.07-0.09

0.150.20*

-0.25$

0.54*

0.070.040.12

-0.12

0.060.08

-0.10

0.29*

-0.01-0.020.10

-0.27f

*Cell sample size varies from 117 to 203 patients; p values are adjusted accordingly; X=redundant cell; short horizontal line (.)=nonpredictive cell.AROUSALS=microarousal index; MIN02=minimum oxygen saturation; SUBJUSE=subjective CPAP use; PRE-ESS=pre-CPAP Epworth sleepinessscore; POST-ESS=post-CPAP Epworth sleepiness score; NUISANCE=nuisance-type problems with CPAP use; MASK PROBS=mask-relatedproblems with CPAP use; SIDE EFF=CPAP side effects; dDAYFUNC^change in daytime function; dNOCTSYMP=change in nocturnal symptoms;dGENFUNC=change in general function/symptoms; BNMSLP=sleep-related near-miss incidents before CPAP; BNMNON=nonsleep-relatednear-miss incidents before CPAP; BMINSLP=sleep-related minor collisions before CPAP; BMINNON=sleep-related minor collisions before CPAP.

fp=0.001.*p=0.05.*p=0.01.

score and reported CPAP use, and negatively withEpworth sleepiness score on CPAP. The frequency ofdriving incidents before treatment was correlated withEpworth sleepiness score, the frequency ofmicroarou¬sals, and the extent of nocturnal hypoxemia.

DiscussionThis study documents experience and perceptions

of CPAP in a large sample of unselected CPAP users

with a wide range of illness severity. Although neces¬

sarily limited by its use of mainly self-reported andretrospective information, the study provides evidenceof patient-perceived, CPAP-induced improvementacross a wide range of function, including sleepiness,driving competence, cognitive function, work effi¬ciency, well-being, and nocturnal symptoms. Further¬more, coherent correlations linked pre-CPAP drivingcompetence, use of CPAP, and benefit from CPAP to

predictive variables.The study is based on retrospective, self-reports

from CPAP users, which may be compromised by poormemory of past events occurring prior to treatment or

by a tendency to overreport improvements, havingpsychologically "invested" in CPAP therapy. Despitethese drawbacks, retrospective ratings ofpretreatmentstatus may provide counterbalancing benefits. A lack ofawareness of impairment before treatment, noted byothers,16,35 may result from the absence of a normalframe of reference in constantly sleepy, untreated pa¬tients. In these, retrospective measures may be more

reliable. The reversal of sleepiness and driving impair¬

ment with treatment may also encourage greaterfrankness on the part of patients regarding their pre¬vious deficits, who need no longer fear being bannedfrom driving. The results of this study, showing wide-ranging improvements in daytime function and noc¬

turnal symptoms, are unlikely to be due only to retro¬

spective inaccuracies and placebo-like effects, havingbeen shown also in a prospective, randomized, place¬bo-controlled study of the effects of CPAP.1CPAP Use

Self-reported CPAP use was significantly associatedwith outcome measures assessing posttreatment sleep¬iness and improvement in function and symptoms,confirming and extending findings of others showingcorrelations between compliance and subjectivechange in sleepiness with treatment.18'19 The observedtriangular association among pre-CPAP sleepiness,subsequent compliance, and posttreatment sleepiness(Table 5) is suggestive of a positively reinforcing loop.Poorer compliance was associated with greater rates ofsleep-related collisions after treatment, providing ad¬ditional corroboration of the benefits of CPAP.Mean objective CPAP compliance demonstrated in

this study (5.1 h a night) was closely in agreement withthat of others in cross-sectional CPAP clinic se¬

ries,15'18,20 which included both new and long-termusers. Patients overestimated CPAP use by around 1 hper night, a finding that is consistent across this andother studies.918,20 Degree of CPAP compliance was

linked only to prior sleepiness and not to illness sever-




ity, confirming the hypothesis that sleepiness is theprimary determinant of CPAP acceptance.5'6 Apartfrom the nuisance problem factor (see below), prob¬lems with CPAP use were not associated with reducedcompliance.Sleepiness

Subjective sleepiness, assessed by the Epworthsleepiness scale, was significantlyimprovedwith CPAP,with average scores after treatment falling within thenormal range.30 The Epworth sleepiness scale, al¬though subjective and completed retrospectively, maybe a relatively robust measure of sleepiness, dealingwith memorable behavior of napping rather thantransient mood states of sleepiness. Sleep onset latencyon the MSLT was improved but was not normalized byCPAP. This observation has been reported previouslyby ourselves and others,1"3'7'810 with only one studyshowing normalization of sleep onset latency withCPAP.4 The small magnitude of change in objectivedaytime sleepiness, with posttreatment scores lying inthe range associated with moderate sleepiness,28 mayindicate only partial resolution of sleepiness withCPAP. Alternatively, it may reflect insensitivity of theMSLT to treatment-induced changes in sleepiness, as

has been suggested by others.10

Road Traffic IncidentsThe survey documents a high prevalence of sleep-

related road traffic incidents in untreated patients, with39% of all driving patients being aware of sleep-relatednear-miss incidents before treatment (Fig 1). Theseresults are compatible with the findings of others ofincreased accident rates in SAHS patients.2122 Self-reported mileage-adjusted rates of near-miss incidentswere significantly improved after CPAP. Thus, thestudy shows significant improvement in actual drivingcompetence with CPAP, consistent with studies sug¬gesting that treatment may improve driving skills on

simulators.1,13'25 These findings confirm previoussmaller-scale reports of lowered driving incidents fol¬lowing CPAP.26'27 Although no significant differencein sleep-related collision rate after CPAP was ob¬served, the low frequency of such events before CPAPin a small population (Fig 1) may contribute to thisfinding. Road traffic incidents before treatment were

significantly correlated with sleepiness and polysom¬nographic measures of sleep fragmentation and hy¬poxemia. These findings of putative predictors fordriving competence in both sleepiness and illness se¬

verity in SAHS extend those of Findley et al.23

Function and SymptomsCPAP-treated patients reported highly significant

improvements in all symptom and function items, ex¬

cept sex drive. The high frequency of reported im¬provements in daytime function items, especially thoserelating to concentration, work efficiency, absencefrom work, and ability to drive distances safely, sug¬gests that these areas of function are compromised ina significant proportion of patients with SAHS (Table4). Together with the data on road traffic incidents, theabove findings suggest a high cost to community andindustry from SAHS and a substantial preventive valuefor CPAP. The magnitude ofreported improvement indaytime function and nocturnal symptoms was relatedto severity of initial illness. Greater reported improve¬ments in daytime function and in nocturnal symptomswere associated with greater reported CPAP use,greater sleepiness before treatment, and lesser sleep¬iness after treatment.

Problems With CPAP Use

Reported problems with CPAP use, which most

patients classified as "minor" in nature, were remark¬ably frequent, despite intervention during patient fol¬low-up. Problems with CPAP use have been associatedpreviously with reduced compliance by ourselves andothers,818 but significant relationships between prob¬lems and reported CPAP use were limited to the nui¬sance problem complex. In contrast to nuisance prob¬lems, mask problem and side effect scores were notassociated with lower SAHS severity, reported CPAPuse, or satisfaction with treatment.

Nuisance ProblemsThe nuisance complex, describing complaints relat¬

ing to noise, frequent awakenings, and sore eyes withCPAP treatment, exhibited an interesting pattern ofassociation with putative determinants and effects(Table 5). This problem complexwas weakly correlatedwith milder polysomnographic illness, lower subse¬quent CPAP use, and lesser perceived benefit. One ofthe nuisance complex items, noise from CPAP units,has been associated with lower SAHS severity,20 butnot previously with lesser CPAP compliance.

Although high scorers for nuisance problems hadmilder initial illness and poorer subsequent CPAPcompliance, recent research suggests that even pa¬tients with milder indexes of illness severity receiveobjective benefits for cognitive function from CPAP.1Thus patients' unawareness of the benefits ofCPAP isinsufficient justification for withholding treatment.The lack of correlation between polysomnographicindexes of illness severity and CPAP use confirms thevalue ofCPAP therapy in "heavy snorers disease"36 or

"upper airway resistance syndrome"37 as well as SAHS.It may be that patient education can aid insight intoillness-induced impairment and thus promote im¬proved compliance with and benefit from CPAP.

CHEST/109/6/JUNE, 1996 1475



ACKNOWLEDGMENT: We thank the nursing, technical, andadministrative staff of the SNSL for their contributions to thisproject.

References1 Engleman HM, Martin SE, Deary IJ, et al. Effect of continuous

positive airway treatment on daytime function in sleep apnoea/hypopnoea syndrome. Lancet 1994; 343:572-75

2 Kribbs NB, Pack AI, Kline LR, et al. Effects of one night with¬out nasal CPAP treatment on sleep and sleepiness in patients withobstructive sleep apnea. Am Rev Respir Dis 1993; 147:1162-68

3 Bedard M-A, Montplaisir J, Malo J, et al. Persistent neuropsy¬chological deficits and vigilance impairment in sleep apnea syn¬drome after treatment with continuous positive airways pressure(CPAP). J Clin Exp Neuropsychol 1993; 15:330-41

"

4 Lamphere J, Roehrs T, Wittig R, et al. Recovery of alertness af¬ter CPAP in apnea. Chest 1989; 96:1364-67

5 Waldhorn RE, Herrick TW, Nguyen MC, et al. Long-termcompliance with nasal continuous positive airway pressure ther¬apy of obstructive sleep apnea. Chest 1990; 97:33-8

6 Rauscher H, Popp W, Wanke T, et al. Acceptance of CPAPtherapy for sleep apnea. Chest 1991; 100:1019-23

7 Engleman HM, Cheshire KE, Deary IJ, et al. Daytime sleepiness,cognitive performance and mood after CPAP therapy for thesleep apnoea hypopnoea syndrome. Thorax 1993; 48:911-14

8 Engleman HM, Martin SE, Douglas NJ. Compliance with CPAPtherapy in patients with the sleep apnoea/hypopnoea syndrome.Thorax 1994; 49:263-66

9 Kribbs NB, Pack AI, Kline LR, et al. Objective measurement ofpatterns of nasal CPAP use by patients with obstructive sleepapnea. Am Rev Respir Dis 1993; 147:887-95

10 Sangal RB, Thomas L, Mitler MM. Disorders of excessive

sleepiness: treatment improves ability to stay awake but does notreduce sleepiness. Chest 1992; 102:699-703

11 Nino-Murcia G, McCann CC, Bliwise DL, et al. Compliance andside effects in sleep apnea patients treated with nasal continuous

positive airway pressure. West J Med 1989; 150:165-6912 Hoffstein V, Viner S, Mateika, et al. Treatment of obstructive

sleep apnea with nasal continuous positive airway pressure:patient compliance, perception of benefits, and side effects. AmRev Respir Dis 1992; 145:841-45

13 Findley LJ, Fabrizio MJ, Knight H, et al. Driving simulator per¬formance in patients with sleep apnea. Am Rev Respir Dis 1989;140:529-30

14 Fletcher EC, Luckett RA. The effect ofpositive reinforcement onhourly compliance in nasal continuous positive airway pressureusers with obstructive sleep apnea. Am Rev Respir Dis 1991;143:936-41

15 Krieger J. Long-term compliance with nasal continuous positiveairway pressure (CPAP) in obstructive sleep apnea patients andnonapneic snorers. Sleep 1992; 15:S42-46

16 Reeves-Hoche MK, Meek R, Zwillich CW. Nasal CPAP: an ob¬jective evaluation of patient compliance. Am J Respir Crit CareMed 1994; 149:149-54

17 Fleury B, Rakatonanahary D, Tehindrazanarivelo AD, et al.Long-term compliance to continuous positive airway pressure

therapy (nCPAP) set up during a split-night polysomnography.Sleep 1994; 17:512-15

18 Rauscher H, Formanek D, Popp W, et al. Self-reported vs mea¬

sured compliance with nasal CPAP for obstructive sleep apnea.Chest 1993; 103:1675-80

19 Meurice J-C, Dore P, Paquereau J, et al. Predictive factors oflong-term compliance with nasal continuous positive airwaypressure treatment in sleep apnea syndrome. Chest 1994;105:429-33

20 Pepin JL, Leger P, Veale D, et al. Side effects of nasal continu¬ous positive airwaypressure in sleep apnea syndrome: study of 193patients in two French sleep centers. Chest 1995; 107:375-81

21 George CF, Nickerson PW, Hanly PJ, et al. Sleep apnoea patientshave more automobile accidents [letter]. Lancet 1987; 2:447

22 Findley LJ, Unverzagt ME, Surratt PM. Automobile accidentsinvolving patients with obstructive sleep apnea. Am Rev RespirDis 1988; 138:337-40

23 Findley LJ, Fabrizio M, Thommi G, et al. Severity of sleep ap¬nea and automobile accidents. N Engl J Med 1989; 320:868-69

24 Home JA, Reyner LA. Sleep related accidents. BMJ 1995;310:565-67

25 Haraldsson P-O, Carenfelt C, Persson HE, et al. Simulated long-term driving performance before and after uvulopalatopharyn¬goplasty. ORL 1991; 53:106-10

26 Suratt P, Findley L. Effect of nasal CPAP treatment on automo¬bile driving simulator performance and on self-reported auto¬mobile accidents in subjects with sleep apnea. Am Rev Respir Dis1992; 145:A169

27 Minemura H, Akashiba T, Yamamoto H, et al. Traffic accidentsin obstructive sleep apnea patients and effects of nasal CPAPtreatment. Jpn J Thorac Med 1993; 31:1103-08

28 Thorpy MJ. The clinical use of the multiple sleep latency test.

Sleep 1992; 15:268-7629 Johns MW. A new method for measuring daytime sleepiness: the

Epworth sleepiness scale. Sleep 1991; 14:540-4530 Johns MW. Reliability and factor analysis of the Epworth sleep¬

iness scale. Sleep 1992; 15:376-8131 Childs D. Essentials of factor analysis, 2nd ed. London: Cassell,

199032 Norusis MJ/SPSS Inc. SPSS/PC+ advanced statistics V2.0 guide.

Chicago: SPSS Inc, 198833 Gould GA, Whyte KF, Rhind GB, et al. The sleep hypopnea

syndrome. Am Rev Respir Dis 1988; 137:895-9834 Cheshire K, Engleman H, Deary I, et al. Factors impairing day¬

time performance in patients with sleep apnea hypopnea syn¬drome. Arch Intern Med 1992; 152:538-41

35 Dement WC, Carskadon MA, Richardson G. Excessive daytimesleepiness in the sleep apnea syndrome. In: Guilleminault C,Dement WC, eds. Sleep apnea syndromes. New York: Alan R.Liss, 1978; 23-46

36 Lugaresi E, Coccagna G, Cirignotta F. Snoring and its clinicalimplications. In: Guilleminault C, Dement WC, eds. Sleep apneasyndromes. New York: Alan R. Liss, 1978; 13-21

37 Guilleminault C, Stoohs R, Clerk A, et al. A cause of excessive

daytime sleepiness: the upper airway resistance syndrome. Chest1993; 104:781-87




DOI 10.1378/chest.109.6.1470 1996;109; 1470-1476Chest

Crichton F. Ramsay, Ian J. Deary and Neil J. DouglasHeather M. Engleman, Nima Asgari-Jirhandeh, Andrew L. McLeod,

Patient SurveySelf-Reported Use of CPAP and Benefits of CPAP Therapy : A

July 5, 2010This information is current as of

http://chestjournal.chestpubs.org/content/109/6/1470Updated Information and services can be found at:

Updated Information & Services

Citations

470#related-urlshttp://chestjournal.chestpubs.org/content/109/6/1articles:This article has been cited by 42 HighWire-hosted

Cited Bys

http://www.chestpubs.org/site/misc/reprints.xhtmlbe found online at: Information about reproducing this article in parts (figures, tables) or in its entirety canPermissions & Licensing

http://www.chestpubs.org/site/misc/reprints.xhtmlInformation about ordering reprints can be found online:

Reprints

"Services" link to the right of the online article.Receive free e-mail alerts when new articles cite this article. To sign up, select the

Citation Alerts

PowerPoint slide format. See any online figure for directions. articles can be downloaded for teaching purposes inCHESTFigures that appear in Images in PowerPoint format



http://chestjournal.chestpubs.org/content/109/6/1470#related-urls

http://www.chestpubs.org/site/misc/reprints.xhtml









Doc 11 –

Noninvasive Ventilation for Critical Care

DOI 10.1378/chest.06-2643 2007;132;711-720Chest

Erik Garpestad, John Brennan and Nicholas S. Hill

*Noninvasive Ventilation for Critical Care

http://chestjournal.chestpubs.org/content/132/2/711.full.html


ISSN:0012-3692)http://chestjournal.chestpubs.org/site/misc/reprints.xhtml(

of the copyright holder.may be reproduced or distributed without the prior written permission Northbrook, IL 60062. All rights reserved. No part of this article or PDFby the American College of Chest Physicians, 3300 Dundee Road,

2007Physicians. It has been published monthly since 1935. Copyright CHEST is the official journal of the American College of Chest

© 2007 American College of Chest Physicians by guest on April 13, 2010chestjournal.chestpubs.orgDownloaded from


http://chestjournal.chestpubs.org/site/misc/reprints.xhtml


Noninvasive Ventilation for CriticalCare*

Erik Garpestad, MD, FCCP; John Brennan, MD; andNicholas S. Hill, MD, FCCP

Noninvasive ventilation (NIV), the provision of ventilatory assistance without an artificial airway,has emerged as an important ventilatory modality in critical care. This has been fueled byevidence demonstrating improved outcomes in patients with respiratory failure due to COPDexacerbations, acute cardiogenic pulmonary edema, or immunocompromised states, and whenNIV is used to facilitate extubation in COPD patients with failed spontaneous breathing trials.Numerous other applications are supported by weaker evidence. A trial of NIV is justified inpatients with acute respiratory failure due to asthma exacerbations and postoperative states,extubation failure, hypoxemic respiratory failure, or a do-not-intubate status. Patients must becarefully selected according to available guidelines and clinical judgment, taking into accountrisk factors for NIV failure. Patients begun on NIV should be monitored closely in an ICU or othersuitable setting until adequately stabilized, paying attention not only to vital signs and gasexchange, but also to comfort and tolerance. Patients not having a favorable initial response toNIV should be considered for intubation without delay. NIV is currently used in only a selectminority of patients with acute respiratory failure, but with technical advances and new evidenceon its proper application, this role is likely to further expand. (CHEST 2007; 132:711–720)

Key words: acute respiratory failure; COPD; mechanical ventilation; noninvasive ventilation

Abbreviations: ALI � acute lung injury; APACHE � acute physiology and chronic health evaluation;CHF � congestive heart failure; CI � confidence interval; CPAP � continuous positive airway pressure;CPE � cardiogenic pulmonary edema; DNI � do not intubate; Fio2 � fraction of inspired oxygen; NIV � noninvasiveventilation; PEEP � positive end-expiratory pressure

O ne of the most important developments in thefield of mechanical ventilation over the past 15

years has been the emergence of noninvasive venti-lation (NIV) as an increasing part of the critical carearmamentarium. Although similar techniques such

as intermittent positive pressure breathing were usedwidely during previous decades, unlike NIV theywere used mainly to provide intermittent aerosoltherapy. The term NIV includes other forms ofventilatory assistance that avoid airway invasion, suchas negative pressure ventilation, but the vast majorityof NIV applications now use positive pressure. Non-invasive application of continuous positive airwaypressure (CPAP) will be considered a form of “NIV”here when used to treat certain types of respiratoryfailure, but it is not a “true” form of ventilatoryassistance because the positive pressure does notincrease intermittently to assist inspiration.

