Post on 04-Dec-2023
transcript
DOI 10.1378/chest.97.1.170 1990;97;170-181Chest
M S Niederman, D E Craven, A M Fein and D E Schultz Pneumonia in the critically ill hospitalized patient.
http://chestjournal.chestpubs.org/content/97/1/170.citation
can be found online on the World Wide Web at: The online version of this article, along with updated information and services
) ISSN:0012-3692http://chestjournal.chestpubs.org/site/misc/reprints.xhtml(without the prior written permission of the copyright holder.reserved. No part of this article or PDF may be reproduced or distributedChest Physicians, 3300 Dundee Road, Northbrook, IL 60062. All rights
ofbeen published monthly since 1935. Copyright1990by the American College is the official journal of the American College of Chest Physicians. It hasChest
© 1990 American College of Chest Physicians by guest on July 10, 2011chestjournal.chestpubs.orgDownloaded from
I
170 Pneumonia in Critically III Hospitalized Patient (Niederman et a!)
clinical conferencePneumonia in the Critically III Hospitalized PatientMichael S. Niedernian, M.D., FC.C.P;* Donald E. Craven, M.D.;t
Alan Al. Fein, M.D. , FC.G.P4 and Douglas F. Schultz, M.D.�
CASE PRESENTATION
Dr Douglas Schultz: A 59-year-old white man with a three-,nonth
history of systemic lupus erytheinatost,s and a history of insulin-
dependent diabetes mellitus was admitted to the hospital with a
three-day history offever to 38.8#{176}Cand dysuria. Evaluation revealed
a urinary tract infection with F coli and benign prostatic hypertro-
phy. One month later, the patient underwent transi,rethral resection
ofthe prostate. On the fourth postoperative da)� he developed fever
to 39.4#{176}C, shortness of breath, weakness, confusion, and reported
a slightly productive cough. His medications at the time included
prednisone, 10 mg dail); and NPII insulin, 20 units sulxs,taneosislv
daily.On initial examination, the patient was found to be mildly short
ofbreath with a respiratory rate of32, blood pressure of 120/76 mm1-1g. and a temperature of 39.3#{176}Crectall}: His on1y remarkal)le
physical findings were crackles ii� the posterior aspect of the right
lung in the upper two-thirds ofthe chest. The remainder ofhis lung
examination was unremarkable and flO new extrapulmonary findings
were noted.
Laboratory evaluation showed a white blood cell count of 8,600
with 85 percent po1�s and 10 percent bands. Blood glucose was 347
mg/(1l and arterial blood gases, while breathing room air, showed a
p11 of 7.4, Pco� 38 mm 11g. and Pu, of 56 mm 11g. A chest
radiograph (Fig 1) showed a right i,pper lobe infiltrate. Sputum was
evaluated b� Gram stain and showed copious white blood cells with
(;ram-negative rods. Sputum �md blood cultures were obtained.
The remainder of the lalxratory data was i,,iremarkahle.
The patient was treated with a presumptive diagnosis of �u)5(x’o-
mial Gram-negative pneumonia and therapy included intravenous
hydration, oxygen via a 40 Iwrcent Venti-mask, aH(l iI1traveIu)L�s
ceftazidime. lii addition, the patient received appropriate increases
in steroid therapy for stress, and insulin coverage for hyperglycemia.
Two days later, the patient was fom�nd to he more short of breath
and he was transferred to Winthrop-University hospitals intensive
care m�nit for further care. On arrival, the patient had a hhxxl
pressure of 70 palpable. Arterial blood gases, with the p�itieiit
breathing via 40 percent Venti-mask, showed a p11 of7.36, Pco� = 22
4Ditect�r Medical and Respiratory Intensive Care Unit, Pulmonary
and Critical Care Medicine Division, Winthrop-University Ilos-pital; Assistant Professor of Medicine, Health Science Center,State University of New York at Stony Brook.
tProfessor of Medicine and Microbiolog� Boston University Schoolof Medicine; Infectious Disease Division, Boston City Hospital,Boston.
lDirector, Pulmonary and Critical Care Medicine Division, Win-throp-University Hospital; Associate Professor ofMedicine, healthScience Center, State University of New York at Stony Brook.
§Fellow, Pulmonary and Critical Care Medicine, Winthrop-Univer-sity Hospital, Mineola, New York
Reprint requests: Dr. Nieder,nan, 222 Station Plaza North, Muwola,New York 11501
mm 11g. and Po2 = 42 mm Hg. At that time, his blood cultures from
the other hospital were reported to show two strains of Pseudomonas
aeruginosa. A repeat chest radiograph (Fig 2) showed diffuse
infiltration of the right lung and faint infiltration at the left base.The patient was then endotracheally intubated and placed on
mechanical ventilation. Antibiotic therapy was directed towards
Pseudonwnas aeruginoca with amikacin added to ceftazidime.
Bronchosecmpy was performed with a protected specimen brush and
the brush was cut into 1 ml of tryptic soy broth and cultured
quantitatively, revealing greater than 10� Pseudomonas aeruginosaper ml in pure culture.
The patient was treated with mechanical ventilation for a period
of three weeks. his therapy also included supplemental oxygen,
titrated to maintain oxygenation, and he had gradual improvement
in his oxygenation status and control of his respiratory infection.
After three weeks of mechanical ventilation, the patient was
successfully extubated. His course had been complicated by diar-
rhea, felt to be related to antibiotic therapy and tube feedings,
intermittent supraventricular tachycardia, tracheostomy for airway
care, and intermittent return of fresh blood from his nasogastric
tube, in spite of prophylaxis of intestinal bleeding with cimetidine.
At the time of discharge from the intensive care unit, the patient
was felt to have recovered from his pneumonia and all of his
antibiotics were discontinued. Sputum cultures continued to reveal
Pseudomonas aeruginosa, but no therapy was prescribed.
Approximately one week after discharge from the intensive care
L,nit, the patient again developed dyspnea, cough productive of
green sputum, and fever to 38.3#{176}C.A repeat chest radiograph (Fig
3) demonstrated persistent infiltrates in the right chest along with
FIGURE 1. Chest radiograph demonstrating right-sided pneumonia
developing postoperativel):
© 1990 American College of Chest Physicians by guest on July 10, 2011chestjournal.chestpubs.orgDownloaded from
FIGURE 2. Progression of pneumonia to a radiographic pattern
consistent with ARDS.
FIGURE 3. A second episode of pneumonia developed, with
infiltration in the left lung.
CHEST I 97 I 1 I JANUARY, 1990 171
a new left perihilar infiltrate. Fiberoptic I)ronchosc()py with pro-
tected specimen brush was repeated and again Pseudonumas
aeruginosa was recovered in greater than 1&’ organisms per ml.
However, at this time, the organism demonstrated a different
sensitivity pattern than the original organism and appropriate
antibiotic therapy was initiated. In addition, bronchoscopy with
bronchoalveolar lavage demonstrated Pneurnocystis carinii on
Giemsa staining and therapy with trimethoprim-sulfamethoxizole
was added to anti-Pseudomonal antibiotics. The patient again
required mechanical ventilation, but after a two-week course of
antibiotics, the patient had defervescence and clearing of his chest
radiograph and was eventually weaned offthe ventilator to ultimately
he discharged from the hospital.
Dr. Michael S. Niederman: Dr. Fein, can you describe
how systemic sepsL� may be complicated by ARDS?
Dr. Alan ?%I. Fein: Our patient exemplifies the close
clinical relationship which ties severe infection to the
adult respiratory distress syndrome (ARDS). The pa-
tient, a 59-year-old man with systemic lupus erythe-
matosus and diabetes mellitus, who was receiving
corticosteroids, presented with Gram-negative pneu-
monia which was uncontrolled by the prescribed
antibiotics and he then developed septic ARDS . ARDS
may be conceptualized as the pulmonary component
ofa generalized inflammatory injury, especially to the
endothelium. When initiated by severe infection, it
often progresses to involve in a sequential fashion
other major organ systems, including kidneys, gastro-
intestinal tract and liver leading to a syndrome of
multisystem organ failure. ARDS may have both direct
and indirect causes which often overlap. Pneumonia
is the most common direct cause of the adult respira-
tory distress syndrome, particularly if it involves a
large enough segment of parenchyma to severely
impair the mechanical and gas exchange function of
the lung.’ While viral pneumonia typically involves
the lung parenchyma in a diffuse fashion, most other
classes of organisms, including bacteria (most corn-
monly), fungi and mycobacteria, also have this capa-
1 Pnetnnocystis carinii has emerged as a common
cause of severe hypoxic respiratory failure, usually in
patients with AIDS.2
Infection may also initiate ARDS indirectly by
activation of systemic inflammation.3 Bacteria, viruses,
fungi, or even traumatized tissue induce a similar
systemic response characterized by hypermetabolism
(even in the absence of blood stream invasion) and
subsequent multiorgan failure.�’5 The link between
this systemic response, organ failure and infection is
under complex and active investigation. Recent work
has emphasized the central role of lymphocyte/mac-
rophage-derived tumor necrosis factor (cachectin) and
interleukin-1, in modulating the response to sepsis.6
Other potential mediators in this chain are activated
neutrophils, complement and the various components
of the thrombotic system, which may directly injure
tissue and amplify inflammation.3
Dr. Niederman: It is possible to predict which septic
patients will develop ARDS?
Dr. Fein: ARDS usually follows onset of the septic
syndrome within 72 hours.5 While isolated bacteremia
carries a relatively low risk for progressive lung injury
(6-15 percent), when combined with shock and throm-
bocytopenia, the risk rises dramatically to 65 and 46
percent, respective1y.�’7 Additional insults, such as
hypertransfusion, aspiration or burns will increase the
likelihood of sepsis leading to ARDS to approximately
50 percent.
At the present time, clinical signs in patients at risk
are the best predictors of progressive lung injury.4 In
a recent analysis combining injury severity score (155),
individual risk factors and initial oxygenation, Pepe et
al� reported that ISS, numbers of transfusions, the
presence of the septic syndrome and initial oxygena-
new tion best predicted ARDS, in that order. Our owndata suggest that severity of presenting hypoxemia
© 1990 American College of Chest Physicians by guest on July 10, 2011chestjournal.chestpubs.orgDownloaded from
172 Pneumonia in Critically III Hospitalized Patient (Niede,man et a!)
and metabolic acidosis are better predictors of pro-
gression to ARDS in septic patients than are any
measurements of mediators in serum.9
In many studies, the measurement ofplasma factors
to predict ARDS has been disappointing.4’5’9 While
various components of activated complement are ele-
vated in septic patients, none has been shown to
consistently predict progressive lung 1q12 Re-
cently, a prospective study by Langlois and Gawryl’3
demonstrated the terminal complement complex was
increased by an average of 100 percent in septic
patients who developed ARDS prior to the onset of
clinical lung injury. While interesting, these findings
await confirmation at other clinical centers. The neu-
trophil is also an important participant in the initial
phases of lung ry’4 Despite this, our own data
suggest that neither neutrophil turnover nor neutro-
phil chemoattractant activity can identify which septic
patients will develop ARDS.9
Dr. Niederman: How does sepsLs affect the outcome of
ARDS?
Dr. &in: Advances in the management of respiratory
failure have reduced the number of patients who
succumb to hypoxemia. Rather, deaths occurring 72
hours or more after the initial insult usually result
from uncontrolled pulmonary �4 Nosocomial
pneumonia is evident in more than 70 percent of
patients with ARDS.’5 Infection in a damaged lung
potentially influences the outcome by several mecha-
nisms. The additional septic burden further enhances
the release ofinflammatory mediators including tumor
necrosis factor (TNF) and interleukin-1, directly po-
tentiating organ damage and recruiting more neutro-
phils and macrophages to the lung. These effector
cells release proteases and oxidants which promote
colonization by micro-organisms and lead to disor-
dered repair of connective tissue matrix and fibrosis.
Among ARDS patients with infection, 67 percent died
compared to only 7 percent of non-infected patients
and those with nosocomial pneumonia may have an
even higher 14 Most infected patients have
failure ofother organ systems in addition to the lung.’6
It is interesting that following the recovery from
ARDS, our patient developed a second episode of
respiratory failure, with recovery ofboth Pseudomonas
aeruginosa and Pneumocystis carinii. Both of these
organisms are indicators of a markedly compromised
host. Host defense impairments are common in the
multiorgan failure syndrome and clearly influence
outcome.’7 In our patient these compounded the
already significant effects of systemic lupus erythe-
matosus, diabetes, and corticosteroid therapy. Proteo-
lytic enzymes, especially neutrophil elastase, are re-
leased into the alveolar lining fluid in ARDS.’8 In
addition to its effects on connective tissue matrix,
elastase may inhibit ciliary function, and degrade
immunoglobulin IgG and secretory IgA, thereby in-
hibiting bacterial clearance and phagocytosis.’7”9’�#{176}
Likewise, the formation of pulmonary edema fluid
may directly inhibit bacterial clearance by macro-
phages2’ and washout surfactant, which has antimicro-
bial activity.� While there is no mention in our patient
of liver dysfunction, this is a frequent accompaniment
of ARDS and multiorgan failure.ss Liver dysfunction
limits clearance of gut bacteria, endotoxin and circu-
lating inflammatory mediators such as TNF, thus
enhancing injury.tm latrogenic factors may also impair
host defenses. High inspired oxygen tensions used in
the patient’s treatment as well as cortiscosteroids,
impair the phagocytic function of macrophages and
neutrophils.2#{176}’�’m It is clear that failure of multiple
host defenses contributed to the second episode of
acute respiratory failure which he sustained.