The emergence of NIV has been fueled by itsrelative ease of application compared to alternativeforms of noninvasive ventilation, as well as its dem-onstrated ability to improve patient outcomes incertain forms of acute respiratory failure comparedto previously standard therapy, including endotra-

*From the Division of Pulmonary, Critical Care, and SleepMedicine, Tufts-New England Medical Center, Boston, MA.Dr. Hill has received honoraria and research grants from andserved on the medical advisory boards of ResMed, Inc., andRespironics, Inc. He has received a research grant from Ver-samed, Inc. Drs. Garpestad and Brennan have no conflicts ofinterest to disclose.Manuscript received October 28, 2006; revision accepted March12, 2007.Reproduction of this article is prohibited without written permissionfrom the American College of Chest Physicians (www.chestjournal.org/misc/reprints.shtml).Correspondence to: Nicholas S. Hill, MD, FCCP, Division ofPulmonary, Critical Care and Sleep Medicine, Tufts-New En-gland Medical Center, 750 Washington St, Boston, MA 02111;e-mail: [email protected]: 10.1378/chest.06-2643

CHEST Postgraduate Education CornerCONTEMPORARY REVIEWS IN CRITICAL CARE MEDICINE

www.chestjournal.org CHEST / 132 / 2 / AUGUST, 2007 711



cheal intubation.1 This update will focus on recentdevelopments regarding acute applications of NIV,including the expanding evidence base, technicaladvances, and assessment of current utilization. Weemphasize techniques for proper patient selectionand implementation that are critical if success ratesreported in the literature are to be duplicated.

NIV for Acute Respiratory Failure

Recommended Indications

Many applications of NIV have been tried in thecritical care setting, but as of yet, only four aresupported by multiple randomized controlled trialsand metaanalyses.

COPD Exacerbations

The strongest level of evidence, including multiplerandomized controlled trials,2–7 supports the use ofNIV to treat exacerbations of COPD. Also, meta-analyses by Ram et al8 and Keenan et al9 demon-strate more rapid improvements in vital signs and gasexchange as well as reductions in the need forintubation (relative risk, 0.41; 95% confidence inter-val [CI], 0.33 to 0.53; risk reduction, 28%), de-creased mortality (relative risk, 0.52; 95% CI, 0.35 to0.76; risk reduction, 10%), and decreased hospitallength of stay (� 3.24 days; 95% CI, � 4.42 to� 2.06 days and � 4.57 days, respectively). TheCochrane analysis8 also noted more rapid improve-ments in vital signs, pH, and gas exchange, andreduced complication rates and hospital lengths ofstay. Based on these observations, NIV should nowbe considered the ventilatory modality of first choiceto treat acute respiratory failure caused by exacerba-tions of COPD.

Acute Cardiogenic Pulmonary Edema

Similarly strong evidence supports the use ofnoninvasive positive pressure techniques to treatacute cardiogenic pulmonary edema (CPE).10–17 Re-cent metaanalyses18–20 on the use of NIV to treatacute pulmonary edema have shown that both CPAPand NIV lower intubation and mortality rates com-pared to conventional therapy with oxygen, althoughthe reduction in mortality rate was statistically sig-nificant only in one of the metaanalyses.20. A ran-domized trial17 comparing CPAP directly to NIVshowed no difference in outcomes between the twoto treat CPE, a finding confirmed in a recentmetaanalysis by Ho and Wong.21 Accordingly, byvirtue of its greater simplicity and lesser expense,CPAP has been suggested as the initial noninvasive

choice for acute CPE. However, some studies22 haveobserved more rapid improvements in gas exchangeand vital signs with NIV than with CPAP alone, soNIV may be preferable for patients with persistingdyspnea or hypercapnia after initiation of CPAP.

Facilitating Extubation in COPD Patients

Another NIV application supported by multiplerandomized trials is to facilitate extubation in COPDpatients. Candidates for early extubation are thosewho were intubated for COPD exacerbations be-cause they were poor candidates for or failed NIVinitially and are unable to pass a T-piece trial eventhough they have improved sufficiently to tolerateNIV. Ferrer et al23 confirmed earlier findings ofNava et al24 in such patients, randomizing 43 patientswith “persistent” weaning failure (failure of threeconsecutive T-piece trials) to be extubated to NIV orweaned using conventional methods. They observedthat NIV-treated patients had shorter durations ofintubation (9.5 days vs 20.1 days) and ICU (14 daysvs 25 days) and hospital stays (14.6 days vs 40.8 days),decreased incidence of nosocomial pneumonia (24%vs 59%), and improved ICU and 90-day survivals(80% vs 50%) [all p � 0.05]. These studies stronglysupport the use of NIV to facilitate extubation inpatients with hypercapnic respiratory failure and toavoid the complications of prolonged intubation. Butit must be applied cautiously: only in patients whoare otherwise good candidates for NIV and were notdifficult intubations.

Immunocompromised Patients

The use of NIV is also well supported for immu-nocompromised patients who are at high risk forinfectious complications from endotracheal intuba-tion, such as those with hematologic malignancies,AIDS, or following solid-organ or bone marrowtransplants. In a randomized trial25 of patients withhypoxemic respiratory failure following solid-organtransplantation, NIV use decreased intubation rate(20% vs 70%, p � 0.002) and ICU mortality (20% vs50%, p � 0.05) compared with conventional therapywith oxygen. Hilbert et al26 observed fewer intuba-tions (46% vs 77%) and a lower mortality rate (50%vs 81%) [both p � 0.05] among immunocompro-mised patients (mainly hematologic malignancies,but some after solid-organ transplantation or withAIDS) with acute respiratory failure randomized toNIV as opposed to conventional therapy. The sizablereductions in mortality in these studies stronglysupport the early use of NIV as the initial ventilatorymodality in immunocompromised patients withacute respiratory failure, although morbidity andmortality rates are still likely to be high.

712 Postgraduate Education Corner



Conditions for Which NIV Should Be Considered

These applications are supported by a single ran-domized controlled trial, historically controlled trials,or multiple trials yielding conflicting evidence. NIVcan be tried if patients are selected and monitoredcarefully.

Asthma

Several uncontrolled series27–29 and one random-ized trial30 support the use of NIV for acute asthma.In the randomized trial,30 a pilot, 33 patients withacute dyspnea but not in respiratory failure wererandomized to standard therapy with nasal bilevelventilation for 3 h or standard therapy with shamNIV. NIV improved expiratory flow rates morerapidly (80% of patients had a � 50% increase inFEV1 in the first hour, compared to only 20% in thesham group) and reduced the need for hospitaliza-tion.30 The authors speculated that the positivepressure had a salutary effect on airway dilatation,although these results have yet to be replicated. Acase report31 raised concerns about the use of NIVfor asthma, and a recent metaanalysis by Ram et al32

concluded that routine use of NIV in severe acuteasthma could not be recommended. We believe thata cautious trial should still be considered in asthmat-ics not responding to the first hour of conventionaltherapy, but more study is warranted.

Postoperative Respiratory Failure

Either NIV or CPAP may be helpful in avertingpostoperative respiratory failure by preventing atel-ectasis and/or improving gas exchange as suggestedby three randomized controlled trials in patientsundergoing different surgical procedures. Followingthoracoabdominal aneurysm repair, prophylactic useof CPAP reduces overall pulmonary complications.33

Squadrone et al34 compared CPAP vs conventionaloxygen therapy in patients with hypoxemic respira-tory failure after major elective abdominal surgery.The CPAP group had lower rates of intubation,pneumonia, and sepsis. In the only randomized,controlled trial of NIV in postoperative patients,Auriant et al35 found that NIV reduced intubationand mortality rates in patients with hypoxemic respi-ratory failure following lung resection. These trialsindicate that either CPAP or NIV should be stronglyconsidered to prevent or treat postoperative respira-tory failure, mainly after lung resection in patientswith underlying COPD or congestive heart failure(CHF). Although multiple studies support this ap-plication, further studies need to focus on the use ofNIV following specific surgical procedures beforefirmer recommendations can be made.

Do-Not-Intubate Patients

Use of NIV for patients with acute respiratoryfailure who have a do-not-intubate (DNI) status hasaroused debate. Concerns have been raised that themodality might merely prolong the dying processwhile mask discomfort outweighs any palliative ef-fect.36 However, a prospective observational study byLevy et al37 showed that patients with reversiblediagnoses such as COPD and CHF had a better-than-even chance of surviving the hospitalization iftreated with NIV (52% and 75%, respectively),whereas those with pneumonia or cancer had muchlower likelihoods of hospital survival. Schettino etal38 reported similar findings in their observationalcohort and noted low success rates for NIV inpostextubation respiratory failure, hypoxemic respi-ratory failure, and end-stage cancer.

Some have proposed resolving the conflict aboutthe use of NIV in DNI patients by specifying thegoals of therapy.39 Patients with reversible processessuch as COPD exacerbations or CHF may wish tosurvive the acute illness and thus use NIV as a formof life support. They are willing to endure somediscomfort to achieve this aim. Others desire pallia-tion, aiming to alleviate dyspnea or briefly prolongsurvival to settle affairs. In these latter circum-stances, excessive mask discomfort would justifystopping therapy because the goal of palliation is notbeing met. Differentiating between and agreeing onthese aims requires close and effective communica-tion between caregivers, patient and family.

Hypoxemic Respiratory Failure

Randomized controlled trials suggest that patientswith hypoxemic respiratory failure (ie, severe respi-ratory distress, Pao2/fraction of inspired oxygen(Fio2) � 200 and a non-COPD cause for respiratoryfailure) benefit from use of NIV.40,41 In 105 suchpatients, Ferrer et al41 found that compared toconventional therapy, NIV reduced the intubationrate (52 to 25%), the incidence of septic shock (31 to12%) and ICU mortality (39 to 18%), and improved90-day mortality (all p � 0.05). Notably, almost onethird of the patients had CPE, which would beexpected to respond favorably to NIV, but patientswith pneumonia were the ones that benefited themost in this study.

This latter finding contrasts with earlier trials thatshowed an association between pneumonia and NIVfailure42–44 and the need for intubation in approxi-mately two thirds of patients with pneumonia treatedinitially with NIV.45 One randomized controlledtrial46 studying patients with severe community-acquired pneumonia showed that NIV improvedoutcomes including survival at 2 months but only in




patients with underlying COPD. In one prospectivecohort study,47 risk factors for NIV failure in patientswith acute hypoxemic respiratory failure includedthe diagnoses of ARDS or severe community-ac-quired pneumonia, more severe hypoxemia (Pao2/Fio2 � 146 after the first hour of treatment), andage � 40 years. In another study48 of NIV to treathypoxemic respiratory failure, shock, severe hypox-emia, and severe metabolic acidosis were associatedwith poor outcomes. In a prospective analysis49 ofselected patients with acute lung injury (ALI)/ARDStreated with NIV as the initial ventilator modality, asimplified acute physiology score II � 34 and aPao2/Fio2 � 175 after 1 h of therapy predictedfailure. The outcome of NIV is also very poor whenused to treat hypoxemic respiratory failure in pa-tients with idiopathic pulmonary fibrosis,48 and thisapplication should be discouraged.

A concern that must be stressed is that thediagnostic category of hypoxemic respiratory failureis very broad and benefits accruing to certain subsetsof patients within the larger diagnostic categorycould obscure adverse consequences in smaller sub-groups. Thus, with the exception of use for CPE,which is supported by strong evidence, NIV shouldbe used only with caution in carefully selectedpatients with hypoxemic respiratory failure, andthose at high risk for failure should be considered forearly intubation, especially if oxygenation fails toimprove substantially within the first hour or two(Fig 1).

Extubation Failure

The use of NIV to prevent or treat extubationfailure has also raised concerns. Respiratory failurefollowing extubation imparts a poor prognosis; theduration of mechanical ventilation is lengthened, thelikelihood of discharge to a chronic care facility isincreased, and mortality may reach 40%.50 Estebanet al51 evaluated the ability of NIV to avoid extuba-tion failure by randomizing patients to NIV orconventional therapy if risk factors developed, in-cluding hypercapnea, hypoxemia, acidemia, or tachy-pnea, after a routine extubation. Surprisingly, NIVnot only failed to lower the reintubation rate com-pared to conventional therapy (approximately 50% inboth groups) but it also increased ICU mortality.This was thought to be related to delayed reintuba-tions in the NIV group, an average of 10 h later thanin the conventional therapy group. The authors51

concluded that NIV is “not effective in averting theneed for reintubation in unselected patients in whomrespiratory failure develops after extubation” andthat it “may in fact be harmful.” However, only 10%of the enrolled patients had COPD. Also, approxi-

mately 25% of the conventional therapy group wascrossed over to NIV when they met criteria forrespiratory failure, and only 25% of these patientsrequired reintubation. Thus, results of the study51 donot apply to COPD patients and suggest that ratherthan initiating NIV early in patients deemed at riskfor postextubation failure, one should wait untilthere are clear indications for NIV so that appropri-ate patients can be selected.

Two subsequent trials support these latter infer-ences. Ferrer et al52 identified patients at risk forpostextubation failure by virtue of age � 65 years, ahistory of CHF, or an APACHE (acute physiologyand chronic health evaluation) II score � 12. Al-though postextubation respiratory failure and needfor intubation were significantly reduced by NIVoverall, most of the benefit was attributable to thehypercapnic subgroup, amounting to about one thirdof the patients, who also had a significantly lowermortality rate than control subjects (4% vs 41%,p � 0.05). Nava et al53 used NIV in patients at “highrisk” for extubation failure, using criteria similar tothose of Esteban et al,51 but risk was higher becausepatients had to have failed at least one T-piece trial.Once again, the need for reintubation (8% vs 18%,p � 0.027) was reduced compared to the conven-tional therapy group. Also, ICU mortality was 10%less in the NIV group (p � 0.01) mediated by thereduced need for reintubation. Thus, a trial of NIVappears to be warranted in patients at high risk forextubation failure, particularly if they have hypercap-nic respiratory failure.

Selection of Patients for NIV

Selection of appropriate patients is crucial for theoptimization of NIV success rates and resourceutilization. Often, NIV must be started before labo-ratory data are available because patients may dete-riorate during the delay and increase the risk of NIVfailure. As depicted in Figure 1, patients with respi-ratory distress and an appropriate diagnosis shouldbe considered for NIV. At the bedside, the clinicianmust make two fundamental judgments: (1) whetherthe patient needs ventilatory assistance based onsymptoms and signs of increased work of breathingor arterial blood gas derangements, and (2) whethersuch patients are candidates for NIV or should bepromptly intubated. These determinations are key tothe appropriate application and outcome of NIV andare based on the diagnosis, bedside observations, theclinician’s experience, and consideration of availableguidelines (Table 1). The timing of NIV initiation isimportant too. NIV should be started early, as soon




Figure 1. Algorithm illustrating the principles of patient selection and practical application of NIV.Patients are started on NIV if respiratory distress develops in the setting of de novo or acute-on-chronicrespiratory failure, or following surgery or extubation. They should have an appropriate diagnosis andmeet guidelines demonstrating the need for ventilatory assistance and absence of contraindications.After starting NIV, they should be closely monitored and checked at 1 to 2 h to establish that they areresponding favorably. If they have ALI/ARDS and are not good candidates, have contraindications orfail the 1- to 2-h checkpoint, they should be intubated unless they have a DNI status, in which casesome patients might still benefit from palliation of respiratory distress. NIV to facilitate weaning shouldbe considered for intubated patients. Patients who respond favorably to NIV should be monitoredclosely and reassessed periodically to determine whether they are ready to attempt weaning, which isusually accomplished by temporary discontinuation. If they have persisting respiratory failure aftertemporary discontinuation, long-term nocturnal NIV should be considered.




as indications appear, because delay may permitfurther deterioration and increase the likelihood offailure.54

Coma has been considered a contraindication toNIV in the past, but in a prospective cohort study,Gonzalez Diaz et al55 observed a high success rate ofNIV in patients with hypercapnic coma. Also, Scalaet al56 showed in a case study that NIV may besuccessfully used in COPD patients with acute re-spiratory failure and altered consciousness, althoughmore severely impaired consciousness was associatedwith higher failure rates.

Predictors of NIV success or failure may also behelpful in selecting patients (Table 2). The bestpredictor of success is a favorable response to NIVwithin the first 2 h. In a prospective cohort study ofnearly 800 COPD patients treated with NIV, Con-falonieri et al57 identified four factors—APACHE IIscore, pH, respiratory rate, and Glasgow comascore—that, when combined in a chart, showed goodpredictive value at baseline. These factors had evenbetter predictive value after 2 h of NIV use; if all fourfactors were favorable, the chance of success was97%; whereas if all were unfavorable, failure was avirtual certainty (99%).

Antonelli et al47 made similar observations inpatients with hypoxemic respiratory failure: if Pao2/Fio2 failed to increase � 146 after the first hour ofNIV therapy, or if the patient had pneumonia andARDS, the risk of NIV failure was increased. Theseobservations, combined with those of Esteban et al,51

demonstrating worse outcomes in NIV-treated pa-tients having delayed reintubations, emphasize theimportance of carefully reassessing patients soonafter NIV initiation (1- to 2-h checkpoint as depictedin Fig 1). If they fail to improve sufficiently, they

should be promptly intubated because a delay inneeded intubation permits the development of arespiratory crisis, requiring emergent intubation andincreasing the likelihood of morbidity or mortality.