Dr. Niederman: The patient being discussed today had
Gram-negative colonization of the tracheobronchial
tree and Gram-negative pneumonia, two findings that
are closely interrelated. In addition, the pathogenesis
of airway colonization is related to many of the host
impairments that our patient demonstrated. He had
two eposides of pneumonia, both acquired in the
hospital, and at the conclusion of the first episode, he
continued to grow Pseudomonas aeruginosa in his
lower airway, in the absence of clinical signs of
infection. This represented airway colonization which
had persisted after an initial pneumonic episode.
While still colonized with Pseudomonas aetvginosa,
he developed a second episode of pneumonia which
was due to both Pseudomonas aeruginosa and Pneu-
mocystis carinii. The features of interest in this case
are the relationship of airway colonization to the
subsequent occurrence of pneumonia, the mecha-
nisms that allow risk factors to lead to colonization of
the airway by Gram-negative bacteria, and reasons for
persistent lower airway colonization by Pseudomonas
aeruginosa.
Colonization ofboth the upper and lower respiratory
tract are frequently associated with the occurrence of
nosocomial pneumonia. Rates of Gram-negative cob-
nization of the oropharynx are directly related to the
severity of illness for a given patient. In a study by
Johanson and colleagues,� utilizing multiple cultures
of the oropharnyx, it was found that no more than 6
percent of normal individuals had oropharyngeal cob-
onization by Gram-negative bacteria. However, as
patients became progressively ill, the incidence of
Gram-negative colonization of the upper airway in-
creased, such that with multiple cultures, nearly three-
quarters of the sickest patients in the hospital were
colonized with enteric Gram-negative bacteria. The
relevance of this finding to the occurrence of pneu-
monia was shown in a follow-up study of213 intensive
care unit patients, 26 of whom developed nosocomial
© 1990 American College of Chest Physicians by guest on July 10, 2011chestjournal.chestpubs.orgDownloaded from
CHEST I 97 I 1 I JANUARY, 1990 173
Table 1-Risk Factorafor Gram-Negative AirwayColonization
Oropharynx Lower Respiratory Tract
Antibiotic therapy Antibiotic therapy
Azotemia* Chronic bronchitis
Coma Ciiary dysfunctionDiabetes Corticosteroid therapy
Hypotension Bronchiectasis
Advanced age Cystic fibrosis
Endotracheal intubation4 Endotracheal intubation4
Pre-existing lung disease Tracheostomy4
Smoking4 Malnutntion*
Surgery* Surgery*
Serious illness4 Acute lung injury
Malnutntion* Viral infection*
Gastric acid-neutralization SmokingGastric acid-neutralization
�ffisk factor has been shown to operate, at least in part, by altering
bacterial adherence.
pneumonia.� Among the patients with pneumonia, 22
of 26 had prior oropharyngeal colonization by Gram-
negative bacteria, indicating a high co-association
between upper airway colonization and the subse-
quent occurrence of parenchymal lung infection. The
reason for the close relationship between airway
colonization and pneumonia may be the result of two
factors. First, once the airway is colonized by Gram-
negative bacteria, these organisms are available for
aspiration into the tracheobronchial tree. In addition,
it is possible that patients who have become colonized
in the oropharynx have host impairments which pre-
dispose not only to upper airway colonization, but also
to lower airway colonization and subsequent pneu-
monia. Thus, colonization ofboth the upper and lower
respiratory tract can be viewed as a “marker” of a
seriously ill patient who has multiple impairments in
respiratory host defenses.
The patient discussed today has multiple risk factors
for both upper and lower airway colonization (Table
1). These included his serious degree of underlying
illness, recent surgery, the use of multiple antibiotics,
endotracheal intubation and tracheostomy, malnutri-
tion, the use of corticosteroids, recurrent hypotensive
episodes, and the use of histamine type-2 receptor
blocking agents for the prophylaxis of gastrointestinal
bleeding. Each of these factors increased the risk of
airway colonization through a variety of mechanisms.
Antibiotics can increase the risk of infection by
interfering with the normal flora in both the upper
and lower respiratory 17�� The normal upper
respiratory tract rarely harbors Gram-negative bacte-
ria and is usually colonized by Gram-positive organ-
isms, particularly anaerobes, which may “interfere”
with the growth of Gram-negative organisms. It has
been conceptualized that the normal flora ofthe upper
airway occupy bacterial binding sites in the oropha-
ryngeal muscosa and, thereby, block subsequent cob-
nization by Gram-negative bacteria. This type of
interference can be eliminated with the use of sys-
temic antibiotics. A similar mechanism may also appby
in the bower airway, although in most individuals the
lower airway does not have normal flora and is ster-
Endotracheal intubation can increase the risk of
upper and lower airway colonization and pneumonia.
Patients with tracheostomy have nearby a 70 percent
incidence ofnosocomial pneumonia and this is bargeby
related to two factors.3’ First, patients with tracheos-
tomy have the capacity for organisms to directly enter
the lower respiratory tract, thereby bypassing upper
airway defenses. In addition, the presence of a trach-
eostomy or an endotracheal tube may traumatize the
tracheobronchial mucosa and alter its integrity, making
it more susceptible to invasion and cobonization.�’�
Corticosteroids may predispose to colonization by
interfering with a variety of host defense mecha-
nisms.’7 Histamine type 2 (H2) bbockers similarly may
predispose to colonization by the mechanisms to be
discussed by Dr. Craven. In our own studies of
mechanically ventilated patients, it has been observed
that Pseudomonas species colonization of the bower
airway was more common in patients receiving H2-
blockers than in patients who did not receive this
therapy.�
The major pathogenetic mechanism that unifies
many of the risk factors for upper and bower airway
colonization involves the cell-cell interaction termed
bacterial adherence. At multiple mucosab sites
throughout the body, the binding of bacteria to the
epithelial surface has been demonstrated to be an
important mechanism that beads to colonization. In
the respiratory tract, bacterial adherencem has been
demonstrated to mediate Gram-negative colonization
of both the upper and the bower airway.�’ Many of
the risk factors for airway colonization have been
shown to act by enhancing the ability of respiratory
epithelial cells to bind bacteria, thereby albowing
bacteria to establish a foothold on the respiratory
mucosa (Table 1). Studies to date have demonstrated
that general surgery, renal failure, malnutrition and
cardiac bypass surgery all have the ability to make
oropharyngeab epithebiab cells express more binding
sites for Gram-negative bacteria.�
In studies by Johanson and cobleagnes,� it has been
demonstrated that patients whose bungs became cob-
onized after general surgery had a serial rise in the
ability of their buccal epithebial cells to bind bacteria.
Such a serial rise was not observed in patients who
did not develop colonization following general surgery.