Advances in Technology

Interfaces

A well-fitting and comfortable interface (or mask)is crucial to the success of noninvasive ventilation.Although nasal masks have certain advantages overoronasal (or full face) masks including greater com-fort, less likelihood of causing claustrophobia, andeasier speech and expectoration, they also permitmore air leakage through the mouth and have beenassociated with a higher rate of initial intoleranceduring acute applications of NIV.58 Thus, oronasalmasks are preferred initially for most critical careapplications, although a nasal mask should still beconsidered for patients with claustrophobia or fre-quent expectoration or for long-term applications.

Other mask types that are receiving attentioninclude the Total Face Mask (Respironics; Murrys-ville, PA), which seals around the perimeter of theface and may enhance mask tolerance in somepatients, and the helmet, which has not yet beenapproved by the Food and Drug Administration forNIV in the United States but has been investigatedin Europe.59,60 The latter device consists of a plasticcylinder that fits over the head and seals around theneck and shoulders. Compared to the full face maskin a case-control study61 of NIV to treat COPD

Table 1—Selection Guidelines for NIV in theAcute Setting

Appropriate diagnosis with potential reversibilityEstablish need for ventilatory assistance

Moderate-to-severe respiratory distress andTachypnea (respiratory rate � 24/min for COPD, � 30/min for

CHF); accessory muscle use or abdominal paradoxBlood gas derangement (pH � 7.35, Paco2 � 45 mm Hg, or

Pao2/Fio2 � 200)Exclude patients with contraindications to NIV

Respiratory or cardiac arrestMedical instability (hypotensive shock, myocardial infarction

requiring intervention, uncontrolled ischemia or arrhythmias)Unable to protect airwayUnable to fit maskUntreated pneumothoraxRecent upper airway or esophageal surgeryExcessive secretions*Uncooperative or agitated*

*Relative contraindications.

Table 2—Factors Associated With NIV Success in theAcute Setting

Synchronous breathing with ventilatorDentateLess air leakingFewer secretionsGood toleranceRespiratory rate � 30/min*Lower APACHE II score (� 29)*pH � 7.30*Glasgow coma score 15*Pao2/Fio2 � 146 after first hour if hypoxemic respiratory failureCOPD, CPENo pneumonia, ARDSBest predictor of success is a good response to NPPV within 1 to

2 h:Reduction in respiratory rateImprovement in pHImprovement in oxygenationReduction in Paco2

*If all four are present in COPD patients at baseline, the likelihoodof success is 94%; if present after 2 h of therapy, the likelihood ofsuccess is 97%.46




patients with acute respiratory failure, the helmetachieved similar improvements in vital signs, equiv-alent intubation and mortality rates and caused fewercomplications, but Paco2 tended to be higher at theend of the treatment period despite a higher level ofpressure support. Also, noise levels within the hel-met may be as high as 100 decibels, compared to 70decibels with a full-face mask.62 Thus, although thehelmet has some advantages over the full face maskwith regard to comfort and complications, it hasother disadvantages including less efficient CO2removal and noisiness that limit its current utility.

Dead Space and Rebreathing

By virtue of its single-ventilator-circuit design,bilevel ventilation has raised concerns about re-breathing.63 A lung model study64 demonstrated thatmasks with smaller volumes were associated with lessrebreathing and an in-mask exhalation port mini-mized rebreathing compared to an in-line port.Another lung model study65 used a mannequin faceto demonstrate that an in-mask exhalation port overthe bridge of the nose minimized dynamic deadspace, sometimes to levels below physiologic, pre-sumably by flushing CO2 from the nose and mouth.The correlation between dynamic dead space andactual mask volume was poor, probably because ofair streaming, and dead space was also minimized ifpositive expiratory pressure flushed CO2 from theventilator tubing. Whether these differences in deadspace and rebreathing are clinically important re-mains unclear, but these studies support the use ofin-mask exhalation ports and positive expiratorypressure during bilevel ventilation.

Ventilators

NIV is usually delivered either by blower-basedportable positive pressure “bilevel” ventilators de-rived from home-based CPAP systems or “criticalcare” ventilators designed to deliver invasive me-chanical ventilation. No study has shown better NIVsuccess rates for one type of ventilator than theother, but the ventilator mode used and specificsettings are important for patient comfort and de-creased work of breathing. Pressure support ventila-tion is rated as more tolerable by patients thanassist-control modes,66 and some studies67–69 havedemonstrated greater comfort with proportional as-sist ventilation than with pressure support, presum-ably because it is targeted to inspiratory flow as asurrogate of patient effort and can respond nearlyinstantaneously to changes in demand. Proportional-assist ventilation has only recently been approved bythe Food and Drug Administration, but it has beenavailable elsewhere in the world for almost a decade.

The perceived need for multiple adjustments tocompensate for patient elastance and resistance aswell as added cost have probably limited greater useof this mode despite the finding in one of thecontrolled trials that proportional assist requiredfewer adjustments than pressure support.68

Other desirable attributes of ventilators for NIVinclude the ability to compensate for air leaks, whichhelps to assure delivery of adequate tidal volumes.70

Because NIV lowers humidity of delivered gas,humidification is useful to bring relative humidityback toward the ambient range, possibly enhancingcomfort.71 Heated passover humidifiers have mini-mal effects on delivered pressures, whereas heat andmoisture exchangers are to be avoided because theycan interfere with the ability of NIV to reduce workof breathing.72

Ventilator Settings

L’Her et al73 showed that patients with acute lunginjury treated with NIV require pressure supportlevels of at least 10 to 15 cm H2O to reduce work ofbreathing. Not surprisingly, higher levels of positiveend-expiratory pressure (PEEP) [10 cm H2O vs 5 cmH2O] were more effective at improving oxygenation.Combining higher levels of pressure support withhigh-level PEEP can detract from patient comfort,however, so compromises may be necessary to opti-mize settings; maximal oxygenation may have to besacrificed if patient comfort and reduction in work ofbreathing are prioritized. Many noninvasive ventila-tors now offer adjustable “rise times” or pressuriza-tion rates—the time taken to achieve the targetinspiratory pressure— to optimize patient comfort.Prinianakis et al74 found that a rapid pressurizationrate was most effective at reducing work of breathingin COPD patients, but a slightly slower rate wasassociated with better comfort ratings.

Guidelines, Utilization, and Outcomes

Sinuff et al75 found that a NIV guideline influ-enced caregiver behavior, leading to greater ICUutilization and more ordering of pulmonary consul-tations and arterial blood gases. However, overallmortality rate was unchanged and, of concern, themortality rate actually increased in patients withoutCOPD or CHF as the cause of their acute respira-tory failure, who were excluded from NIV use by theguideline. The results highlight the need for ongoingguideline evaluation and modification because theycould increase resource utilization and the cost ofcare if they mandate ICU use and frequent labora-tory testing among all NIV patients, some of whomcould conceivably be managed in less costly environ-ments.




Studies of NIV utilization in the acute care settinghave found that enormous disparity exists betweendifferent institutions. In a 1997 survey of NIV use inEuropean ICUs, Carlucci et al76 found that 20% ofICUs surveyed used no NIV at all and, overall, NIVwas used in 16% of all ventilator starts. A subsequentUK survey77 found that 52% of hospitals were notusing NIV. A more recent survey78 in the UnitedStates found that although only 1 of 71 respondinghospitals used no NIV, some used it in � 5% ofventilator starts and others in � 50%. Overall aver-age use among ventilator starts was 20%, but only athird of patients with COPD or CHF received NIVas their initial ventilator therapy. Major reasons fornot using NIV more were lack of physician knowl-edge and inadequately trained staff, suggesting thateducation may help to enhance utilization. However,progress is being made, as indicated by Demoule etal,79 who found that the overall percentage for NIVamong ventilator starts in European ICUs had risento 23% by 2002; and Girou et al,80 who showed thatincreasing use of NIV in CHF and COPD patients(from approximately 20 to 90% of ventilator starts)over a 7-year period in a French ICU was associatedwith a decrease in the rate of nosocomial pneumo-nias from 20 to 8% and in ICU mortality rate from 21to 7% (all p � 0.05). The latter study80 also illustratesthe value of increasing experience using NIV afterestablishing an NIV program.

Conclusion

The role of NIV in the management of acuterespiratory failure has been further clarified in re-cent years. Evidence is strong to support the use ofNIV in the initial management of acute respiratoryfailure in patients with COPD exacerbations, acuteCPE, and immunocompromised states, and to facil-itate extubation in patients with COPD with failedspontaneous breathing trials. A trial of NIV is justi-fied in patients with asthma exacerbations, postop-erative respiratory failure, extubation failure, hypox-emic respiratory failure or a DNI status, but becausesupporting evidence is not as strong, they should becarefully selected according to available guidelinesand clinical judgment, taking into account risk fac-tors for NIV failure. Once begun, patients should beclosely monitored in an ICU or step-down unit untiladequately stabilized, paying attention not only tovital signs and gas exchange, but also to comfort andtolerance. If patients do not have a favorable initialresponse to NIV, clinicians should strongly considerintubation without delay. When used appropriately,NIV improves patient outcomes and the efficiency ofcare. Although it is still used in only a select minority

of patients with acute respiratory failure, it hasassumed an important role in the therapeutic arma-mentarium. With technical advances and new evi-dence on its proper application, this role is likely toexpand.

References1 Liesching T, Kwok H, Hill NS. Acute applications of nonin-

vasive positive pressure ventilation. Chest 2003; 124:699–7132 Brochard L, Isabey D, Piquet J, et al. Reversal of acute

exacerbations of chronic obstructive lung disease by inspira-tory assistance with a face mask. N Engl J Med 1990;323:1523–1530

3 Bott J, Carroll MP, Conway JH, et al. Randomised controlledtrial of nasal ventilation in acute ventilatory failure due tochronic obstructive airways disease. Lancet 1993; 341:1555–1557

4 Kramer N, Meyer TJ, Meharg J, et al. Randomized, prospec-tive trial of noninvasive positive pressure ventilation in acuterespiratory failure. Am J Respir Crit Care Med 1995; 151:1799–1806

5 Plant PK, Owen JL, Elliott MW. Early use of non-invasiveventilation for acute exacerbations of chronic obstructivepulmonary disease on general respiratory wards: a multicen-tre randomised controlled trial. Lancet 2000; 355:1931–1935

6 Dikensoy O, Ikidag B, Filiz A, et al. Comparison of non-invasive ventilation and standard medical therapy in acutehypercapnic respiratory failure: a randomised controlledstudy at a tertiary health centre in SE Turkey. Int J Clin Pract2002; 56:85–88

7 Celikel T, Sungur M, Ceyhan B, et al. Comparison ofnoninvasive positive pressure ventilation with standard med-ical therapy in hypercapnic acute respiratory failure. Chest1998; 114:1636–642

8 Ram FS, Picot J, Lightowler J, et al. Non-invasive positivepressure ventilation for treatment of respiratory failure due toexacerbations of chronic obstructive pulmonary disease. Co-chrane Database Syst Rev 2004; 3:CD004104

9 Keenan SP, Sinuff T, Cook DJ, et al. Which patients withacute exacerbation of chronic obstructive pulmonary diseasebenefit from noninvasive positive pressure ventilation? Asystematic review of the literature. Ann Intern Med 2003;138:861–870

10 Bersten AD, Holt AW, Vedig AE, et al. Treatment of severecardiogenic pulmonary edema with continuous positive air-way pressure delivered by face mask. N Engl J Med 1991;325:1825–1830

11 Lin M, Yang Y, Chiany H, et al. Reappraisal of continuouspositive airway pressure therapy in acute cardiogenic pulmo-nary edema: short-term results and long-term follow-up.Chest 1995; 107:1379–1386

12 Nava S, Carbone G, DiBattista N, et al. Noninvasive ventila-tion in cardiogenic pulmonary edema: a multicenter random-ized trial. Am J Respir Crit Care Med 2003; 168:1432–1437

13 Crane SD, Elliott MW, Gilligan P, et al. Randomised con-trolled comparison of continuous positive airways pressure,bilevel non-invasive ventilation, and standard treatment inemergency department patients with acute cardiogenic pul-monary oedema. Emerg Med J 2004; 21:155–161

14 Pang D, Keenan SP, Cook DJ, et al. The effect of positiveairway pressure on mortality and the need for intubation incardiogenic pulmonary edema. Chest 1998; 114:1185–1192

15 Rasanen J, Heikkila J, Downs J, et al. Continuous positiveairway pressure by face mask in acute cardiogenic pulmonary




edema. Am J Cardiol 1985; 55:296–30016 Lin M, Chiang HT. The efficacy of early continuous positive

airway pressure therapy in patients with acute cardiogenicpulmonary edema. J Formos Med Assoc 1991; 90:736–743

17 Park M, Sangean MC, Volpe Mde S, et al. Randomized,prospective trial of oxygen, continuous positive airway pres-sure, and bilevel positive airway pressure by face mask inacute cardiogenic pulmonary edema. Crit Care Med 2004;32:2407–2415

18 Masip J, Roque M, Sanchez B, et al. Noninvasive ventilationin acute cardiogenic pulmonary edema. JAMA 2005; 294:3124–3130

19 Winck JC, Azevedo LF, Costa-Pereira A, et al. Efficacy andsafety of non-invasive ventilation in the treatment of acutecardiogenic pulmonary edema: a systematic review and meta-analysis. Crit Care 2006; 10:R69

20 Collins SP, Mielniczuk LM, Whittingham HA, et al. The useof noninvasive ventilation in emergency department patientswith acute cardiogenic pulmonary edema: a systematic re-view. Ann Emerg Med 2006; 48:260–269

21 Ho KM, Wong K. A comparison of continuous and bi-levelpositive airway pressure non-invasive ventilation in patientswith acute cardiogenic pulmonary oedema: a meta-analysis.Crit Care 2006; 10:R4921

22 Mehta S, Jay GD, Woolard RH, et al. Randomized prospec-tive trial of bilevel versus continuous positive airway pressurein acute pulmonary edema. Crit Care Med 1997; 25:620–628

23 Ferrer M, Esquinas A, Arancibia F, et al. Noninvasiveventilation during persistent weaning failure: a randomizedcontrolled trial. Am J Respir Crit Care Med 2003; 168:70–76

24 Nava S, Ambrosino N, Clini E, et al. Non-invasive mechanicalventilation in the weaning of patients with respiratory failuredue to chronic obstructive pulmonary disease: a randomizedstudy. Ann Intern Med 1998; 128:721–728

25 Antonelli M, Conti G, Bufi M, et al. Noninvasive ventilationfor treatment of acute respiratory failure in patients under-going solid organ transplantation: a randomized trial. JAMA2000; 283:2239–2240

26 Hilbert G, Gruson D, Vargas F, et al. Noninvasive ventilationin immunosuppressed patients with pulmonary infiltrates,and acute respiratory failure. N Engl J Med 2001; 344:481–487

27 Meduri GU, Turner RE, Abou-Shala N, et al. Noninvasivepositive pressure ventilation via face mask: first line interven-tion in patients with acute hypercapnic and hypoxemic respi-ratory failure. Chest 1996; 109:179–193

28 Meduri GU, Cook TR, Turner RE, et al. Noninvasive positivepressure ventilation in status asthmaticus. Chest 1996; 110:767–774

29 Fernandez MM, Villagra A, Blanch L, et al. Non-invasivemechanical ventilation in status asthmaticus. Intensive CareMed 2001; 27:486–492

30 Soroksky A, Stav D, Shpirer I. A pilot prospective, random-ized, placebo-controlled trial of bi-level positive airway pres-sure in acute asthmatic attack. Chest 2003; 123:1018–1025

31 Agarwal R, Malhotra P, Gupta D. Failure of NIV in acuteasthma: case report and a word of caution. Emerg Med J2006; 23:e9

32 Ram FS, Wellington S, Rowe B, et al. Non-invasive positivepressure ventilation for treatment of respiratory failure due tosevere acute exacerbations of asthma. Cochrane DatabaseSyst Rev 2005; 1:CD004360

33 Kindgen-Milles D, Muller E, Buhl R, et al. Nasal continuouspositive airway pressure reduces pulmonary morbidity andlength of stay following thoracoabdominal aortic surgery.Chest 2005; 128:821–828

34 Squadrone V, Coha M, Cerutti E, et al. Continuous positive

airway pressure for treatment of postoperative hypoxemia.JAMA 2005; 293:589–595

35 Auriant I, Jallot A, Herve P, et al. Noninvasive ventilationreduces mortality in acute respiratory failure following lungresection. Am J Respir Crit Care Med 2001; 164:1231–1235

36 Clarke DE, Vaughan L, Raffin TA. Noninvasive positivepressure ventilation for patients with terminal respiratoryfailure: the ethical and economic cost of delaying the inevi-table are too great. Am J Crit Care 1994; 3:4–5

37 Levy MM, Tanios MA, Nelson D, et al. Outcomes of patientswith do-not- intubate orders treated with noninvasive venti-lation. Crit Care Med 2004; 32:2002–2007

38 Schettino G, Altobelli N, Kacmarek RM. Noninvasive posi-tive-pressure ventilation reverses acute respiratory failure inselect “do-not-intubate” patients. Crit Care Med. 2005; 33:1976–1982

39 Curtis RJ, Cook DJ, Sinuff T, et al. Noninvasive positivepressure ventilation in critical and palliative care settings:understanding the goals of therapy. Crit Care Med 2007 (inpress)

40 Antonelli M, Conti G, Rocco M, et al. A comparison ofnoninvasive positive-pressure ventilation and conventionalmechanical ventilation in patients with acute respiratoryfailure. N Engl J Med 1998; 339:429–435

41 Ferrer M, Esquinas A, Leon M, et al. Noninvasive ventilationin severe hypoxemic respiratory failure: a randomized clinicaltrial. Am J Respir Crit Care Med 2003; 168:1438–1444

42 Ambrosino N, Foglio K, Rubini F, et al. Noninvasive mechan-ical ventilation in acute respiratory failure due to chronicobstructive pulmonary disease: correlates for success. Thorax1995; 50:755–757

43 Honrubia T, Garcia Lopez F, Franco N, et al. Noninvasive vs.conventional mechanical ventilation for acute respiratoryfailure. Chest 2005; 128:3916–3924

44 Antonelli M, Conti G, Moro ML, et al. Predictors of failuresof noninvasive positive pressure ventilation in patients withacute hypoxemic respiratory failure: a multi-center study.Intensive Care Med 2001; 27:1718–1728

45 Jolliet P, Abajo B, Pasquina P, et al. Non-invasive pressuresupport ventilation in severe community-acquired pneumo-nia. Intensive Care Med 2001; 27:812–821