The mechanism whereby serious illness and systemic
insult increase the number of binding sites on oral
epithelial cells has been evaluated by the same inves-
tigators. In these studies, it was shown that certain
© 1990 American College of Chest Physicians by guest on July 10, 2011chestjournal.chestpubs.orgDownloaded from
174 Pneumonia ni CritiCally HI Hospftahzed Patient (Niederman et a!)
critically ill patients elaborate proteases in their sail-
vary secretions which have the capacity to digest
fibronectin from the cell surface.� Fibronectin ap-
pears to be a blocking glycoprotein which covers
epitheliab cell receptors for Gram-negative bacteria.
With the release of oral proteases, fibronectin is
removed from the buccal epithelial surface, thereby
exposing more epithelial receptors for subsequent
bacterial binding. It is yet unknown whether similar
mechanisms operate in the lower respiratory tract.
However, it has been shown that patients with trach-
eostomy, colonized by Pseudomonas aeruginosa, had
an overall higher degree of tracheal cell binding
capacity than did tracheostomized patients who were
not colonized by Pseudomonas aeruginosa.32 In addi-
tion, patients with the highest degree of tracheal cell
adherence tended to have the highest levels of neutro-
phil elastase in their tracheal secretions.�#{176} Thus, it is
possible that proteases in high concentrations, in some
manner mediated the increased adherence to tracheal
epithelial cells that was seen in patients colonized in
the lower airway by Pseudomonas species.
Malnutrition is one host factor that has been related
to both colonization and increases in adherence to
both oral and tracheal epithelial cebls.32,�,37 Higuchi
and colleagues37 demonstrated, in an animal model,
that with progressive declines in animal weight, there
was a serial rise in the ability ofbuccal epithelial cells
to bind bacteria. In addition, among tracheostomy
patients, those with the most severe nutritional im-
pairment tended to have the highest degree of tracheal
cell adherence.� In patients with endotracheal intu-
bation who were receiving mechanical ventilation, it
has also been shown that patients who were severely
malnourished had a greaterlikelihood to develop lower
respiratory tract colonization by Pseudomonas species
than intubated patients who were better nourished)”
In studies of bacterial adherence, there has been
the consistent observation, in normal individuals and
in critically ill patients, that tracheal cells have a
greater capacity to bind Pseudomonas aeruginosa than
do buccal epithelial cells)’�”#{176}”'’The clinical relevance
ofthese findings may relate to the occurrence of tissue
“tropisms,” or preferences, ofPseudomonas aeruginosa
in the human respiratory tract. In other words, if
Pseudornonas aeruginosa binds to tracheal cells more
avidly than it binds to buccal cells, then possibly
Pseudomonas aeruginosa would preferentially colonize
the lower airway rather than the upper airway, if the
organism has access to both sites simultaneously. Such
a situation does exist in patients with tracheostomy
and in patients with endotracheal intubation. In both
of these populations, colonization studies have dem-
onstrated that Pseudomonas species colonized the
bower airway more frequently and persistently than
the upper airway)”� The patient discussed today did
not have bacterial adherence directly measured. How-
ever, it is likely that many clinical factors did increase
his tracheal cell capacity to bind Pseudomonas aeru-
ginosa and, thereby, led to persistent lower airway
colonization by Pseudomonas aeruginosa. The persist-
ence of colonization by Pseudomonas aeruginosa in
such patient may be related to multiple factors. The
patient was severely malnourished and if malnutrition
increased tracheal cell binding capacity for bacteria,
then Pseudomonas colonization would persist until
the malnutrition was reversed. In addition, with
colonization, airway inflammation is present and neu-
trophils in inflammatory secretions can release ebas-
tase which can, in turn, interfere with the local
protective function of IgA and further predispose to
airway cobonization.2#{176} A vicious circle of colonization
begetting more colonization is quite likely in this
circumstance.
In addition to more frequent and more persistent
colonization of the lower airway rather than the
oropharynx, the other expression of tissue “tropisms”
in the human respiratory tract may be that the lower
airway becomes colonized by Pseudomonas species
independent of the upper airway. In mechanically
ventilated patients, Pseudomonas species, in contrast
to other Gram-negative bacteria, have the capacity to
colonize the lower airway without first colonizing the
upper airway)’� This finding of primary tracheobron-
chial colonization has been observed by several inves-
tigators. Schwartz et al’#{176}have reported that while
Enterobacteriaceae entered the trachea after initial
oropharyngeal colonization, other organisms such as
Pseudomonas species rarely were found in the upper
airway before colonizing the lower airway. Pingleton
et al”' have also observed that intubated patients may
have primary tracheal colonization by Gram-negative
bacteria.
At least three different factors influence the adher-
ence of bacteria to cells. These include host cell
variables, bacterial variables, and micro-environmen-
tab factors.”� Among host cell variables, the most
important factors include the site of cellular origin
and the type of host from which the cells are taken.
Another cellular variable that may be important is the
presence of cilia. Recently, it has been shown that
Pseudomona3 aeruginosa can bind directly to ciiary
structures.�
Bacterial variables are important because certain
organisms have the capacity to bind epithelial cells
while others do not. Specific features that enhance an
organism’s ability to bind to epithelial cells include
the presence or absence of a capsule, the type of
surface appendages present, and the nature of the
exoproducts released by the bacteria.”�’47’� Certain
Pseudomonas species do have the capacity to bind
epithelial cells directly via their pili, which may serve
© 1990 American College of Chest Physicians by guest on July 10, 2011chestjournal.chestpubs.orgDownloaded from
CHEST I 97 1 1 1 JANUARY, 1990 175
as bacterial adhesions that attach directly to the
epithelial surface.”7
The micro-environment in which bacteria meet
epithelial cells Is also a determinant of adherence . For
example, the composition ofsputum, both its protease
and mucin components, can influence whether or not
bacteria bind epithelial cells. In addition, the pH of
the epitheliab surface may influence the ability of
bacteria to bind.”#{176}As mentioned previously, proteo-
bytic components of sputum may alter the bacterial
binding interaction.� Mucins in sputum can serve as
receptors for bacterial binding and thus have the
capacity to competitively inhibit bacterial binding if
they bind to bacteria, but not directly to cells. On the
other hand, if mucins attach directly the cellular
surface, then they may act as a receptor “bridge” for
bacteria and further increase the capacity of the
epithelium to bind baeteria.’� This may be particularly
important in patients who have impaired mucociliary
clearance where large quantities ofmucins may adhere
to the epithelial surface and, thereby, predispose to
colonization.