46 Confalonieri M, Potena A, Carbone G, et al. Acute respiratoryfailure in patients with severe community-acquired pneumo-nia. Am J Respir Crit Care Med 1999; 160:1585–1591

47 Antonelli M, Conti G, Esquinas A, et al. A multiple-centersurvey on the use in clinical practice of noninvasive ventila-tion as a first-line intervention for acute respiratory distresssyndrome. Crit Care Med 2007; 35:18–25

48 Rana S, Hussam J, Gay P, et al. Failure of non-invasiveventilation in patients with acute lung injury: observationalcohort study. Crit Care 2006; 10:R79

49 Blivet S, Philit F, Sab JM, et al. Outcome of patients withidiopathic pulmonary fibrosis admitted to the ICU for respi-ratory failure. Chest 2001; 120:8–10

50 Nevins ML, Epstein SK. Predictors of outcome for patientswith COPD requiring invasive mechanical ventilation. Chest2001; 119:1840–1849

51 Esteban A, Frutos-Vivar F, Ferguson ND, et al. Noninvasivepositive-pressure ventilation for respiratory failure after extu-bation. N Engl J Med 2004; 350:2452–2460

52 Ferrer M, Valencia M, Nicolas JM, et al. Early noninvasiveventilation averts extubation failure in patients at high risk: arandomized trial. Am J Respir Crit Care Med 2006; 173:164–170

53 Nava S, Gregoretti C, Fanfulla F, et al. Noninvasive ventila-tion to prevent respiratory failure after extubation in high-riskpatients. Crit Care Med 2005; 33:2465–470




54 Nava S, Navalesi P, Conti G. Time of non-invasive ventilation.Intensive Care Med 2006; 32:361–370

55 Gonzalez Diaz G, Carillo A, Perez P, et al. Noninvasivepositive-pressure ventilation to treat hypercapnic coma sec-ondary to respiratory failure. Chest 2005; 127:952–960

56 Scala R, Naldi M, Archinucci I, et al. Noninvasive positive-pressure ventilation in patients with acute exacerbations ofCOPD and varying levels of consciousness. Chest 2005;128:1657–1666

57 Confalonieri M, Garuti G, Cattaruzza MS, et al. A chart offailure risk for noninvasive ventilation in patients with COPDexacerbation. Eur Respir J 2005; 25:348–355

58 Kwok H, McCormack J, Cece R, et al. Controlled trial oforonasal versus nasal mask ventilation in the treatment ofacute respiratory failure. Crit Care Med 2003; 31:468–473

59 Tonnelier JM, Prat G, Nowak E, et al. Noninvasive continu-ous positive airway pressure ventilation using a new helmetinterface: a case prospective pilot study. Intensive Care Med2003; 29:2077–2080

60 Principi T, Pantanetti S, Catani F, et al. Noninvasive contin-uous positive airway pressure delivered by helmet in hema-tological malignancy patients with hypoxemic acute respira-tory failure. Intensive Care Med 2004; 30:147–150

61 Antonelli M, Pennisi MA, Pelosi P, et al. Noninvasive positivepressure ventilation using a helmet in patients with acuteexacerbation of chronic obstructive pulmonary disease. Anes-thesiology 2004; 100:16–24

62 Cavaliere F, Conti G, Costa R, et al. Noise exposure duringnoninvasive ventilation with a helmet, a nasal mask, and afacial mask. Intensive Care Med 2004; 30:1755–1760

63 Ferguson GT, Gilmartin M. CO2 rebreathing during BiPAPventilatory assistance. Am J Respir Crit Care Med 1995;151:1126–1135

64 Schettino GPP, Chatmongkolchart S, Hess D, et al. Positionof exhalation port and mask design affect CO2 rebreathingduring noninvasive positive pressure ventilation. Crit CareMed 2003; 31:2178–2182

65 Saatci E, Miller DM, Sztell IM, et al. Dynamic dead space inface masks used with noninvasive ventilators: a lung modelstudy. Eur Respir J 2004; 23:129–135

66 Vitacca M, Rubini F, Foglio K, et al. Noninvasive modalitiesof positive pressure ventilation improve the outcome of acuteexacerbations of COLD patients. Intensive Care Med 1993;19:450–455

67 Fernandez-Vivas M, Caturia-Such J, de la Rosa JG, et al.Noninvasive pressure support versus proportional assist ven-

tilation in acute respiratory failure. Intensive Care Med 2003;29:1126–1133

68 Gay PC, Hess DR, Hill NS. Noninvasive proportional assistventilation for acute respiratory insufficiency: comparisonwith pressure support ventilation. Am J Respir Crit Care Med2001; 164:1606–1611

69 Wysocki M, Richard JC, Meshaka P. Noninvasive propor-tional assist ventilation compared with noninvasive pressuresupport ventilation in hypercapnic acute respiratory failure.Crit Care Med 2002; 30:323–329

70 Mehta S, McCool FD, Hill NS. Leak compensation inpositive pressure ventilators: a lung model study. Eur RespirJ 2001; 17:259–267

71 Holland AE, Denehy L, Buchan CA, et al. Efficacy of aheated passover humidifier during noninvasive ventilation: abench study. Respir Care 2007; 52:38–44

72 Lellouche F, Maggiore SM, Deye N, et al. Effect of thehumidification device on the work of breathing during non-invasive ventilation. Intensive Care Med 2002; 28:1582–1589

73 L’Her E, Deye N, Lellouche F, et al. Physiologic effects ofnoninvasive ventilation during acute lung injury. Am J CritCare Med 2005; 172:1112–1118

74 Prinianakis G, Delmastro M, Carlucci A, et al. Effect ofvarying the pressurization rate during noninvasive pressuresupport ventilation. Eur Respir J 2004; 23:314–320

75 Sinuff T, Cook DJ, Randall J, et al. Evaluation of a practiceguideline for noninvasive positive pressure ventilation foracute respiratory failure. Chest 2003; 123:2062–2073

76 Carlucci A, Richard JC, Wysocki M, et al. Noninvasive versusconventional mechanical ventilation. Am J Respir Crit CareMed 2001; 163:874–880

77 Doherty MJ, Greenstone MA. Survey of non-invasive venti-lation (NIPPV) in patients with acute exacerbations of chronicobstructive pulmonary disease (COPD) in the UK. Thorax1998; 53:863–866

78 Maheshwari V, Paioli D, Rothaar R, et al. Utilization ofnoninvasive ventilation in acute care hospitals. Chest 2006;129:1226–1233

79 Demoule A, Girou E, Richard JC, et al. Increased use ofnoninvasive ventilation in French intensive care units. Inten-sive Care Med 2006; 32:1747–1755

80 Girou E, Brun-Buisson C, Taille S, et al. Secular trends innosocomial infections and mortality associated with noninva-sive ventilation in patients with exacerbation of COPD andpulmonary edema. JAMA 2003; 290:2985–2991




DOI 10.1378/chest.06-2643 2007;132; 711-720Chest

Erik Garpestad, John Brennan and Nicholas S. Hill*Noninvasive Ventilation for Critical Care

April 13, 2010This information is current as of

& ServicesUpdated Information

htmlhttp://chestjournal.chestpubs.org/content/132/2/711.full.high-resolution figures, can be found at:Updated Information and services, including

References

full.html#ref-list-1http://chestjournal.chestpubs.org/content/132/2/711.accessed free at:This article cites 79 articles, 44 of which can be

Citations

full.html#related-urlshttp://chestjournal.chestpubs.org/content/132/2/711.articles:This article has been cited by 2 HighWire-hosted

Open AccessoptionFreely available online through CHEST open access

Permissions & Licensing

http://www.chestjournal.org/site/misc/reprints.xhtml(figures, tables) or in its entirety can be found online at: Information about reproducing this article in parts

Reprints http://www.chestjournal.org/site/misc/reprints.xhtml

Information about ordering reprints can be found online:

Email alerting service

online article.article. Sign up in the box at the top right corner of the Receive free email alerts when new articles cite this

formatImages in PowerPoint

format. See any online article figure for directions downloaded for teaching purposes in PowerPoint slide Figures that appear in CHEST articles can be



http://chestjournal.chestpubs.org/content/132/2/711.full.html#ref-list-1

http://chestjournal.chestpubs.org/content/132/2/711.full.html#related-urls

http://www.chestjournal.org/site/misc/reprints.xhtml

http://www.chestjournal.org/site/misc/reprints.xhtml








Doc 12 –

APRV Theory and Practice

�Airway pressure release ventilation(APRV) is a relatively new mode ofventilation, that only becamecommercially available in the UnitedStates in the mid-1990s. Airway pressurerelease ventilation produces tidalventilation using a method that differsfrom any other mode. It uses a release ofairway pressure from an elevatedbaseline to simulate expiration. Theelevated baseline facilitates oxygenation,and the timed releases aid in carbondioxide removal.

Advantages of APRV include lowerairway pressures, lower minuteventilation, minimal adverse effects oncardio-circulatory function, ability tospontaneously breathe throughout theentire ventilatory cycle, decreasedsedation use, and near elimination ofneuromuscular blockade. Airway pressurerelease ventilation is consistent with lungprotection strategies that strive to limitlung injury associated with mechanicalventilation. Future research will probablysupport the use of APRV as the primarymode of choice for patients with acutelung injury. (KEYWORDS: acute lunginjury, airway pressure releaseventilation, alveolar recruitment, alveolarderecruitment, lung protective strategies)

Airway pressure release ventilation (APRV) isa mode of ventilation that was first describedin 1987.1,2 It uses a philosophy that differsfundamentally from that of conventionalventilation. Whereas conventional modes of

AACN Clinical IssuesVolume 12, Number 2, pp. 234–246© 2001, AACN

ventilation begin the ventilatory cycle at abaseline pressure and elevate airway pres-sure to accomplish tidal ventilation (Figure1), APRV commences at an elevated baselinepressure (similar to a plateau pressure) andfollows with a deflation to accomplish tidalventilation (Figure 2). In addition, duringAPRV, spontaneous breathing may occur ateither the plateau pressure or deflation pres-sure levels. This article provides a detailedexamination of the terminology, indications,theoretical benefits, advantages, and disad-vantages of APRV as well as a discussion ofapplication and weaning procedures.

� Airway Pressure ReleaseVentilation Defined

Airway pressure release ventilation has beendescribed as continuous positive airway pres-sure (CPAP) with regular, brief, intermittentreleases in airway pressure.3,4 The releasephase results in alveolar ventilation and re-moval of carbon dioxide (CO2). Airway pres-sure release ventilation, unlike CPAP, facili-tates both oxygenation and CO2 clearanceand originally was described as an improvedmethod of ventilatory support in the presenceof acute lung injury (ALI) and inadequate CO2

ventilation.2,5 Airway pressure release ventila-tion is capable of either augmenting alveolar

Airway Pressure ReleaseVentilation: Theory and PracticeP. Milo Frawley, RN, MS,* and Nader M. Habashi, MD†

▪ ▪ ▪ ▪ ▪ ▪ ▪ ▪ ▪ ▪

From *Maryland ExpressCare and †Department ofCritical Care Medicine, University of Maryland MedicalCenter, Baltimore, Maryland.

Reprint requests to P. Milo Frawley, RN, MS, Mary-land ExpressCare, TGR25C, University of MarylandMedical Center, 22 South Greene Street, Baltimore, MD21201.

234

Vol. 12, No. 2 May 2001 APRV: THEORY AND PRACTICE � 235

ventilation in the spontaneously breathing pa-tient or accomplishing complete ventilation inthe apneic patient.6 The CPAP level drivesoxygenation, while the timed releases aid inCO2 clearance.

Technically, APRV is a time-triggered, pres-sure-limited, time-cycled mode of mechanicalventilation. In addition, APRV allows unre-stricted, spontaneous breathing throughout

the entire ventilatory cycle (Table 1). Advan-tages of APRV include: significantly lowerpeak/plateau airway pressures for a giventidal volume; the ability to superimpose spon-taneous breathing throughout the ventilatorycycle; decreased sedation; and near elimina-tion of neuromuscular blockade use.7,8 Fea-tures that distinguish APRV from other modesof mechanical ventilation include sponta-

Figure 1. Conventional volume targeted ventilation, e.g., synchronized intermittent mandatory ventilation(SIMV). Any mechanically delivered breath will be defined by its trigger, limit, and cycle off feature. In SIMV, thebreath will be triggered by either the patient or by time, the volume delivered will limit the breath, and time will cy-cle the breath off into exhalation. Cms of H2O = centimeters of water.

Figure 2. Airway pressure release ventilation: this can also be defined by a trigger, limit, and cycle off feature.However, unlike other modes of ventilation, the trigger (time) initiates a drop in airway pressure. The amount ofpressure change will be the limit. The cycle off will occur because of time. Airway pressure then returns to thebaseline. Cms of H2O = centimeters of water.

236 � FRAWLEY AND HABASHI AACN Clinical Issues

neous breathing throughout the ventilatorycycle and an intermittent pressure releasephase that results in a brief decrease in lungvolume to assist ventilation.1,2

� History of MechanicalVentilation

The basic principles of ventilator design andmanagement were founded upon patientswho developed nonparenchymal respiratoryfailure (e.g., polio). In the absence of ade-quate research, those same principles wereapplied to patients with parenchymal respira-tory failure as well (e.g., ALI). Mode selectionoften was based on availability and simplicityof the ventilator, user experience, and tradi-tion, because little evidence existed to guidemanagement.

In 1993, the American College of ChestPhysicians (ACCP) consensus conferencefailed to “agree on an optimum mode ofventilation for any disease state or an opti-mum method of weaning patients from me-chanical ventilation.”9(p1834) The ACCPagreed that well-controlled clinical trialsthat defined the indications and uses ofspecific modes of ventilation were lacking.New technology must scientifically show adistinct advantage in safety, expense, easeof operation, or therapeutic outcome.10,11

Despite more than 30 years since itsrecognition, acute respiratory distress syn-drome (ARDS) continues to have a 30% to50% mortality rate.12,13 Recently, discovery ofthe potential for mechanical ventilation toproduce ventilator-associated lung injury hasresulted in the development of new lungprotective strategies.14 Lung protective strate-gies include those described in the “theopen lung approach” promoted by Amato etal.15 The open lung approach uses reducedtidal volumes (6 mL/kg) to prevent high-vol-ume lung injury and over-distension of air-spaces. In addition, Amato et al.16 used ele-vated end expiratory pressure (averagepositive end-expiratory pressure [PEEP] 16cm water pressure), to prevent low volumelung injury from cyclic airway reopening.

The recently completed ARDSNet studycompared conventional tidal volume (12mL/kg) to reduced tidal volume (6 mL/kg).13

The results of the ARDSNet trial13 and astudy conducted by Amato et al.16 suggest anassociation between reduced tidal volumeand improved outcome. Although the ARD-SNet trial targeted similar PEEP levels in bothits groups, study protocols for maintainingsaturation resulted in higher levels of setPEEP in the low tidal volume group. In addi-tion, to maintain similar targets for PaCO2,the low tidal volume group had much higherrespiratory frequencies, resulting in the de-

TABLE 1 � Classification of Common Modes of Mechanical Ventilation

SpontaneousMode Trigger Limit Cycle Off Breathing Flow of Gas

A-C Time or patient Volume Time No Constant(volume)

A-C Time or patient Pressure Time No Decelerating(pressure)

SIMV Time or patient Volume Time Yes Constant(volume)

PSV Patient Pressure Flow of gas No Decelerating

PRVC Time or patient* Volume Time No Decelerating

APRV Time Pressure Time Yes Decelerating

Note: A mechanically delivered breath is made up of three distinct phases. The Trigger initiates the breath; the Limit will stop thebreath from increasing, but does not initiate exhalation; and the Cycle Off, that switches the breath from inspiration to exhala-tion. Beyond this, modes may or may not allow unsupported, spontaneous breathing. The inspired gas may be delivered usingeither a constant or decelerating flow of gas.

*This mode is designed for patients with no breathing capacity, though they are able to trigger breaths.

A-C � assist control; SIMV � synchronized intermittent mandatory ventilation; PSV � pressure support ventilation; PRVC � pressure regulated volume control; APRV � airway pressure release ventilation.


velopment of intrinsic PEEP. Therefore, therole of elevated levels of end expiratorypressure (PEEP) on survival of the low tidalvolume group may have been obscured. De-spite improved survival with the low tidalvolumes group, survival was less than that ofAmato’s16 combined approach (tidal volumereduction and PEEP elevation). As a result,the planned ARDSNet Assessment of Lowtidal Volume and Elevated end-expiratoryvolume to Obviate Lung Injury (ALVEOLI)study will evaluate the role of higher levelsof PEEP on survival. ARDSNet ALVEOLI willuse data from the pressure-volume curve todevelop the PEEP scale (PEEP scale = frac-tion of inspired oxygen:PEEP).

However, recent data suggest that deter-mining optimal PEEP from the pressure-vol-ume curve may be inaccurate.17 In addition,recruitment to prevent cyclic airway closure(low volume lung injury) requires pressurein excess of 30 cm of water pressure. Com-plete recruitment exceeds the lower inflec-tion point used by Amato et al.16 to deter-mine optimal PEEP levels. Recruitmentbegins at the lower inflection point andcontinues to the upper inflection point.18–20

Therefore, elevated baseline airway pres-sure during APRV may produce near com-plete recruitment, thus minimizing low vol-ume lung injury from cyclic recruitment.Additionally, APRV is less likely to produceover-inflation or high-volume lung injury,as airway pressures are lowered (released)to accomplish ventilation.

Other lung protective strategies includeoptimization of current modes of ventilationand alteration of ventilator strategies to pre-vent or reduce ventilator-associated lung in-jury. Current goals of ventilation include thefollowing:

• avoiding extension of lung injury,• minimizing oxygen toxicity by using

mean airway pressure (Paw),• recruiting alveoli by raising mean Paw by in-

creasing PEEP and/or prolonging inspira-tion,

• minimizing peak Paw,• preventing atelectasis, and• using sedation and paralysis judiciously.21

Although first described 11 years earlier,1,2

APRV may have benefits for preventing orlimiting ventilator-associated lung injury.

� Terminology

Unfortunately, a consistent vocabulary forAPRV has failed to mature. Four commonlyused terms include: pressure high (P High),pressure low (P Low), time high (T High),and time low (T Low).7 P High is the baselineairway pressure level and is the higher of thetwo airway pressure levels. Other authorshave described P High as the CPAP level,22

the inflating pressure,23 or the P1 pressure(P1). P Low is the airway pressure level re-sulting from the pressure release. Other au-thors may refer to P Low as the PEEP level,22

the release pressure,23 or the P2 pressure(P2). T High corresponds with the length oftime for which P High is maintained; T Lowis the length of time for which the P Low isheld (i.e. for which the airway pressure is re-leased).