Based on our understanding ofbacterial adherence,
it may become possible in the future to devise
prophylactic strategies for airway colonization. As we
better understand the nature of bacterial adherence,
we may be abbe to develop vaccines with the use of
adhesins as antigens and, thereby, form local antibody
which can block bacterial binding. To the extent that
mucins serve as cellular receptors for bacteria, the
use of mucolytic or ciliokinetic agents may be helpful
in modifying the risk of infection. In addition, our
insight into bacterial adherence studies has shown
that primary colonization of the bower airway by
Pseudomonas aeruginosa is possible. Consideration of
this finding may be useful in the design of protocols
that employ prophylactic antibiotics. As Dr. Craven
will discuss, many protocols that are designed to
prevent nosocomial pneumonia involve sterilization of
the oropharynx and the gastrointestinal tract, under
the presumption that these are the only sources of
organisms that enter the lung. If Pseudomonas aeru-
ginosa can enter the tracheobronchial tree without
first colonizing elsewhere, then prevention of infection
by this organism may require the direct application of
antibiotics to the lower airway, and not just to the
oropharynx and gastrointestinal tract.
Dr. Neid.erman: Dr. Craven, can �pu discuss the scope
and causes ofthe problem ofnosocomial pneumonia?
Dr. Craven: Nosocomial pneumonia accounts for ap-
proximately 15 percent of hospital-acquired infec-
tions�#{176}and is the leading cause ofdeath from nosocom-
ial infection.5’ Rates of nosocomial pneumonia are
considerably higher in intensive care unit patients
compared to patients on hospital wards, and mechan-
ically ventilated patients have a risk ofpneumonia that
is several fold higher than nonventilated patients.3152so
Celis et al� examined 120 consecutive episodes of
nosocomial pneumonia and found intubation increased
the risk of nosocomial pneumonia approximately 7-
fold. Cross and Roup&’ found rates of pneumonia in
patients receiving mechanical ventilation via an en-
dotracheal tube were increased 10-fold compared to
patients with no respiratory therapy device.3’ In the
Study on the Efficacy ofNosocomiab Infection Control
(SENIC), only 1 percent of the patients were treated
with continuous ventilatory support, but the rate of
pneumonia was 21-fold higher than patients who were
not receiving mechanical ventilation)’�
Gram-negative bacilli, such as Escherichia coli,
Kkbsiella pneumoniae, and Pseudomonas aeruginosa
are the most common types of bacteria causing noso-
comial pneumonia. Staphylococcus aureu.s�, which can
be methicillin-sensitive or resistant, accounts for a
large portion of hospital-acquired pneumonia in
some hospitals. Nosocornial pneumonia due to Le-
gionella pneumophila has been reported in certain
geographic areas such as Pittsburgh, Burlington (Ver-
mont), and Los Angeles. These infections are usualby
associated with cooling-tower reservoirs or hospital
water which is heavily colonized with the organism.
Anaerobes have been cultured from approximately 30
percent of infected patients, but appear to be less
important than aerobic pathogens. Most nosocomiab
pneumonias are caused by more than one pathogen.
Fatality rates for patients with nosocomial pneu-
monia remain high in many series.� Nosocomial
pneumonia contributed to 60 percent of the fatal
nosocomial infections in a study of 200 consecutive
hospital deaths by Gross et al.�’ Stevens et alM reported
fatality rates of 50 percent for intensive care unit
patients with hospital-acquired pneumonia compared
to 3.5 percent of patients without pneumonia and
rates were higher for patients infected with Pseudo-
monas aeruginosa. In our study of 233 mechanicabby
ventilated patients, there was a 55 percent fatality rate
for patients with pneumonia compared to a rate of 25
percent for patients without pneumonia)’� These data
underscore the need for earlier recognition, treat-
ment, and prevention.
Dr. Niederinan: What are some of the risk factors for
nosocomial pneumonia?
Dr. Craven: Aspiration ofbacteria from the oropharynx
is the primary route of entry into the lung, and a
number of factors affect the type and number of
bacteria that colonize the oropharynx.�’� Most of us
aspirate bacteria into our lower airways daily, but
pneumonia does not devebop)’#{176} The development of
pneumonia is not only related to the numbers and
types ofbacteria that are aspirated, but also ability of
the lung’s mechanical, humorab, and cellular defenses
to contain and prevent infection.
© 1990 American College of Chest Physicians by guest on July 10, 2011chestjournal.chestpubs.orgDownloaded from
176 Pneumonia in CritiCally iii HOspitaliZed Patient (Niede,’man at a!)
Table 2-Factors That May increase the Risk of Nosocomial
Pneumonia by Aliering Colonization, Increasing the Risk ofAspiration, or impairing Host Defenses
Host factors:
Age
Obesity
Coma
Underlying disease-(chronic lung disease, congestive heart failure,
diabetes mellitus, AIDS, systemic lupus, cancer, central nervous
system diseases, seizures, head trauma, uremia, malnutritionDrugs:
Heroin, cocaine, alcohol, sedatives, antibiotics, antacids, histamine
type-2 blockers, steroids, cytotoxic drugs
Invasive Devices:
Intubation, tracheostomy, mechanical ventilation, nasogastnc tube
Surgery:Head and neck, chest, abdomen
Risk factors that may increase the risk of nosocomial
pneumonia are summarized in Table 2. Host factors
such as age and underlying disease may either allow
more bacteria to enter the lung or impair removal of
bacteria by various pulmonary host defense mecha-
nisms. Drugs or medications may either alter con-
sciousness or promote infection by impairing removal
of pathogens from the lung. Invasive devices and
surgery increase the risk of aspiration or removal of
bacteria from the lung.
As mentioned, the use ofan endotracheal tube may
increase the risk of nosocomial pneumonia at least 6-
to 21-fold compared to nonintubated patients.31’�
The presence of the endotracheal tube impairs re-
moval of bacteria, allows leakage of pathogens around
the cuff, and causes local trauma and inflammation
(Table 3). The upper part of the tracheobronchial tree
contains heavily ciliated epithebium and mucus that
can trap and clear bacteria from the lung. The cilia
beat hundreds oftimes per minute, in unison, to move
mucus and bacteria out ofthe trachea. Colonization of
the tracheobronchial tree may decrease or alter this
Table 3-Effect ofintubation on OropharyngealCOlonization and the Ibthogenesis of Pneumonia
Endotracheal Tube:
-bypasses nasopharynx and mechanical trapping of particles-impairs normal temperature and humidity of air
-acts as a foreign body and produces local trauma
-impairs cillary clearance and natural removal of secretions
-impairs swallowing-changes mouth flora-cuffleaks contaminated secretions from the oropharynx
-nasotracheal tube may cause sinusitis
-allows organisms to directly reach the lung
Nasogastric Tube:
-foreign body that impairs swallowing
-causes stagnation oforopharyngeal secretions
-increases reflux through lower esophageal sphincter
-acts as a conduit for bacterial migration
�-: may increase risk of sinusitis
activity, and increase the number of pathogens to be
cleared by alveolar macrophages, polymorpho- nuclear
leukocytes, humoral antibodies (1gM, IgG, IgA), and
complement. In addition, the endotracheal tube can
impair swallowing, and alter the host’s diet and thus
his intestinal flora. The use of a nasogastric tube is
also a widely unappreciated risk factor for pneumo-
nia,u� that may increase the risk of nosocomial pneu-
monia by several different mechanisms (Table 3).