The mean airway pressure can be calcu-lated as follows:7

(P High � T High) + (P Low � T Low)

T High + T Low

Some ventilators may compute this auto-matically, making manual calculation redun-dant. Common terms associated with APRVare summarized in Table 2 and Figure 3.

Somewhat confusing to the understand-ing of APRV have been the subsequent de-scriptions of modes of ventilation that ap-pear very similar to it. Biphasic positiveairway pressure (BIPAP)24,25 differs fromAPRV only in the timing of the upper andlower pressure levels. In BIPAP, T High usu-ally is shorter than T Low. One descriptionof BIPAP25 subdivides it into four categories,one of which is APRV-BIPAP.

Intermittent mandatory pressure releaseventilation (IMPRV),26 another mode of ven-tilation similar to and sometimes confusedwith APRV, synchronizes the release eventwith the patient’s spontaneous effort. The re-lease occurs after the patient’s second, third,fourth, fifth, or sixth spontaneous breath.Further, all spontaneous breaths are pressuresupported to overcome the resistance associ-ated with breathing through the endotra-cheal tube and ventilator tubing. Synchro-nization does not occur with the raising ofairway pressure, only the release. Becausethe concept of dyssynchrony in APRV hasnot been demonstrated clearly—and has


Figure 3. Airway pressure release ventilation terminology. Paw = airway pressure; P High = 30 centimeters of wa-ter (cms of H2O); P Low = 0 cms of H2O, T High = 6.0 seconds; T low = 0.8 seconds; calculated mean Paw = 26.5cms of H2O.

BiLevel ventilation28 is defined as aug-mented pressure ventilation that allows forunrestricted, albeit pressure-supported, spon-taneous breathing throughout the ventilatorycycle. Although similar to APRV, it incorpo-rates the option of pressure support in the air-way pressure waveform to augment sponta-neous breathing.

� Indications for APRV

Airway pressure release ventilation was de-signed to oxygenate and augment ventila-tion for patients with ALI or low-compliance

been stated not to be an issue—the necessityof intermittent mandatory pressure releaseventilation is questionable.10

Intermittent CPAP27 is based on the prin-ciples of APRV but is intended for patientsundergoing general anesthesia. Continuouspositive airway pressure is applied at alevel that will provide an adequate tidalvolume, then removed for 1 second to pro-duce tidal ventilation, then reapplied. Un-like APRV, intermittent CPAP is not in-tended to restore normal functional residualcapacity or improve oxygenation, and it canbe discontinued abruptly.

TABLE 2 � Summary of Airway Pressure Release Ventilation Terminology

Alternative Units of Term Definition Names Measure

Pressure High Baseline airway pressure level CPAP level,22 Cm H2O(P High)7 Higher of the two airway pressures Inflation pressure,23 P1

Pressure Low Airway pressure level resulting from PEEP level,22 Release Cm H2O(P Low)7 pressure release. The lower of the two pressure,23 P2

airway pressures

Time High Length of time for which P High is T1 Seconds(T High)7 maintained

Time Low Length of time for which P Low is T2 Seconds(T Low)7 maintained

Mean Paw (P High � T High) � (P Low � T Low) — Cm H2O

(T High � T Low)7

CPAP � continuous positive airway pressure; Cm H2O � centimeters of water; PEEP � positive end expiratory pressure; Paw � airway pressure.


lung disease.1,5,6 Airway pressure releaseventilation also has been used successfullywith patients with airway disease. Similar toCPAP, APRV can unload inspiratory musclesand decrease the work of breathing associ-ated with chronic obstructive pulmonary dis-ease.29 Unlike PEEP (an expiratory flow re-sistor, which decreases expiratory flow),peak expiratory flow rates are increased dur-ing the release phase of APRV, improvingexpiratory flow limitation. Furthermore, dur-ing APRV, exhalation is not limited to the re-lease phase, as it is permitted throughout therespiratory cycle.

The main causes of hypoxemia associ-ated with ALI are shunting due to alveolarcollapse and reduction in functional resid-ual capacity.1,7,30 Therefore, a primary goalof the treatment of ALI is recruitment ofalveoli and prevention of derecruitment.Sustained plateau pressure is used to pro-mote alveolar recruitment, while beingmaintained at an acceptable level. In addi-tion, the number of respiratory cycles isminimized to prevent both the repetitiveopening of alveoli and alveolar stretch, thatmay result in lung injury.

Patients in early-phase ALI often do nothave impaired respiratory muscle strength orinadequate respiratory drive. Frequently,CPAP alone is sufficient to restore lung vol-ume and increase lung compliance. How-ever, when assistance with ventilation is re-quired, APRV can be used. Intermittent

airway pressure release allows alveolar gasto be expelled via natural lung recoil.1

� Importance of CollateralChannels of Ventilation

Maintaining a constant airway pressure maybe advantageous for several reasons. Con-stant airway pressure facilitates alveolar re-cruitment; enhances diffusion of gases; al-lows alveolar units with slow time constantsto fill, preventing over-distension of alveoli;and augments collateral ventilation.31

Van Allen et al32 noted that complete ob-struction of an airway unit did not always re-sult in collapse of the alveoli and, therefore,hypothesized that alternative pathways mustexist. The pores of Kohn, located in the septaof the alveoli and open only during inspira-tion,33 first were believed to be responsible.However, two additional pathways were latercredited with playing a role: (1) Lambert’scanals connect terminal and respiratory bron-chioles with adjacent peribronchial alveoli,and (2) channels of Martin interconnect respi-ratory bronchioles and serve to bypass themain pathway (Figure 4).34

In normal, healthy lungs, collateral venti-lation may barely occur at the functionalresidual capacity level, i.e. end exhalation.However, alternative pathways may beopened at a higher lung volume.35 The roleof alternative pathways in healthy lungs isvery limited; but in disease states may be im-

Figure 4. Collateral channels of ventilation.


portant.36 Although collateral ventilation istypically lost in pulmonary edema, collateralpathways may be reopened and oxygena-tion improved by increasing functional resid-ual capacity.36 Sustained airway pressure,rather than intermittent periods of airwaypressure, is more beneficial in the edema-tous, collapsed lung. Sustained breathsmaintain a constant airway pressure and al-low collateral channels to assist in producingventilation. Collateral ventilation efficiencydrops as respiratory frequency increases.37

Airway pressure release ventilation usesthese concepts, maintaining sufficient airwaypressure for an adequate duration to opencollapsed alveoli, thus improving recruit-ment of alveoli and increasing oxygenation.

� Advantages

In patients with severe acute respiratory fail-ure, the use of APRV results in significantlylower peak Paw, when compared with con-tinuous positive pressure ventilation (CPPV).Lower airway pressures are thought to be as-sociated with a reduced risk of ventilator-as-sociated lung injury.14 Further, APRV re-quires lower minute ventilation than CPPV,suggesting less dead-space ventilation.23

Studies of patients with ALI have shown thatAPRV supports oxygenation and ventilation,while producing lower peak Paw than volumeassist-control ventilation5 and intermittentmandatory ventilation.8,11 Similarly, animalstudies of injured lungs suggest lower airwaypressure, reduced dead space ventilation,and improved oxygenation and ventilation,when compared with intermittent positivepressure ventilation.2

Airway pressure release ventilation re-cruits lung units by optimizing end-inspira-tory lung volume. Ideally, the end-inspira-tory pressure, which equates to P High orplateau pressure, should be kept beneath 35cm of water pressure.9 This protective lungstrategy has several positive effects. First, thepreset pressure limit prevents, or limits,over-distension of alveoli and high-volumelung injury. Second, APRV affects tidal venti-lation by decreasing rather than increasingairway pressure. Decreasing lung volume forventilation further limits air space over-dis-tension and the potential for high-volume

lung injury. Third, maintaining airway pres-sure optimizes recruitment and prevents orlimits low-volume lung injury by avoidingthe repetitious opening of alveoli.14

High-volume lung injury occurs as a re-sult of tidal ventilation above the upper in-flection point of the pressure-volume curve.Low-volume lung injury results from ventila-tion beginning beneath the lower inflectionpoint.17 Airway pressure release ventilationbegins on the pressure-volume curve be-tween these two points and uses a release,not an increase, of pressure from its base-line. Therefore, oxygenation and ventilationoccur predominantly within the upper andlower inflection points (Figure 5).

Calzia and Radermacher,38 in their 10-yearliterature review of APRV, were unable todocument any severe adverse effects ofAPRV and BIPAP on cardio-circulatory func-tion. One case report39 demonstrated an in-crease in cardiac output and blood pressurewhen APRV was used. Further, the authorssuggested that it should be considered as analternative therapy to pharmacologic or fluidtherapy in the hemodynamically compro-mised, mechanically ventilated patient.

Animal studies indicate that APRV doesnot compromise circulatory function and tis-sue oxygenation, whereas CPPV can impaircardiovascular function significantly.40 Spon-taneous ventilation has a positive effect onthe venous thoracic pump mechanism. Sup-pressing spontaneous breathing during CPPVcan compromise cardiac function by decreas-ing venous return, thus cardiac output.4

The main advantage of APRV is that it al-lows for spontaneous breathing to occur atany point in the respiratory cycle. Depend-ing on the patient’s need, spontaneousbreathing may involve only exhalation, onlyinspiration, or both.

The distribution of ventilation is signifi-cantly different when a spontaneous breathis compared with a mechanically controlledor assisted breath. Spontaneous breaths tendto improve ventilation-perfusion matchingby preferentially aerating well-perfused, de-pendent lung regions. Mechanically deliv-ered breaths primarily ventilate areas awayfrom those receiving maximal blood flow.This phenomenon is consistent with earlierresearch, which demonstrated that sponta-neous ventilation opens more alveoli, im-


proves regional gas exchange, and reducesatelectasis.41

Putensen et al.42,43 found that by allowingunsupported, spontaneous breathing (usingBIPAP or APRV) in both dogs42 and humans43

with ALI, ventilation-perfusion matching im-proved, as seen by a marked decrease in in-trapulmonary shunt. In humans, however,pressure support ventilation preferentiallyventilated poorly or nonperfused lung unitsthat already were well ventilated. Further-more, pressure support ventilation did notconvert shunted areas to normal ventilation-perfusion units.43

Decreased need for sedation use or neuro-muscular blockade use with APRV7,8 and BI-PAP25 has been reported. Judicious use of se-dation and paralysis in the mechanicallyventilated patient was recommended at theAmerican-European Consensus Conference onARDS.21 Unintentional, prolonged paralysis isnow recognized as a complication of the long-term use of paralytics. In addition, a paralyzeddiaphragm moves very differently with posi-tive pressure ventilation compared with an ac-tive contraction. The paralyzed diaphragm isdisplaced preferentially along the path of leastresistance, that is, into the abdomen of the nondependent region. This displacement leads tofavored ventilation of the nondependent lung

regions.41 All of this contributes to both ventila-tion-perfusion mismatch and possible over-dis-tension of healthy alveoli, leading to furtherhypoxemia (Table 3).

� Disadvantages

Consistent with other pressure-targetedmodes of ventilation, APRV is affected bychanges in lung compliance and/or resis-tance. Clinicians need to identify the scenar-ios that affect lung volume and monitor pa-tients for changes in their tidal volumes.

Because APRV is time-cycled, synchronywith the patient’s spontaneous respirationdoes not occur. If a release phase is not syn-chronous with the patient’s effort, discomfortmay result. However, because APRV has a dy-namic pneumatic system, inspiration and exha-lation are facilitated at any time. Dyssynchronywith APRV has not been identified as a prob-lem in the majority of the literature to date.5,6,11

As with any new technology, staff stressand subsequent increased risk to the patientmay be noted with the implementation ofAPRV. Adequate and appropriate on-site train-ing, coupled with off-site support services andbackup, will help resolve some of the stressand decrease the risks associated with the in-

Figure 5. Pressure-volume curve. Conceptual drawing of airway pressure release ventilation occurring below theupper inflection point and above the lower inflection point, achieving goals of lung protective strategies.


troduction of APRV. Transferring patients tosubacute areas as their disease processes im-prove, may cause these issues to be revisited.Further, these areas may not have access toventilators capable of delivering APRV, whichwill require switching the patient to a differentmode of ventilation. Similarly, traveling toother departments (e.g., radiology, hyperbaricoxygenation chamber) may require temporarydiscontinuation of APRV, causing undue anxi-ety or discomfort in some patients.

Finally, only limited research exists re-garding the clinical practice of APRV and itscomparison with other modes. For example,APRV is suitable for ventilator weaning,though its superiority to mainstream modes,e.g. pressure support ventilation, has notbeen demonstrated. Weaning, in general,lacks a consensus, and this absence of ab-solutes exemplifies the great confusionwithin clinical practice and within the studyof mechanical ventilation (Table 3).

� Application of APRV

Little direction on the application of APRV canbe found in the literature, other than as sug-gested by vendors and limited study proto-cols. However, based on an understanding ofpulmonary physiology and pathophysiology,coupled with the theoretical understanding ofmechanical ventilation44 and current recom-mendations from consensus conferences,9,21

the following technique has evolved.When changing a patient’s mode of venti-

lation to APRV, the initial settings are partlydeduced from values of conventional ventila-tion. The clinician converts the plateau pres-sure of the conventional mode to P High andseeks an expired minute ventilation of 2 to 3L/minute, less than when on conventionalventilation. This is accomplished by setting PHigh at the plateau pressure, with a ceilinglevel for the P High normally at 35 cm of wa-ter pressure. P Low is set at 0 cm of waterpressure. A P Low of zero produces minimalexpiratory resistance, thus accelerating expi-ratory flow rates, facilitating rapid pressuredrops. T High is set at a minimum of 4.0 sec-onds. A T High of less than 4.0 seconds be-gins to impact mean Paw negatively. T Low isset between 0.5 and 1.0 seconds (often at 0.8seconds). With these settings (P High = 35 cmof water pressure, P Low = 0 cm of water

pressure, T High = 4.0 seconds, T Low = 0.8seconds), the mean Paw will equal 29.2 cm ofwater pressure. It is not possible for conven-tional volume targeted modes to maintain amean Paw of 29 cm of water pressure and limitthe peak or plateau pressures to 35 cm of wa-ter pressure, and still produce sufficient tidalventilation.44

Application of APRV to newly intubatedpatients usually involves using standard pa-rameters and adjusting the settings accord-ingly. Commonly, in the patient with moder-ate to severe ALI we default to P High/P Lowof 35/0 cm of water pressure and T High/TLow of 4.0/0.8 seconds and allow sponta-neous breathing to take place.44

When attempting to avoid alveolar over-distension, the clinician must be cognizantof the plateau pressure, as this is the bestclinically available estimate of average alveo-lar pressure.9 Although based primarily onanimal data, a plateau pressure (or P High)greater than 35 cm of water pressure is asso-ciated with lung injury and, therefore,should be kept beneath this level.

Rarely, an elevated P High (40–45 cm ofwater pressure) may be indicated, especially

TABLE 3 � Advantages and Potential Disadvantagesof Airway Pressure Release Ventilation

AAddvvaannttaaggeess

1. Lower Paw for a given tidal volume comparedwith volume-targeted modes, e.g., AC, SIMV

2. Lower minute ventilation, i.e., less dead spaceventilation

3. Limited adverse effects on cardio-circulatoryfunction

4. Spontaneous breathing possible throughoutentire ventilatory cycle

5. Decreased sedation use6. Near elimination of neuromuscular blockade

usePPootteennttiiaall DDiissaaddvvaannttaaggeess

1. Volumes change with alteration in lungcompliance and resistance

2. Process of integrating new technology3. Limited access to technology capable of

delivering APRV4. Limited research and clinical experience

Paw � airway pressure; A-C � assist control; SIMV � synchronized intermittent mandatory ventilation; APRV � airway pressure release ventilation.


for patients with low-compliance respiratorysystems, (e.g., individuals with morbid obe-sity, abdominal distension, or chest walledema), either for the purpose of oxygena-tion or ventilation.44 Although not optimal,the increased P High would be less than thepressure generated by conventional modesto produce a similar response.22

The P Low of zero is selected because min-imal resistance to exhalation is the goal.Higher pressures may impede expiratory gasflow during passive lung recoil. The validconcern of collapsing alveoli with a P Low ofzero is negated with the use of a short T Low(0.5–0.8 seconds) to maintain end expiratorylung volume.

The minimum T High duration is 4.0 sec-onds. The goal is to create a nearly continu-ous airway pressure level, which serves torecruit collapsed alveoli and maintain re-cruitment, thus optimizing oxygenation andcompliance. As a patient’s lung mechanicsimprove, T High is progressively lengthenedto 12 to 15 seconds, usually in 0.5 to 2.0 sec-ond increments.44 A further advantage of thelong T High is the reduction in the numberof opening and closings of the small airways,one of the mechanisms implicated in the de-velopment of iatrogenic ALI.14

The T Low probably is the most closelystudied of the 4 parameters. Early writings5,22

suggested a T Low of 1.5 seconds as thenorm, which allows for complete emptying ofthe lungs. A longer T Low (3.0–4.0 seconds)in animals with ALI was associated with a de-crease in arterial oxygenation and the accu-mulation of hemorrhagic fluid in the endotra-cheal tube.5 An excessively long T Lowencourages alveolar derecruitment, atelecta-sis, and airway closure during the releasephase. Alternatively, an insufficient T Low po-tentially may result in inadequate exhalation,leading to dead space ventilation, hypercap-nia, and hemodynamic compromise.45 In-deed, an appropriately timed T Low is vital.44

Optimal release time allows for adequateventilation while minimizing lung volumeloss. Essentially, release time should impedecomplete exhalation in the slower compart-ments of the lung (i.e., areas of high compli-ance or high resistance to exhalation) andgenerate regional intrinsic PEEP. Theoreti-cally, this will enhance alveolar recruitment.4,7

Calculation of T Low depends on expira-tory time constants (T), which are a productof the compliance of the respiratory system(CRS) and the resistance of the airways(RAW); that is, T = CRS � RAW.4,45 Low-compli-

Figure 6. Inspiratory and expiratory flow of gas in airway pressure release ventilation.44 In this example, T Lowterminates at 40% of the peak expiratory gas flow. Baseline airway pressure is then rapidly re-established. T High= 6.0 seconds; T Low = 0.8 seconds.


ance states, such as ARDS, will have lower(or shorter) expiratory time constants andtherefore a lower (or shorter) T Low. Highresistance diseases, such as asthma, willhave longer time constants and requirelonger release times.45 Determining the cor-rect multiple of time constants to calculate TLow is a challenge of future research.

In practice, however, the clinician does notcalculate the time constants for each patient,but rather relies on an approximation of therestriction of expiratory flow, as indicated bythe expiratory flow of gas waveform (Figure6). When expiratory flow falls to approxi-mately 25% to 50% of peak expiratory flow,the clinician stops the release time and allowsthe airway pressure to return to P High.44

The transition to APRV may not result ininstant improvement in oxygenation. Consis-tent with observations of inverse ratio venti-lation,4 the positive effects may take severalhours to be realized. It appears that the re-cruitment of alveoli occurs “one by one.”Sydow et al.7 demonstrated that the maximalbeneficial effect of APRV upon oxygenationoccurred 8 hours after implementation, withno further improvement after 16 hours. Inearlier studies, data were collected withinthe first 60 minutes after transition to APRVand thus the full effect of time on alveolarrecruitment was not appreciated.

� Weaning From APRV

The current technique of weaning from APRVis guided by general principles of weaningused in clinical practice today. Knowledge ofthe signs of respiratory failure, as well as exclu-sion or correction of contributing factors pre-venting successful weaning, such as excessivesecretions, bronchospasm, sepsis, anxiety, anddiameter of endotracheal tubes and other deadspace devices, are paramount. The approachin APRV is to maintain lung volume, improvingboth oxygenation and ventilation. As such,rarely does a specific point in time occur whenweaning is “officially” commenced.

Primarily, the method to reduce support isthrough manipulation of P High and T High. PHigh will be lowered 2 to 3 cm of water pres-sure at a time, and T High will be lengthenedin 0.5- to 2.0-second increments, dependingon patient tolerance. The goal is to arrive atstraight CPAP—usually at 12 cm of water pres-sure—and then the clinician either weansCPAP or simply extubates the patient at 6 to 12cm of water pressure. Before switching toCPAP, P High often is approximately 14 to 16cm of water pressure and T High is at 12 to 15seconds (Table 4).44 Patients with more severeforms of ALI or ARDS are weaned on a slowerbasis. Changes in mean Paw are monitoredclosely for their effect on oxygenation. Simi-

TABLE 4 � Example of Airway Pressure Release Ventilation Settings in an Uncomplicated Case of Acute Lung Injury43*

Calculated MeanP High T High P Low T Low Airway Pressure

(cm H2O) (seconds) (cm H2O) (seconds) (cm H2O)

35 4.0 0 0.8 29.2

33 4.5 0 0.8 28.0

30 5.0 0 0.8 25.9

28 5.5 0 0.8 24.4

26 6.0 0 0.8 22.9

23 7.0 0 0.8 20.6

20 8.0 0 0.8 18.2

18 10.0 0 0.8 16.7

15 12.0 0 0.8 14.1

*Following the final settings, the patient was transitioned to CPAP of 12 cm of water pressure.

CPAP � continuous positive airway pressure; Cm H2O � centimeters of water; P High � pressure high; T High � time high; Plow � pressure low; T low � time low.


larly, exhaled minute ventilation is tracked inconjunction with CO2 removal.

� Conclusion

Airway pressure release ventilation canmaintain oxygenation and ventilation at alevel comparable to CPPV. Airway pressurerelease ventilation is associated with signifi-cantly lower peak airway pressures anddead space ventilation. Airway pressure re-lease ventilation uses almost constant airwaypressure that not only facilitates alveolar re-cruitment but also sustains that recruitmentonce it has occurred. Spontaneous, unsup-ported breathing during APRV may occur atany point in the ventilatory cycle. Sponta-neous breathing is advantageous because itdecreases intrapulmonary shunting and im-proves venous return. The ability to avoidneuromuscular blockade and decreased useof sedation have resulted in fewer complica-tions and decreased drug costs. Finally, ven-tilator-associated lung injury, which can re-sult from both high- and low-volume lungventilation, may be balanced and averted.

Few clinicians believe that any single, iso-lated treatment can be responsible for a ma-jor improvement in the outcome for patientswith ARDS. Combination therapy is expectedto be the standard, including such conceptsas prone positioning and permissive hyper-capnia. Part of that therapy may include theventilator strategy of APRV, which incorpo-rates the advantages listed above. The au-thors believe that future research will supportthe use of APRV as the mode of choice forpatients with ALI and ARDS.

Acknowledgments

The authors thank Jill Kuramoto for her sug-gestions and constructive criticism, and KrisDorman for reviewing the manuscript.

References1. Downs JB, Stock MC. Airway pressure release

ventilation: a new concept in ventilatory sup-port. Crit Care Med. 1987;15:459–461.

2. Stock MC, Downs JB, Frolicher D. Airwaypressure release ventilation. Crit Care Med.1987;15:462–466.

3. Rasanen J. Airway pressure release ventila-tion. In: Tobin MJ, ed. Principles and Practice

of Mechanical Ventilation. New York: Mc-Graw-Hill, Inc.; 1994:341–348.

4. Burchardi H. New strategies in mechanicalventilation for acute lung injury. Eur Respir J.1996;9:1063–1072.

5. Stock MC, Downs JB. Airway pressure releaseventilation: a new approach to ventilatorysupport during acute lung injury. Respir CareClin N Am. 1987;32:517–524.

6. Garner W, Downs JB, Stock MC, Rasanen J.Airway pressure release ventilation (APRV): ahuman trial. Chest. 1988;94:779–781.

7. Sydow M, Burchardi H, Ephraim E, ZielmannS, Crozier TA. Long-term effects of two differ-ent ventilatory modes on oxygenation inacute lung injury: comparison of airway pres-sure release ventilation and volume-controlled inverse ratio ventilation. Am JRespir Crit Care Med. 1994;149:1550–1556.

8. Davis K, Johnson DJ, Branson RD, CampbellRS, Johannigman JA, Porembka D. Airwaypressure release ventilation. Arch Surg. 1993;128:1348–1352.

9. Slutsky AS (chairman). ACCP consensus con-ference: mechanical ventilation. Chest. 1993;104:1833–1859.

10. Rasanen J. IMPRV: Synchronized APRV, ormore [editorial]? Intensive Care Med. 1992;18:65–66.

11. Rasanen J, Cane RD, Downs JB, et al. Airwaypressure release ventilation during acute lunginjury: a prospective multicenter trial. CritCare Med. 1991;19:1234–1241.

12. Milberg JA, Davis DR, Stienberg KP, HudsonLD. Improved survival of patients withacute respiratory distress syndrome (ARDS):1983–1993. JAMA. 1995;273:306–309.

13. Brower RG, Matthay MA, Morris A, et al. Ven-tilation with lower tidal volumes as comparedwith traditional tidal volumes for acute lunginjury and the acute respiratory distress syn-drome. N Engl J Med. 2000;342:1301–1308.

14. Dreyfuss D, Saumon G. Ventilator-inducedlung injury: lessons from experimental stud-ies. Am J Respir Crit Care Med. 1998;157:294–323.

15. Amato MBP, Barbas CSV, Medeiros DM, et al.Beneficial effects of the “open lung ap-proach” with low distending pressures inacute respiratory distress syndrome: aprospective randomized study on mechanicalventilation. Am J Respir Crit Care Med. 1995;152:1835–1846.

16. Amato MBP, Barbas CSV, Medeiros DM, et al.Effect of a protective-ventilation strategy onmortality in the acute respiratory distress syn-drome. N Engl J Med. 1998;338:347–354.

17. Hickling KG. The pressure-volume curve isgreatly modified by recruitment: a mathemat-


ical model of ARDS lungs. Am J Respir CritCare Med. 1998;158:194–202.

18. Jonson B, Richard J-C, Straus C, Mancebo J,Lemaire R, Brochard L. Pressure-volumecurves and compliance in acute lung injury:evidence of recruitment above the lower in-flection point. Am J Respir Crit Care Med.1999;159:1172–1178.

19. Gattinoni L, Pelosi P, Crotti S, Valenza F. Ef-fects of positive end-expiratory pressure onregional distribution of tidal volume and re-cruitment in adult respiratory distress syn-drome. Am J Respir Crit Care Med. 1995;151:1807–1814.

20. Carney DE, Bredenberg CE, Schiller HJ, et al.The mechanism of lung volume change dur-ing mechanical ventilation. Am J Respir CritCare Med. 1999;160:1697–1702.

21. Artigas A, Bernard GR, Carlet J, et al. TheAmerican-European consensus conferenceon ARDS, part 2: ventilatory, pharmacologic,supportive therapy, study design strategiesand issues related to recovery and remodel-ing. Intensive Care Med. 1998;24:378–398.

22. Valentine DD, Hammond MD, Downs JB,Sears NJ, Sims WR. Distribution of ventilationand perfusion with different modes of me-chanical ventilation. Am Rev Respir Dis. 1991;143:1262–1266.

23. Cane RD, Peruzzi WT, Shapiro BA. Airwaypressure release ventilation in severe acuterespiratory failure. Chest. 1991;100:460–463

24. Baum M, Benzer H, Putensen C, Koller W,Putz G. Biphasic positive airway pressure (BI-PAP): a new form of assisted ventilation.Anaesthesist. 1989;38:432–458.

25. Hormann C, Baum M, Putensen C, Mutz NJ,Benzer H. Biphasic positive airway pressure(BIPAP): a new mode of ventilatory support.Eur J Anaesthesiol. 1994;11:37–42.

26. Rouby JJ, Ben Ameur M, Jawish D, et al. Con-tinuous positive airway pressure (CPAP) vs.intermittent mandatory pressure release ven-tilation (IMPRV) in patients with acute respi-ratory failure. Intensive Care Med. 1992;18:69–75.

27. Bratzke E, Downs JB, Smith RA. IntermittentCPAP: a new mode of ventilation during gen-eral anesthesia. Anesthesiology. 1998;89:334–340.

28. Puritan Bennett Company. Two ventilatingstrategies in one mode: BiLevel. St. Louis,MO: Puritan Bennett Company; 1999.

29. Petrof BJ, Legare M, Goldberg P, Milic-Emili J,Gottfried SB. Continuous positive airwaypressure reduces work of breathing and dys-pnea during weaning from mechanical venti-lation in severe chronic obstructive pul-monary disease. Am Rev Respir Dis. 1990;141:281–289.

30. Smith RA, Smith DB. Does airway pressurerelease ventilation alter lung function afteracute lung injury? Chest. 1995;107:805–808.

31. Davis K Jr., Branson RD, Campbell RS,Porembka DT. Comparison of volume con-trol and pressure control ventilation: is flowwaveform the difference? J Trauma. 1996;41: 808–814.

32. Van Allen CM, Lindskog GE, Richter, HG. Col-lateral respiration: Transfer of air collaterallybetween pulmonary lobules. J Clinical Invest.1941;10:559.

33. Brashers VL, Davey SS. Alterations of pul-monary functions. In: McCance KL, HuetherSE, eds. Pathophysiology: The Biologic Basisfor Disease in Adults and Children. 3rd ed. StLouis, MO; Mosby; 1998:1165–1166.

34. Corrin B. Pathology of the Lungs. New York,NY: Churchill Livingstone; 2000:4.

35. Terry PB, Traystman RJ, Newball HH, BatraG, Menkes HA. Collateral ventilation in man.N Engl J Med. 1978;298:10–15.

36. Delaunois L. Anatomy and physiology of col-lateral respiratory pathways. Eur Respir J. 1989;2:893–904.

37. Menkes HA, Traystman RJ. Collateral venti-lation. Am Rev Respir Dis. 1977;116:287–309.

38. Calzia E, Radermacher P. Airway pressure re-lease ventilation and biphasic positive airwaypressure: a 10-year literature review. ClinicalIntensive Care. 1997;8:296–301.

39. Falkenhain SK, Reilley TE, Gregory JS. Im-provement in cardiac output during airwaypressure release ventilation. Crit Care Med.1992;20:1358–1360.

40. Rasanen J, Downs JB, Stock MC. Cardiovascu-lar effects of conventional positive pressureventilation and airway pressure release venti-lation. Chest. 1988;3:911–915.

41. Froese AB, Bryan AC. Effects of anesthesiaand paralysis on diaphragmatic mechanics inman. Anesthesiology. 1974;41:242–255.

42. Putensen C, Rasanen J, Lopez FA. Ventilation-perfusion distributions during mechanicalventilation with superimposed spontaneousbreathing in canine lung injury. Am J RespirCrit Care Med. 1994;150:101–108.

43. Putensen C, Norbert JM, Putensen-Himmer G,Zinserling J. Spontaneous breathing duringventilatory support improves ventilation-per-fusion distributions in patients with acute res-piratory distress syndrome. Am J Respir CritCare Med. 1999;159:1241–1248.

44. Habashi NM. APRV: Principles and Practice.Luebeck, Germany: Draegerwerk. In press.

45. Martin LD, Wetzel RC. Optimal release timeduring airway pressure release ventilation inneonatal sheep. Crit Care Med. 1994;22:486–493.







Doc 13 –

PLV

DOI 10.1378/chest.108.4.1041 1995;108;1041-1048Chest

Jacques-Andre Romand, Weizhong Shi and Michael R. Pinsky Injury

LungPressure Ventilation During Acute Cardiopulmonary Effects of Positive



) ISSN:0012-3692http://chestjournal.chestpubs.org/site/misc/reprints.xhtml(without the prior written permission of the copyright holder.

distributedNo part of this article or PDF may be reproduced or reserved.3300 Dundee Road, Northbrook, IL 60062. All rights

Copyright 1995 by the American College of Chest Physicians,Physicians. It has been published monthly since 1935.

is the official journal of the American College of ChestCHEST




Cardiopulmonary Effects of PositivePressure Ventilation During Acute LungInjury*Jacques-Andre Romand, MD;f Weizhong Shi, MD; andMichael R. Pinsky, MD, FCCP

Study objectives: To assess the gas exchange andhemodynamic effects of pressure-limited ventilation(PLV) strategies in acute lung injury (ALI). We hy¬pothesized that in ALI, the reduction ofplateau airwaypressure (Paw) would be associated with less alveolaroverdistention and thus have better hemodynamic andgas exchange characteristics than larger tidal volume(Vt) ventilation.Setting: Laboratory.Design: Prospective time-controlled sequential animalstudy.Measurements: Right atrial, pulmonary artery, leftatrial, arterial, lateral pleural (Ppl), and pericardial(Ppc) pressures, Paw, ventricular stroke volume, meanexpired CO2, and arterial and mixed venous oxygencontents. Airway resistance and static lung compliancewere also measured.Interventions: Intermittent positive pressure ventila¬tion (IPPV) given before (control) and after inductionofALI by oleic acid infusion (0.1 mL/kg). IPPV at FIo2of 1, Vt of 12 mL/kg, and frequency adjusted to main¬tain normocarbia. ALI PLV was given during ALI anddefined as that Vt which gave a similar plateau Paw tothat of control IPPV. High-frequency jet ventilation(HFJV) and ALI HFJV were also given and defined as

frequency within 10% of heart rate and mean Pawsimilar to that during control IPPV.

Results: After ALI, static lung compliance, PaC>2, andpH decreased, whereas airway resistance and PaCC>2increased. For a constant lung volume, Ppl and Ppcwere not different between control and ALI. Both ab¬solute dead space (Vd) and intrapulmonary shuntfraction increased afterALI, but absoluteVd was lowerwith ALI PLV andALI HFJVwhen compared with ALIIPPV. Ventilation did not alter hemodynamics duringALI.Conclusions: Changes in lung volume determine Ppcand Ppl. PLV strategies do not alterhemodynamics butresult in less of an increase in Vd/Vt than would bepredicted from the obligatory decrease in Vt.

(CHEST 1995; 108:1041-48)

ALI=acute lung injury; HFJV=high-frequency jet venti¬lation; IPPV=intermittent positive pressure ventilation;Paw=airway pressure; PeCC>2=mean expired C02;PEEP=positive end-expiratory pressure; PLV=pressure-limitedventilation; Ppc=pericardialpressure; Ppl=pleuralpressure; Qpa=pulmonary artery flow; Qs/Qt=intrapul-monaryshunt fraction; Raw=airway resistance; SV=strokevolume; VD=dead space; Vr=tidal volume

Key words: ARDS; dead space; dog model; heart-lung in¬

teraction; mechanical ventilation; pericardial pressure; pleu¬ral pressure

HPhe ventilatory support of patients with acute lung¦*¦ injury (ALI) has evolved over recent years as our

understanding of the distribution of lung injury and itsinteractions with positive-pressure ventilation has in¬creased. The distribution of lung consolidation in pa¬tients with ALI is nonhomogeneous1 with aerated lungunits displaying normal specific compliance.2 Thus, as

airway pressure (Paw) increases, aerated lung units inpatients with ALI expand to the same extent as would

*From the Cardiopulmonary Research Laboratory, Department ofAnesthesiology and Critical Care Medicine, University of Pitts¬burgh.fCurrently at the Departement d'Anesthesiologie des Soins Inten¬sify de Chirugie, Hopital Cantonal Universitaire de Geneve,Geneva, Switzerland.This study was supported in part by a research award from theDepartment ofAnesthesiology and Critical Care Medicine and bythe Veterans Affairs Medical Center.

Manuscript received January 25,1995; revision accepted March 31.

normal lungs for the same increase in Paw. Because thetotal amount of aerated and recruitable lung unitsavailable to be ventilated is reduced in patients withALI, tidal volumes (Vts) that would otherwise be nor¬

mal in non-ALI condition will overdistend smaller to¬tal aerated lung units in these patients. One manifes¬tation ofthis regional overdistention will be an increasein end-inspiratory plateau Pawwhen otherwise normalVts are given. Clinically, increased peak inspiratoryPaw is commonly seen in ventilated patients with ALIand probably reflects regional alveolar overdistention.Assuming that high plateau Paw occurs during me¬

chanical ventilation of patients with ALI, severalpathophysiologic effects may be seen. First, markedoverdistention oflung units will decrease their regionalblood flow because of increasing regional pulmonaryvascular resistance, and, thus, dead space (Vd) venti-

CHEST / 108 / 4 / OCTOBER, 1995 1041



lation will increase. Second, such regional increases inaerated lung pulmonary vascular resistance will divertblood flow to nonoverdistended consolidated or col¬lapsed lung units, thereby increasing shunt blood flow.3Third, to the extent that increases in local lung units

compress the cardiac fossa, local intrathoracic pressurewill rise, which may decrease venous return and thuscardiac output.4"9 Finally, repetitive overexpansion ofaerated lung units will injure respiratory epithelial tightjunctions, inducing capillary leaks, ALI, and even

death.10'11Based on all these considerations, increased interest

has developed in minimizing lung overdistention in

patients with ALI while maintaining adequate arterialoxygenation.31213 Numerous novel ventilatory strate¬

gies have been developed to address this problem,14"16all ofwhich involve the use ofa limitedVt breath whilemaintaining mean Paw elevations. It is not clear, how¬ever, what effect pressure-limited ventilation (PLV)has on shunt, Vd, or cardiovascular performance dur¬ing ALI.