Dr. Niederman: Dr Craven, could �pu comment on the
role ofrespiratory therapy equipment in the pathogen-
esis of pneumonia?
Di� Craven: When a patient is mechanically ventilated,
the bower respiratory tract may become colonized
more readily with bacteria. In the 1960s, the Dallas
group reported that ventilators with contaminated
mainstream nebulizers could generate bacterial aero-
sobs that infiltrated the terminal bronchioles and
alveoli, resulting in a necrotizing Gram-negative pneu-
monia.�#{176}’6’Most ventibators now heat and humidify the
inspiratory phase gas with humidifiers that do not
generate significant bacterial aerosols.
While ventilators rarely infect patients, secretions
from patients are a common source of bacterial con-
tamination of the mechanical ventilator tubing and
condensate present in the tubing.t1� If the condensate
is inadvertenfly flushed back into the patient, pneu-
monia can result, because the patient receives a large
inocubum of bacteria in a volume of fluid which can
be inoculated directly into the lungs, thereby over-
whelming host defenses. Several devices have been
advocated to remove condensate from the circuit. One
of these devices, the heat moisture exchanger, elimi-
nates the condensate problem, but may not provide
optimal humidity to the patient.6”M
In-line medication nebulizers are a potential source
ofbacterial aerosols that may penetrate to the terminal
bronchioles and avoid host defenses. These devices
may become contaminated by reflux of condensate
from the tubing. Humidifying cascades appear to be
un unbikely risk factor for pneumonia and ventilator
circuit coIonization.� Humidifying cascades should be
filled with sterile water and because of the high
temperature in the cascade, bacterial growth of the
most nosocomiab pathogens is limited.
Dr. Niederman: Please explain how gastric colonization
may serve as a risk factor for pneumonia and discuss
how we could approach intestinal bleeding prophylaxis
so as to minimize pneumonia risk.
Dr. Craven: First, bet me briefly discuss the subject
of stress ulcer prophylaxis. In contrast to the 1970s,
when stress bleeding was a serious complication for
intensive care patients, the incidence in the 1980s
appears to be less frequent, perhaps because of
improvements in mechanical ventilation, nutritional
support, and early treatment ofshock. For this reason,
© 1990 American College of Chest Physicians by guest on July 10, 2011chestjournal.chestpubs.orgDownloaded from
CHEST I 97 I 1 I JANUARY, 1990 177
it is critical to carefully assess the needs of each
specific patient and the risk-benefit ratio of any pro-
phylactic agent.
Many critically ill patients receive prophylaxis
against stress bleeding with antacids which may neu-
tralize gastric acid or H2-blockers that block gastric
acid secretion.#{176}� Antacids and/or H2-blockers such
as cimetidine and ranitidine have been used for
prophylaxis against stress bleeding in critically ill
patients with variable effectiveness. The efficacy de-
pends on the criteria used to assess stress bleeding,
the doses administered, and the patient population
studied.� Antacids have been most effective in
studies where they were administered every 2 hours
to maintain gastric pH above 3#{149}5#{149}�,67 H2-blockers have
been most effective in earlier studies ofstress bleeding
prophybaxis, and for critical care patients who have a
moderate risk of bleeding.�’� When macroscopic
bleeding is used as the criterion for efficacy and when
combining results of all studies, antacids and H2-
blockers appear to have similar efficacy and appear
more effective than placebo.�
Prophylaxis against stress bleeding can also be
achieved with sucrose octasulfate which acts by a
different mechanism. In contrast to the potential
effects of antacids and H2-blockers on gastric pH,
sucrose octasulfate (sucralfate) activity is independent
ofhydrogen diffusion or neutralization.m70 In addition,
sucralfate has littbe buffering capacity and appears to
act by adhesion to the mucosa, altering gastric mucus,
increasing PGE2 in the gastriclumen, and by absorbing
pepsin. In studies reported to date, sucralfate appears
to provide protection against stress bleeding that is
similar to antacids and H2-blockers.�’7#{176} Although
suspensions ofsucralfate are only available in Europe,
a useable preparation for critically ill patients can be
made by putting the tablet in solution, making a
suspension, and then flushing it down the nasogastric
tube.7’ It is very important that the sucralfate tablet
not be crushed before adding the saline solution or
sterile water, and that the nasogastric tube be flushed
after the sucralfate is given, to avoid clogging the tube.
Let me now briefly discuss the role of nasogastric
tubes and gastric colonization in the pathogenesis of
oropharyngeal colonization and pneumonia. Gastric
colonization may be an important prerequisite for
retrograde colonization of the oropharynx and the
development of nosocomial pneumonia. The bacteri-
cidal activity of hydrochloric acid (pH-i) and gastric
secretions was first demonstrated in 1939 by Garrod.72
Normally, the stomach maintains near-sterility by its
acid pH. Changes in the gastric flora may occur in
patients with increased age,73 malnutrition,74 achbor-
hydria,�’76 or other gastrointestinal diseases.�
Reduced gastric acid in the intubated intensive care
unit patients may result from intrinsic disease of
gastric acid production or from the use of antacids or
histamine type 2-blockers which neutralize or block
gastric acid secretion. Correlation between levels of
bacteria in the gastric juice and treatment of patients
with peptic ulcer disease with cimetidine was reported
by Ruddebb et �
du Moulin et � described gastric overgrowth with
Gram-negative bacilli in mechanically ventilated pa-
tients and related these finding to increasing the gastric
pH . These observations have been corroborated by
other investigators.57’� The level ofgastnc overgrowth
noted in critically ill ventilated patients is a concern.
Some patients with high gastric pH had colonization
with Gram-negative bacilli that reached 100 million
organisms/mb. Colonization was usually considerably
lower when the pH was less than 3.5. Gastric over-
growth with aerobic Gram-negative bacilli was most
common,m’�’�’�#{176} but high numbers of Gram-positive
bacteria and fungi may occur as
Several investigators have studied the time se-
quence of cobonization)”�’57’79’�#{176} du Moulin et alas
showed that 52 of 58 post-surgical patients with
respiratory failure had gastric and/or tracheal coboni-
zation with Gram-negative bacilli and a clear sequence
of transmission could be demonstrated in 17 of 52
patients. In 11 (65 percent of the patients), gastrjc
colonization preceded tracheal colonization. In a sim-
ilar study, using pharyngeal and gastric specimens
from 40 medical and surgical patients, Goularte et al�
showed a clear sequence ofcobonization in ten patients
in whom four (40 percent) had gastric colonization
that preceded colonization of the pharynx. Daschner
et � reported retrograde colonization of the trachea
from the stomach in 32 percent of 142 patients who
were receiving stress ulcer prophylaxis and mechani-
cal ventilation. When a nasogastric tube is in place, it
may facilitate the transfer ofbacteria from a colonized
stomach to the oropharynx and organisms may then
be aspirated into the lung.