Accordingly, we studied the hemodynamic and gasexchange effects of two different types of PLV as

compared with fixed Vt positive-pressure ventilation inan animal model of ALL Different ventilatory strate¬

gies can be used to limit end-inspiratory Paw whilemaintaining mean Paw at a defined level. High-frequency jet ventilation (HFJV) accomplishes this bycombining high inspiratory gas flow and frequencywith small Vt, whereas, low-frequency PLV accom¬

plishes this by decreasing inspiratory flow rate and Vt.It is not clear, however, ifthese two different strategieshave similar hemodynamic and gas exchange effects,because both the degree ofchange oflung volume andintrathoracic pressure differ between these two formsof ventilation. Thus, we studied both HFJV and low-frequency PLV.

Materials and Methods

PreparationAfter approval of our protocol by the Animal Care and Use

Committee, seven male mongrel dogs weighing 18.3 to 25 kg (mean,23.3 kg) were anesthetized with IV pentobarbital sodium (30 mL/kg). Their tracheas were intubated with a cuffed endotracheal tube(9-mm-internal diameter; Hi-Lo National Catheter; Argyle, NY)equipped with a 2.4-mm-internal diameter jet ventilation injectionport (open 5 cm from the distal orifice) and an open port at the distalend for measuring Paw. The dogs were placed in the supine posi¬tion during the entire experiment. Anesthesia was maintained witha continuous IV infusion of pentobarbital sodium at a rate of 4

mg . kg-1 . fr1 supplemented by a bolus of 50 to 100 mg IV as

needed. Ventilation during the surgical procedure was provided at

a respiratory rate of20 breaths/min, a Vt of 10 mL/kg, and a forcedinspired oxygen of 1.0 (Siemens Servo 900 B Ventilator; Siemens;Elema AB, Sweden). Arterial blood gas values were monitored pe¬riodically (ABL-30; Radiometer; Copenhagen). Corrections in theacid-base balance during the surgical procedure were made by ad¬ministration of sodium bicarbonate IV to maintain a pH between7.35 and 7.45, and by increasing the Vt to 15 mL/kg and

subsequently the respiratory rate as necessary to maintain a PaC02between 35 and 40 mm Hg. A standard lead 2 ECG was used tomonitor the heart rate. A calibrated infrared CO2 detector (Cap-nometer 47210A; Hewlett Packard; Palo Alto, Calif) was connectedto the endotracheal tube to monitor the end-expiratoryCO2 andwasused as an initial guide to adjust the respiratory rate.A saline solution-filled polyethylene catheter with end and mul¬

tiple side holes was placed in the descending aorta and in the rightatrium via the peripheral cutdown sites to measure aortic and rightatrial pressures, respectively. A 7.5F balloon-tipped flow-directedthermodilution catheter, with an injection port 15 cm from the distalend, was placed in the pulmonary artery to measure the pulmonaryarterial pressure (Baxter-Edwards; Irvine, Calif). The blood tem¬

perature was monitored continuously via the thermistor of thepulmonary artery catheter, and temperature was maintained above35° C using external heating pads.A midline sternotomy was performed, and the heart was

suspended in a pericardial cradle. A saline solution-filled polyeth¬ylene catheter with end and multiple side holes was placed in theleft atrium via its appendage to measure the left atrial pressure. Acircumferential electromagnetic flow probe (Carolina MedicalElectronics; King, NC) was placed snugly around the root of thepulmonary artery. In two dogs, because of anatomic limitations, theflow probe could not be placed around the pulmonary artery so itwas placed around the aortic root. During steady state, mean aor¬

tic and pulmonary blood flow were assumed to be equal. The flowprobe signal was linear to ±5% over the range of flow studied. Zeropulmonary artery or aortic flow was taken as the diastolic plateauof the flow signal. Absolute pulmonary artery flow (Qpa) was

quantified in vivo during an apneic steady state by the thermodi¬lution technique (Edwards 9520 cardiac output computer; Amer¬ican Edwards Laboratory; Santa Ana, Calif) using the average ofthree 5-mL iced-saline solution injections (each value had to bewithin 10% of each other to be accepted).

Right or left ventricular stroke volume (SV) was derived by inte¬

gration ofthe respective flow signals. Cardiac output was calculatedas the product of SV and heart rate. A 10x1.5-cm thin-walled air-filled latex balloon attached to a polyethylene catheter with end andmultiple side holes was placed over the left lateral aspect ofthe heartin the long axis direction and secured with stay sutures to measurethe pericardial pressure (Ppc). The pericardium was then approx¬imated with multiple sutures. A second identical air-filled ballooncatheterwas positioned on the right lateral mid chest wall in the longaxis direction at the height ofthe atria and secured in place with staysutures to measure the pleural pressure (Ppl). Chest tubes were

placed bilaterally. The positions of all the catheters and balloonswere confirmed by palpation before chest closure. The left atrial,pericardial, and pleural balloon catheters and the flow-probe cablewere exteriorized. The sternum was approximated, and the fasciaand skin were closed in three layers to ensure an airtight seal.The chest tubes were connected to continuous suction at -15 cm

H20 and a positive end-expiratory pressure (PEEP) of 5 cm H2Owas added to ensure lung reexpansion. As previously described andvalidated,17 in situ pressure-volume curves were then generated foreach balloon, and the volume of air left in the balloon was alwayslower than the in situ stressed volume ofthe system. The chest tubeswere transiently occluded to assess any potential effects of suction

applied on the drainage system on Ppc and Ppl. No effects of chesttube clamping were observed. All pressures were referenced to themidthorax. The vascular catheters were connected to low-displace¬ment transducers (Gould Statham P-50; Gould Inc; Cleveland), andthe air-filled Paw, pericardial, and pleural catheters were connectedto high-sensitivity transducers (Bell and Howell 4-3271; Gould Inc).Airway, aortic, left and right atrial, Ppl, Ppc, pulmonary arterypressure, and Qpa were continuously recorded on an eight-channelstrip-chart recorder (Gould Inc), digitized on line (Advantage A to

D; Menlo Park, Calif), and stored on disk for subsequent analysis

1042 Clinical Investigations in Critical Care



(IBM AT and customized software). This surgical procedure tookapproximately 2.5 h.

At the end of the experiment, while still anesthetized, each an¬

imal was killed by an IV injection of potassium chloride and a

necropsy was performed. Correct placement of all catheters was

confirmed. The tips of the vascular catheters were then dissectedfree of surrounding structures to record the zero hydrostatic pres¬sure reference. All catheters were patent, and no leaks were foundin the balloon catheters. The lungs were then examined grossly toascertain the distribution and degree of pulmonary parenchymalinvolvement induced by oleic acid injection.

ProtocolAfter stabilization for 30 min, defined as hemodynamic stability

in the absence of bleeding, arrhythmias, or ongoing metabolic ac¬

idosis, the protocol was begun (usually 1 h after the completion ofsurgery). Mechanical ventilation was provided using a constant in¬spiratory flow pattern by a positive-pressure ventilator (Siemens 900B) during intermittent positive pressure ventilation (IPPV) runs andby a jet ventilator (Acutronic MK800; Medical Systems; Basel,Switzerland) during the HFJV runs. During HFJV, a one-way valve(exhalation only) was placed at the end of the endotracheal tube to

prevent entrainment of room air, which allowed us to deliver anexact amount of gas per HFJV inspiration as previously describedby our laboratory.14 HFJV was delivered asynchronously, but at a

frequency within 10% of the heart rate to minimize cardiovascularinstability.The protocol consisted of observing the effects of various types

of positive-pressure ventilation during control and ALI conditions.Hemodynamic variables were averaged over the entire ventilatorycycle and a minimum of three breaths was used to derive mean

pressures and SV. Muscle paralysis (vecuronium bromide, 0.01mg/kg IV) was induced at the beginning of each condition to abol¬ish spontaneous movement. The control condition consisted offoursequential ventilatory stages: (1) IPPV (IPPV 1); (2) HFJV; (3) a

second episode of IPPV (IPPV 2) to serve as a time control; and (4)apnea used as a baseline minimal heart-lung interaction. The ALIcondition consisted of the same ventilatory stages as with the con¬

trol plus an additional PLV stage? consisting of an IPPV-like run inwhich the Vt was adjusted downward until the plateau Paw was

similar to the IPPV plateau Paw during the control. These five ALIstages are referred to in the text as ALI IPPV 1, ALI PLV, ALIHFJV, ALI IPPV 2, and ALI apnea. IPPV was defined as a

frequency of 20 breaths/min, the inspiratory/expiratory (I/E) ratiowas 1/3, and the Vt was adjusted for a PaCC>2 between 35 and 45mm Hg, with an average Vt of 12 mL/kg. The plateau Paw wasdefined using the flow interrupter technique at end-inspiration.Mean Paw was also recorded during IPPV 1 and was used to de¬fine HFJV ventilator settings because estimates of plateau Pawduring HFJV are difficult to determine. During HFJV, the venti¬latory rate was fixed within 10% of the heart rate, the I/E ratio was1/4, and the driving pressure of the HFJV air flow was adjusted tomatch the mean Paw as during IPPV 1. Ventilatory settings duringIPPV 1 and 2 were identical. By using these three ventilatory modes,we could compare the effects of plateau Paw (IPPV vs PLV), lungvolume (IPPV vs PLV vs HFJV), and mean Paw (IPPV vs PLV andHFJV) on gas exchange and hemodynamics.At the beginning and end of control and ALI conditions, static

inflation compliance curves were generated in 100-mL incrementsup to 500 mL above the resting lung volume using a 1-L supersy-ringe. Airway resistance (Raw) was estimated at end-inspiratorylung volume using an end-inspiratory hold maneuver, wherein theratio ofthe immediate decline in airway pressure at end-inspirationto inspiratory gas flow reflected Raw at the end-inspiratory lungvolume. This maneuver also allowed for the measurement of pla¬teau Paw defined as Paw at 6 s of end-inspiratory hold. Thirty-sec¬

ond timed expired gas was collected during each mechanical ven¬

tilation step of the protocol in a 15-L polyester film (Mylar) plasticbag, which was then assayed for mean expired C02 (PeCC^). Theescaped volume was also measured during HFJV runs to calculatetrue Vt. Paired mixed venous and arterial O2 contents were alsomeasured using a co-oximeter adjusted for dog blood (Co-oximeterIL-282; Instrumentation Laboratories; Lexington, Mass). The ratioof Vd to Vt was estimated by the following formula:

(1) VD/VT=(PaC02-PeC02)/PaC02Intrapulmonary shunt fraction (Qs/Qt) was estimated by theformula:

(2) Qs/Qt=(Cc02-Ca02)/(Cc02-Cv02)where CcC>2, CaC>2, and CvC>2 were the O2 content of the pulmo¬nary capillary, systemic arterial, and mixed venous blood, respec¬tively, and assuming a fully saturated end-capillary blood sampleand a plasma oxygen solubility of0.003 mL O2 permm Hg of PaC>2.ALI was induced by injection ofoleic acid (0.1 mL/kg) in the right

atrium over 5 min after the solution was vigorously agitated (Vor¬tex; Fisher Scientific Industries; Bohemia, NY) for 30 s in 20 mLof 0.9% sodium chloride. The ALI stages of the protocol beganapproximately 90 min after induction of ALI and after hemody¬namic stabilization, as previously suggested.1819 During ALI IPPV,the respiratory rate usually had to be increased in an attempt tomaintain normocapnia. Measures for each stage were made onlyafter measures of PeCC>2 were stable for over 5 min and frequencywas held constant across the condition. During ALI HFJV, thesame settings as during the control stage of HFJV were used.Measurements taken during IPPV 2 and ALI IPPV 2 served astime controls for their respective conditions. Hemodynamic mea¬surements were performed after stabilization (usually after 3 min)at each ventilatory step for control and ALI conditions. Hemody¬namic variables were averaged over the ventilatory cycle takingapproximately ten beats. After induction of ALI, hypoxemia occa¬

sionally was severe enough to require supplemental PEEP tomaintain a minimal level of arterial oxygenation (PaO2>50 mm Hg)during the stabilization period and the intervals between the ven¬

tilatory stages. All measurements were taken after the applicationof supplemental PEEP was discontinued and hemodynamic stabi¬lization was achieved, usually in 1 or 2 min. Gas exchange measureswere not made during apneic steps of the protocol. In five dogs,frothy pulmonary edema occurred during the ALI conditionrequiring intermittent suctioning of the endotracheal tube andPEEP in between the protocol steps. The entire control and ALIsequences took approximately 20 min and 30 min, respectively, tocomplete. An infusion of dextran (60 g/L) was given IV during thecontrol condition at 2 to 4 mL . kg"1 . h"1 and during the ALI con¬dition at 10 mL . kg"1 . h"1 to maintain a constant apneic transmu¬ral left arterial pressure (defined as left arterial pressure minus Ppc).Statistical Analysis

Analysis was performed on group data by ventilatory stages andexperimental conditions using a two-way analysis of variance forrepeated measures. Post hoc analysis was done using the Scheffetest. Paired comparisons between control and ALI conditions forcompliance curves and Raw were done by a two-tailed t-test. SincePaC>2 values among conditions were not distributed normally, wecompared these values between conditions by a nonparametricWilcoxon signed rank test. A p value <0.05 was considered signif¬icant. All data are shown as mean±SD.

ResultsModel CharacteristicsThe model was stable throughout the control and

CHEST / 108 / 4 / OCTOBER, 1995 1043



Table 1.Ventilatory Effects of Positive-Pressure Ventilation*

StageVt,mL

Frequency,Beat/min

Plateau Paw,mm Hg

Mean Paw,mm Hg

Pa02,mm Hg

PaC02,mm Hg

Vd/Vt,PH

ControlIPPVHFJV

ALIALI IPPVALI PLVALI HFJV

288±20138±33f

285±20184+52*147±60f

21±0125±19f

24±436±6*125+19f

7.4±0.9

9.9±1.1**7.8±1.4

2.8+0.32.1±0.6f

4.4±1.1*§3.4±0.7*3.2±1.0*

542±43559±64

60±24*48+08*38±llf*

36-35:

42±3*53+8*44±8*

7.38±0.037.39±0.06

7.35 ±0.07*7.22 ±0.06*7.29+0.06*

19±769±9f

52±9*55±9*73±10f

*Data=mean±SD; for abbreviations see text.

fp<0.01 vs non-HFJV ventilatory modes.*p<0.02 ALI vs IPPV.

*p<0.05 ALI IPPV vs ALI PLV.

the ALI conditions. Besides the arterial pH, whichdemonstrated a moderate metabolic acidosis duringALI IPPV 2 compared with ALI IPPV 1 (7.23±0.12and 7.35±0.07, respectively; p<0.03), comparisons oftime control between IPPV 1 and IPPV 2 and betweenALI IPPV 1 and ALI IPPV 2 demonstrated no differ¬ence in any measured variable between pair time-controlled steps of the protocol. Thus, only the initialIPPV sequences were used for subsequent compari¬sons. Oleic acid-induced ALI was characterized by a

significant decrease in Pa02 and pH, and an increasein PaC02 for all stages (Table 1). Furthermore, statictotal thoracic compliance decreased from 62±15 to

100200 300 400 500600

Static lung inflation volume above FRC (ml)

32±8 mL/cm H2O (p<0.01) and Raw incrased com¬

pared with the control (7.1±1.4 to 18.6±8.6 cm

H20/min; p<0.01). Transpulmonary pressure, whichwas defined as Paw-Ppl, also increased for a givenchange in lung volume after ALI (p<0.0001) with ev¬

ery inflation step (Fig 1, top). However, Ppl and Ppcincreased by similar amounts during both the controland ALI conditions for similar increases in lungvolume(Fig 1, center and bottom, respectively), although theincrease in Ppc was proportionally less than theincrease in Ppl (p<0.05).Ventilation CharacteristicsWhen Vt was maintained similar to control IPPV

during ALI (ALI IPPV), both plateau and mean Pawincreased (by 25 and 57% respectively; Table 1); whenplateau Paw was maintained constant (ALI PLV), bothmean Paw and Vt decreased. ALI HFJV had a lowerVt than either IPPV or ALI IPPV and a lower meanPaw than ALI IPPV. During HFJV, the gas flow

Figure 1. Effects of static lung inflation on the following: top:transpulmonary pressure (Paw-Ppl); center: Ppl; and bottom: Ppefor both control (closed circles) and ALI (open diamonds). Data are

presented as mean±SD. Asterisk denotes differences betweencontrol and ALI (p<0.001).


100200 300 400500





IPPV HFJV AlilPPV ALIPLV ALIHFJV

Figure 2. Effect of both induction of ALI and different modes ofventilation on (Vt solid bars) and calculated (Vd open bars). See textfor description ofdifferent modes ofventilation. Data are presentedas mean±SD. Asterisk denotes p<0.01 ALI IPPV vs ALI PLV.

(driving pressure) had to be decreased to prevent hy-pocapnia, which resulted in a significant decrease inmean Paw compared with IPPV. PaC>2 was constant

during the different modes of ventilation within theseparate control and ALI conditions with the exceptionof ALI HFJV during which it was reduced (Table 1).The Pa02 did not change between the two ALI IPPVruns (60±24 to 48±8 mm Hg; p=0.17, paired t test;p=0.1, Wilcoxon signed rank test). After the inductionof ALI and despite increasing respiratory frequencyfrom 21 to 36 breaths/min, PaC02 increased by 17%during ALI IPPV (p<0.0001, ALI vs control). DuringALI PLV, PaC02 was higher by 30% than during ALIIPPV and ALI HFJV (p<0.05, respectively). Theseacute increases in PaC02 fully account for the de¬creases in pH. Despite increasing respiratory fre¬quency and Raw during ALI conditions, we observedno increase in end-expiratory pleural pressure duringany ventilatory mode. Compared with IPPV, both ALIIPPV and ALI PLV had doubled Vd/Vt (Table 1).Calculated physiologic Vd decreased significantly dur-

IPPV HFJV ALI IPPV ALIPlv ALI HFJV

Figure 3. Delta Ppl and Ppc before and after the induction ofALIfor inspiratory and expiratory pressure swings during positive pres¬sure ventilation for both Ppl (solid bars) and Ppc (dashed bars). Dataare expressed as mean±SD. Asterisk denotes difference betweencontrol and ALI (p<0.03), whereas dagger denotes differences be¬tween HFJV and non-HFJV ventilation values (p<0.04).

ing ALI PLV as compared with ALI IPPV (p<0.05)(Fig 2). The increase in Vd/Vt was 174 and 189% forALI IPPV and ALI PLV, respectively, when comparedwith IPPV. These changes in Vd/Vt were associatedwith a 17 and 47% increase in PaCC>2, respectively.Although Vd/Vt was greater during control HFJV as

compared with IPPV, it did not change during ALI,although PaCC>2 did increase 27% above control HFJVvalue.