The possible robe of gastric colonization in the
pathogenesis of pneumonia was supported in our
prospective study of risk factors for pneumonia in
mechanically ventilated patients)’�’ Overall, 21 percent
of the 233 patients receiving mechanical ventilation
developed pneumonia. Pneumonia occurred in 38
percent of 18 patients who received antacids and
cimetidine, 36 percent of 48 patients who received
cimetidine alone, and in 18 percent of 135 patients
who received antacids alone. Although the numbers
ofpatients with pneumonia in each group were small,
H2-blockers with and without antacids were inde-
pendently associated with the development of pneu-
monia (p = .01).
In a study of 153 critical care patients receiving
antacids and/or cimetidine therapy, Donowitz et al�#{176}
showed that 59 percent of gastric cultures with a pH
© 1990 American College of Chest Physicians by guest on July 10, 2011chestjournal.chestpubs.orgDownloaded from
178 Pneumonia in CritiCally III Hospitalized Patient (Niederman eta!)
of �4 were positive for Gram-negative bacilli. In
contrast, only 14 percent of gastric cultures at a pH
�4 were positive for these organisms (p <.001). As
gastric pH increased, the proportion of specimens
with Gram-negative bacilli rose, and those with normal
flora decreased.
A more recent report of nosocomial pneumonia in
mechanically ventilated patients receiving prophylaxis
against stress bleeding concluded that rates of pneu-
monia directly correlated with increasing gastric pH)’�
In patients whose gastric pH was <3.4, the rate of
pneumonia was 41 percent compared to a rate of 69
percent in patients who had a pH >5.0.
Tube feeding and aspiration is particularly common
in critically ilb patients. Enteral feeding and prepara-
tions have a pH range from 6.4 to 7.0. Pingbeton et
at” demonstrated gastric colonization in 100 percent
ofventilated patients receiving enteral feeding without
antacid or H2-blocker therapy, and 1 1 (63 percent)
subsequently developed nosocomial pneumonia.
Two recent randomized trials81,�3�� of mechanically
ventilated intensive care unit patients given stress
ulcer prophylaxis with sucralfate compared to conven-
tional agents, found lower rates of pneumonia in
patients given sucralfate. In Tryba’s study,8’ rates of
pneumonia were increased 3-fold for patients receiv-
ing antacids compared to those given sucralfate. In
the study by Driks et al,� 61 patients were randomized
to sucralfate and 69 patients to conventional therapy;
antacids (N = 39), H2-blockers (N = 17), and antacids
and H2-blockers (N = 13). Rates of pneumonia were
12 percent in the sucralfate group compared to 23
percent for patients treated with antacids and/or H2-
blockers. Of note, pneumonia occurred in only 1 of
the 17 patients who received an H2-blocker alone as
prophylaxis against stress bleeding. The low rate of
pneumonia in the H2-blocker alone group suggests
the need for additional randomized trials to assess risk
and benefit compared to sucralfate. The H2-blocker
group was comprised mostly of medical intensive care
unit patients who may be at less risk for pneumonia.
Driks et al82 also reported that qualitative and
quantitative gastric colonization with Gram-negative
bacilli was significantly bower in patients given sucral-
fate compared to patients given conventional therapy.
Laggener et al� also reported colonization was signif-
icantly lower in the patients who received sucralfate
compared to patients who were treated with raniti-
dine)’�’ Although the changes in gastric colonization
were most likely rebated to gastric pH, two recent
reports have suggested that sucralfate may have intrin-
sic bactericidal activity)’�’M
Dr. Niederman: In addition to careful attention to
gastrointestinal bleeding prophylaxis, how can pneu-
monia be prevented in patients at risk?
Dr. Craven: Another approach to the prevention of
pneumonia and other nosocomial infections has been
selective decontamination ofthe trachea, oropharynx,
and gastrointestinal tract with antobiotics. Stouten-
beek et ab”� applied polymyxin B, tobramycin, and
amphotericin B (PTA) paste four times daily in the
oropharnyx, along with the use of a solution of these
antibiotics given via the nasogastric tube to eliminate
colonization in the stomach and upper gastrointestinal
tract. Systemic cefotaxime was also administered for
variable periods to treat incubating infections. Using
this regimen, nosocomial infections were reduced
from 81 percent in 59 historic control subjects to 16
percent in 63 patients receiving the prophylactic
antibiotic regimen, and “respiratory tract infection”
was reduced from 59 percent in the control group
compared to 8 percent for patients receiving prophy-
laxis,
Unertb et al� administered a solution of polymyxin
B, gentamicin, and amphotericin B to the nose,
oropharynx, and stomach of 19 intubated patients who
were expected to receive more than six days of
mechanical ventilation and compared infection rates
to 20 control patients. Colonization of the oropharynxand trachea were significantly bower (p <.001) in the
group given antibiotic prophylaxis. Nine cases of
pneumonia were identified in the control group com-
pared to one case in the group given antibiotic
prophylaxis.
In a more recent study ofintensive care unit patients
by Ledingham et al,� selective decontamination with
a regimen similar to that of Stoutenbeek et al reduced
the number of respiratory infections 6-fold compared
to historic control subjects (18 vs 3). There was also a
significant reduction in the colonization rates with
aerobic Gram-negative bacilli in the oropharynx and
rectum of patients given prophylaxis compared to
historical controls.
Flaherty et al� randomized patients in a cardiac
intensive care unit to receive sucralfate compared
with selective decontamination of the oropharynx and
stomach with gentamicin, nystatin, and polymyxin; no
systemic third generation cephalosporin was included.
Overall, rates of infection were 27 percent in the
sucralfate group vs 12 percent in the selective decon-
tamination group. Overall, there were five eposides of
pneumonia in the sucralfate group compared to one
in the antibiotic prophylaxis group. Patients in the
sucralfate group had three Gram-negative pneumonias
compared to none in the antiobiotic group. Although
there were no significant differences in ICU stay or
fatality rates, antibiotic use was increased 3-fold in the
sucralfate group. Antibiotic resistance was not ob-
served, but the duration ofthe study was short.
With the exception of the study of Flaherty et al,m
most of the data on selective decontamination of the
oropharynx and gastrointestinal tract has been tried
© 1990 American College of Chest Physicians by guest on July 10, 2011chestjournal.chestpubs.orgDownloaded from
CHEST I 97 I 1 I JANUARY, 1990 179
in multiple trauma patients with results compared to
historic controls. Of note is the absence ofany change
in fatality rates despite the marked decrease in noso-
comial infections. In addition, it is not clear if selective
decontamination will be applicable to chronically ill
patients who may develop colonization with more
resistant nosocomial pathogens. Although the selection
of antibiotic-resistant organisms was not encountered
in these studies, more data collected over a longer
time period are needed to definitively assess the risks,
benefits, and cost effectiveness of antibiotic prophy-
laxis in the intensive care unit patient.