Hemodynamic CharacteristicsALI was associated with a lower aortic pressure and

a higher transmural pulmonary artery pressure (pul¬monary artery minus Ppc or right ventricular ejectionpressure) than the control (Table 2). No differenceswere seen in SV, heart rate, right atrial pressure, or

transmural left atrial pressure (left atrial pressure mi¬nus Ppc, or left ventricular filling pressure) across dif¬ferent conditions or modes ofventilation. Interestingly,hemodynamic values during ventilatory modes were

not dissimilar to apneic values. The induction of ALIwas associated with an increase in Qs/Qt. There were

no differences, however, in Qs/Qt across ventilatoryTable 2.Hemodynamic Effects of Positive-Pressure Ventilation*

StageSV,mL

HR,Beats/min

CO,L/min

Pa,mm Hg

Platm,mm Hg

Ppatm,mm Hg

Qs/Qt,

ControlIPPVHFJVApnea

ALIALI VtALI PawALI HFJVALI apnea

14.5±3.213.3±2.113.2±3.0

13.8±3.113.7±3.717.3±5.116.0±4.7

128±30123±28125±27

129±22121±17129±20131±17

1.9:!1.6d1.6d

0.50.50.5

1.8±0.51.7±0.52.2±0.82.1±0.6

144±32140±29137±30

105±33f102±34f121±43f115±27f

7.7±3.76.5±2.88.6±5.0

5.4±3.46.1±4.16.3±3.36.2±4.0

14.8±3.613.8±3.715.9±3.5

16.4±2.7f16.6±2.7f19.3±5.1f17.9±3.3f

11+39±3

48±5f52+4f55±7f

*Data=mean±SD; HR=heart rate; CO=cardiac output; Pa=mean arterial pressure; Platm=left atrial pressure relative to pericardial pressure;Ppatm=pulmonary artery pressure relative to pericardial pressure.

fP<

CHEST /108 / 4 / OCTOBER, 1995 1045



modes in either control or ALI conditions.Figure 3 shows the maximal Ppl and Ppc changes

(delta) between peak-inspiratory and end-expiratorynadir values. We found no significant changes in deltaPpl or delta Ppc between IPPV and ALI IPPV. DeltaPpl and delta Ppc were smaller during ALI PLV thancontrol IPPV and ALI IPPV. The delta Ppl and Ppcdata were quantitatively similar to predicted changesfrom static compliance data (Fig 1). A delta pressure,defined as the maximum range of pressure changesover time measured at the nadir of the insufflationcurve for Ppl and Ppc was also analyzed during theHFJV runs as peak minus trough pressure values. Al¬though significantly different delta Ppl and Ppc valueswere seen in the control and ALI conditions, unlike inall other ventilatory models, delta Ppc was greater thandelta Ppl during HFJV runs (p<0.05).

Inspection of the lungs of all animals at necropsyrevealed diffuse patchy areas of hemorrhagic consoli¬dation, which tended to be greatest in dependent re¬

gions and in the lower lobes. One animal had primaryupper-lobe, lingula, and right middle-lobe consolida¬tion. All animals had clearly defined regions of aeratedlung units that expanded easily with slight increases in

airway pressure. A small quantity of watery blood-tinged pleural effusion was present in all animals andthe tracheas were filled with pink to red frothy fluid.

Discussion

This study demonstrates that positive-pressure ven¬

tilation has differential effects on gas exchange andhemodynamics, such that selective changes in eithermay occur with changes in ventilatory pattern. Overthe range of airway pressures analyzed, PLV strategieshad similar effects to volume-controlled IPPV on he¬modynamics and shunt in this canine model of ALLThis was true even though the Vt delivered with PLVwas almost half that of volume-controlled ventilation.Interestingly, following the induction ofALI, PLV was

associated with a lower physiologic Vd, thus minimiz¬

ing the potential alveolar hypoventilation-induced hy¬percarbia that would be predicted to occur. Indeed, fora 35% decrease in Vt, we measured only a 5% increasein Vd/Vt (Table 1). Furthermore, the primary deter¬minant of increases in both lateral chest Ppl and Ppcduring positive pressure ventilation is the Vt, and nei¬ther the compliance nor resistance characteristics ofthe lungs alter this relationship. Finally, ventilationduring ALI was not associated with increased end-ex¬piratory Ppl, suggesting that dynamic hyperinflationwas not a component of our model.A decrease in lung compliance has been shown to

decrease the transmission of Paw to the pleuralspace,5,6,20"22 potentially minimizing the effect of in¬creased airway pressure on right atrial pressure, andsubsequently on the pressure gradient for venous re¬

turn. These findings have been challenged by O'Quinet al23 who measured juxtacardiac Ppl and found thatthe fractional change ofPpl vs airway pressure was onlyslightly decreased after ALI in a canine model. Fur¬thermore, Potkin et al24 and Viquerat et al25 showed,in patients with ARDS, that the stepwise reduction incardiac chamber size with increasing PEEP was asso¬

ciated with a decrease in cardiac output. Similarly,Scharf and Ingram26 measured Ppl during incremen¬tal increases in PEEP before and after induction ofALI in a canine model and found that decreases incardiac output occurred independently of lung com¬

pliance, but were dependent on Ppl. Finally, Venus et

al,27 studying an ALI model in swine, demonstratedthat the transmission of the Paw to the pleural spacewas reduced by ALI, whereas, with a constant Vt, thechanges in Ppl were unaltered by changes in lungcompliance. These conclusions are in accordance withour present study. Neither Cabrera et al,8 analyzing theeffects ofchanges in airway pressure on Ppc in an ALIdog model, nor Pinsky and Guimond,19 examining theeffect of ALI on the transmission of PEEP to thepleural and pericardial spaces, examined the effects oflung volume itself on Ppc. This study allowed us to

unify these previous studies and to explain the appar¬ent discrepancies. The primary determinant of changein Ppl and Ppc during positive-pressure breathing isthe amount oflung inflation. IfVt is not held constant

throughout the course of an entire experiment, thenapparent decreases in transmission of Paw may incor¬

rectly be assumed to be due to decreased lungcompliance related to the ALI condition when, in fact,decreased Vt alone was responsible for this effect.We attribute the apparently higher compliance of

Ppc as opposed to Ppl, reflected by a significantly loweramplitude change in Ppc compared with Ppl duringHFJV conditions (Fig 3), to the effect of lung inflationon the heart. According to this model,19 increasing lungvolume would compress the heart limiting its size, suchthat the associated increase in intrathoracic pressurewould impede venous return, resulting in a decrease inthe right ventricular volume and a smaller increase in

Ppc. This is consistent with the hypothesis that whenlung volume changes are rapid enough, as with HFJV,mechanical heart-lung interactions are accentuatedbecause insufficient time is allowed for the blood toleave the heart when compared with "conventional"positive pressure breathing. In support of this hypoth¬esis, we saw that the swings in Ppc during ventilationincreased as frequency increased while Ppl swingswere unaltered.When we controlled the airway pressure by decreas¬

ing the Vt after induction ofALI (ALI PLV) (Table 2),we did not find any beneficial hemodynamic effects ofPLV compared with volume-limited positive-pressureventilation in the range of airway pressures explored in




our canine model. PLV was associated with smallerpleural and pericardial excursions during ventilation,however, and cardiac output tended to improve at theextremes oflow Vt ventilation (jet ventilation). It is notclear if these changes would have resulted in hemo¬dynamic differences if the intravascular volume statushad not been well maintained. We vigorously resusci¬tated these animals during ALI to keep left-side fillingpressures constant. This approach was occasionallyassociated with marked pulmonary edema formation.If resuscitative efforts had been more restrained, wemay speculate that decreased mean Ppl induced byPLV might have been associated with a higher venousreturn.The observation that the physiologic Vd is moder¬

ately affected by changes in Vt has already beendescribed28 and is attributed to a decrease in the an¬

atomic Vd in normal lungs. When nonhomogeneouslung injury conditions prevail, however, as in our

model, decreasing Vt is also likely to allow a bettermatching of ventilation and perfusion secondary todecreased overinflation. Because overdistention is a

potential cause of further lung injury, limiting the air¬way pressure during mechanical ventilation is a logicalventilatory strategy. The obligatory reduction in Vtusually induces alveolar hypoventilation and has givenrise to the term permissive hypercapnia to denote theinevitable increase in arterial Pco2.15'16 The goal ofventilatory therapy in patients with ALI is to maximizegas exchange while minimizing the detrimental effectsof positive-pressure breaths on the lungs. Our resultssuggest that PLV compared with large Vt positive-pressure breaths may be better tolerated than previ¬ously thought. The improved ventilatory efficiency dueto the decreases in physiologic Vd ventilation is alsoassociated with lower inspiratory-increase in Ppl. Thus,pressure-limited ventilation strategies may not worsen

hemodynamics or gas exchange.Limitations of the StudyOur model of ALI follows oleic acid infusion with

microembolism of lipids and diffuse endothelial injury.This may alter lung perfusion by mechanisms not as¬

sociated with changes in alveolar pressure. Endothelialinjury models, however, reflect the more common

sepsis-type lung injury seen clinically in patients withALL29 Moreover, IV oleic acid induced an ALI in ourcanine model with a nonhomogeneous distribution ofthe lesions, decreased lung compliance, pulmonaryventilation-perfusion mismatch, and systemic vasodi¬lation, mimicking many of the pathophysiologic events

occurring in patients with ALL Furthermore, even ifmean airway pressure did not increase to the levelsusually seen during human ALI, the lesions were se¬

vere enough to produced a 50% decrease in static lungcompliance, making the findings ofour model relevant.

Another potential limitation ofour studywas thatwevaried minute ventilation during our ALI conditions to

keep PaC02 as close to control values as possible. Thisresulted in a greater frequency of ventilation duringPLV than during conventional ventilation. Despite thismaneuver, we measured a significant increase inPaC02 during ALI conditions. If CO2 flux was not inequilibrium, then calculated Vd/Vt could be inaccu¬rate. We continuously monitored end-tidal CO2, how¬ever, and took measurements only after baseline shiftshad disappeared. Thus, although we may not have ad¬equate data to calculate CO2 excretion, the assump¬tions made in the calculation of Vd/Vt by the Bohrequation are still valid.Our study did not use PEEP, which is used in pa¬

tients to maintain adequate Pa02, because we wantedto analyze the direct effects of changing the ventilatoryparameters on hemodynamics and gas exchange.Clearly, increased levels of PEEP would distend fur¬ther aerated alveoli and, if anything, would have exag¬gerated the differences in physiologic Vd betweenpressure-limited and volume-limited ventilation. A fi¬nal limitation ofour study was that extreme hypoxemiaand marked decrease in airway compliance were pro¬duced by the oleic acid injection. This, however,resulted in only minor changes in the mean airwaypressure. The absence ofbeneficial effects ofpressure-controlled ventilation on hemodynamics may be re¬

lated to the relatively small decrease in plateau airwaypressure achieved.ACKNOWLEDGMENTS: The authors wish to thank Brian On-dulick for his technical assistance, F. Donald, MD, for reviewing thismanuscript, and Wendy J. Bouton for her help in preparing themanuscript.

References1 Gattinoni L, Mascheroni D, Torresin A, et al. Morphological re¬

sponse to positive end expiratory pressure in acute respiratoryfailure: computerized tomography study. Intensive Care Med1986; 12:137-42

2 Gattinoni L, Pesenti A, Avalli L, et al. Pressure-volume curve oftotal respiratory system in acute respiratory failure: computedtomographic scan study. Am Rev Respir Dis 1987; 136:730-36

3 Abraham E, Yoshihara G. Cardiorespiratory effects of pressurecontrolled inverse ratio ventilation in severe respiratory failure.Chest 1989; 96:1356-59

4 Morgan BC, Martin WE, Hornbein TF, et al. Hemodynamic ef¬fects of intermittent positive pressure respiration. Anesthesiology1966; 27:584-90

5 Chapin JC, Downs JB, Douglas ME, et al. Lung expansion, air¬way pressure transmission, and positive end-expiratory pressure.Arch Surg 1979; 114:1193-97

6 Cheney FW Jr., Martin WE. Effects of continuous positive-pressure ventilation on gas exchange in acute pulmonary edema.J Appl Physiol 1971; 30:378-81

7 Cournand A, Motley HL, Werko L, et al. Physiological studies ofthe effects of intermittent positive pressure breathing on cardiacoutput in man. Am J Physiol 1948; 152:162-74

8 Cabrera MR, Nakamura GE, Montague DA, et al. Effect of air¬

way pressure on pericardial pressure. Am Rev Respir Dis 1989;

CHEST /108 / 4 / OCTOBER, 1995 1047



140:659-679 Braunwald E, Binion JT, Morgan WL, et al. Alterations in cen¬

tral blood volume and cardiac output induced by positive pressurebreathing and counteracted by metaraminol (Aramine). Circ Res1957; 5:670-75

10 Dreyfuss D, Bassett G, Soler P, et al. Intermittent positive-pres¬sure hyperventilation with high inflation pressures produces pul¬monary microvascular injury in rats. Am Rev Respir Dis 1985;132:880-84

11 Tsuno K, Muira K, Takeya MM, et al. Histopathologicpulmonarychanges from mechanical ventilation at high peak airway pres¬sures. Am Rev Respir Dis 1991; 143:1115-20

12 Chan K, Abraham E. Effects of inverse ratio ventilation on car¬

diorespiratory parameters in severe respiratory failure. Chest1992; 102:1556-61

13 Lain DC, DiBenedetto R, Morris SL, et al. Pressure control in¬verse ration ventilation as a method to reduce peak inspiratorypressure andprovide adequate ventilation and oxygenation. Chest1989; 95:1081-88

14 Guntapalli KK, Pinsky MR, Marquez J, et al. Determinants ofventilation during high frequencyjet ventilation. J Crit Care 1987;2:93-100

15 Hickling KG, Henderson SJ, Jackson R. Low mortality associatedwith low volume pressure limited ventilation with permissive hy¬percapnia in severe adult respiratory distress syndrome. IntensiveCare Med 1990; 16:372-77

16 Hickling KG, Walsh J, Henderson S, et al. Low mortality rate inARDS using low-volume, pressure-limited ventilation with per¬missive hypercapnea: a prospective study. Crit Care Med 1994;22:1568-78

17 McMahon SM, Permutt S, Proctor DF. A model to evaluatepleural surface pressure measuring devices. J Appl Physiol 1969;27:886-89

18 Ashbaugh DG, Uzawa T. Respiratory and hemodynamic changes

after injection of free fatty acids. J Surg Res 1968; 8:417-2319 Pinsky MR, Guimond JG. The effects of positive end-expiratory

pressure on heart-lung interactions. J Crit Care 1991; 6:1-1120 Ellman H, Dembin H. Lack of adverse hemodynamic effects of

PEEP in patients with acute respiratory failure. Crit Care Med1982; 10:706-11

21 Mirro R, Busija D, Green R, et al. Relationship between mean

airway pressure, cardiac output and organ blood flow with normaland decreased respiratory compliance. J Pediatr 1987; 111:101-06

22 Jardin F, Genevray B, Brun-Ney D, et al. Influence of lung andchest wall compliances on transmission of airway pressure to thepleural space in critically ill patients. Chest 1985; 88:653-58

23 O'Quin RJ, Marini JJ, Culver BH, et al. Transmission of airwaypressure to pleural space during lung edema and chest wall re¬

striction. J Appl Physiol 1985; 59:1171-7724 Potkin RT, Hudson LD, Weaver LJ, et al. Effect of positive

end-expiratory pressure on right and left ventricular function in

patients with the adult respiratory distress syndrome. Am RevRespir Dis 1987; 135:307-11

25 Viquerat CE, Righetti A, Suter PM. Biventricular volumes andfunction in patients with adult respiratory distress syndromeventilated with PEEP. Chest 1983; 83:509-14

26 Scharf SM, Ingram RH Jr. Effects of decreasing lung compliancewith oleic acid on the cardiovascular response to PEEP. Am JPhysiol 1977; 233:H635-41

27 Venus B, Cohen LE, Smith RA. Hemodynamics and intrathoracicpressure transmission during controlled mechanical ventilationand positive end-expiratory pressure in normal and low compli¬ance lungs. Crit Care Med 1988; 16:686-90

28 Nunn JF, Hill DW. Respiratory dead space and arterial to end-tidal CO2 tension difference in anesthetized man. J Appl Physiol1960; 15:383-89

29 Schuster DP. ARDS: clinical lessons from the oleic acid model ofacute lung injury. Am J Respir Crit Care Med 1994; 149:245-60




DOI 10.1378/chest.108.4.1041 1995;108; 1041-1048Chest

Jacques-Andre Romand, Weizhong Shi and Michael R. PinskyAcute Lung Injury

Cardiopulmonary Effects of Positive Pressure Ventilation During

July 22, 2010This information is current as of

http://chestjournal.chestpubs.org/content/108/4/1041Updated Information and services can be found at:

Updated Information & Services

Citations

4/1041#related-urlshttp://chestjournal.chestpubs.org/content/108/articles:This article has been cited by 2 HighWire-hosted

Cited Bys

http://www.chestpubs.org/site/misc/reprints.xhtmlcan be found online at: Information about reproducing this article in parts (figures, tables) or in its entiretyPermissions & Licensing

http://www.chestpubs.org/site/misc/reprints.xhtmlInformation about ordering reprints can be found online:

Reprints

"Services" link to the right of the online article.Receive free e-mail alerts when new articles cite this article. To sign up, select the

Citation Alerts

PowerPoint slide format. See any online figure for directions. articles can be downloaded for teaching purposes inCHESTFigures that appear in Images in PowerPoint format



http://chestjournal.chestpubs.org/content/108/4/1041#related-urls




Date post:	15-Feb-2016
Category:	Documents
Upload:	consejeria-de-empleo-empresa-e-innovacion
View:	221 times
Download:	1 times

1600185-02-85-02 - Dados clínicos, Papers

Documents