ACKNOWLEDGMENT: The authors thank Ms. Ellen Maye forsecretarial assistance in preparing this manuscript.
REFERENCES
1 Dubaybo BA, Carlson RW. Post infections ARDS: mechanisms
oflung injury repair. Crit Care Clin 1988; 4:229-43
2 Maxfield BA, Sorkin B, Fazzini EP, Bapoport DM, StensonWM, Coldring RM. Respiratory failure in patients with acquired
immune deficiency syndmme and Pneumocystis carinii pneu-monia. Crit Care Med 1986; 14:443-49
3 Fein A, Wiener-Kronish JP, Niederman MS. Matthay MA.Pathophysiology of adult respiratory distress syndrome. Whathave we learned from human studies? Crit Care Clin 1986;
2:429-534 Fein A, Lippmann M, Holtzman H, Eliraz A, Goldberg SK.
The risk factors, evidence and prognosis of ARDS following
septicemia. Chest 1983; 83:40-25 Pepe PE, Potkin lIT, Reus DH, Hudson LD, Carrico CJ. Clinical
predictors ofthe adult respiratory distress syndrome. Am J Surg
1982; 144:124-30
6 Sherry B, Cerami A. Cachectin/tumor necrosis factor exertsendocrine, paracrine, and autocrine control of inflammatory
responses. J Cell Biol 1988; 107:1269-777 Fowler AA, Hamman RF, Good j’r, Benson KN, Baird M,
Eberle DJ, et al. Adult respiratory distress syndrome: risk withcommon predispositions. Ann Intern Med 1983; 98:593-97
8 Pepe PE, Thomas RG, Stager MA, Hudson LD, Carrico CJ.
Early predictions of the adult respiratory distress syndrome by
a simple scoring method. Ann Emerg Med 1983; 12:749-55
9 Grant M. Belenko M, Niederman M, JanoffA, Rosen H, FeinA. Hemodynamic status but not PMN activation and turnover
predict ARDS in sepsis [abstract]. Chest 1987; 92:59510 Duchateu M, Haas H, Schreyen H, Radoux L, Spranges I, Noel
F, et al. Complement activation in patient at risk for developingthe adult respiratory distress syndrome. Am Rev Respir INs1984; 130:1058-64
11 Weinberg P, Matthay M, Webster R, Boskos K, Goldstein I,Murray J. Biologically active products ofcomplement and acutelung injury in patients with septic syndrome. Am Rev Besp Dis1984; 130:791-96
12 Solomkin J, Cotta L, Satoh P. Hurst J, Nelson R. Complement
activation and clearance in acute illness and injury: evidence
for C3 as acell directed mediatorofthe adult respiratory distress
syndrome in man. Surgery 1985; 97:668-78
13 Langlois P, Gawryl M. Accentuated formation of the terminal
C5b-9 complement complex in patient plasma precedes devel-opment of the adult respiratory distress syndrome. Am Rev
Bespir INs 1988; 138:368-75
14 Bell RC, Coalson JJ, Smith JD, Johanson WG Jr. Multi-organ
system failure and infection in adult respiratory disease syn-drome. Ann Intern Med 1983; 99:293-98
15 Andrews CP, Coalson JJ, Johansen WG. Diagnosis of nosocomialbacterial pneumonia in acute diffuse lung injury. Chest 1981;
80254-5816 Montgomery AB, Stager MA, Carrico C, Hudson L. Causes of
Mortality in patients with the adult respiratory distress syn-
drome. Am Rev Respir Dis 1985; 132:485-89
17 Niederman M, Fein A. The interaction of infection and the
adult respiratory distress syndrome. Crit Care Cliii 1986; 2:471-95
18 Lee CF, Fein A, Lippmann M, Holtzmann H, Kimbel P.
Weinbaum G. Elastolytic activity in pulmonary lavage fluid
from patient with adult respiratory distress syndrome. N Engi JMed 1981; 304:192-96
19 Fick RB Jr, Baltimore RS, Squier SU, Reynolds HY. IgG
proteolytic activity ofPseudomonas aeruglnosa in cystic fibrosis.
J Infectious Dis 1985; 151:589-98
20 Niederman MS. Merrill WW, Fblomski L, Reynolds HY, Gee
JBL. Influence of sputum IgA and elastase on tracheal cellbacterial adherence. Am Rev Respir Dis 1986; 133:255-60
21 LaForce FM, Mullane JF, Boehme RF, Kelly WJ, Huber GL.The effect of pulmonary edema on antibacterial defenses of the
lung. J Lab Clin Med 1973; 82:634-4822 Juers J, Rogers RM, McCurdy JB. Enhancement of bactericidal
of capacity of alveolar macrophages by human “alveolar lining
material.” Clin Res 1975; 23:348
23 Matuschak GM, Rinaldo JE. Organ interactions in the adult
respiratory distress syndrome during sepsis. Role ofthe liver in
host defense. Chest 1988; 94:404
24 Huber GL, LaForce FM, Mason R. Impairment and recovery
of pulmonary antibacterial defense mechanisms after oxygen
administration. J Clin Invest 1970; 49:47a25 Matthay BA, Wiedemann HP, Niedernian MS. Nosocomial
pulmonary infections in acute respiratory failure. Resp Care1983; 28:634-40
26 Johanson WG, Pierce AK, Sanford JR Changing pharyngeal
flora of hospitalized patients: Emergence of gram-negative
bacilli. N Engl J Med 1969; 281:1137-4027 Johanson WG, Pierce AK, Sanford JP, Thomas GD. Nosocomial
respiratory infections with gram-negative bacilli: The signifi-cance of colonization of the respiratory tract. Ann Intern Med
1972; 77:701-06
28 T,llotson JR. Finland M. Bacterial colonization and clinic’ilsuper infection of the respiratory tract complicating antibiotictreatment ofpneumonia. J Infectious INs 1969; 119:597-624
29 Laurenzi GA, Potter RT, Kass EH. Bacteriologic flora of thelower respiratory tract. N Engl J Med 1961; 265:1273-78
30 Lees AW, McNaught W. Bacteriology of the lower respiratory
tract secretions, sputum, and upper respiratory tract secretions
in “normals” and chronic bronchitics. Lancet 1959; #{252}:1112-15
31 Cross AS, Roup B. Role of respiratory assistance devices inendemic nosocomial pneumonia. Am J Med 1981; 70:681-85
32 Niederman MS, Merrill WV� Ferranti RD. Pagano KM. Palmer
LB, Reynolds HY. Nutritional status and bacterial binding inthelower respiratory tract in patients withchronic tracheostomy.
Ann Intern Med 1984; 100:795-80033 Ramphal R, Small PM, Shands JW Jr. Fischlschweiger W, Small
PA Jr. Adherence of Pseudomonas aeruginosa to tracheal cells
injured by influenza infection or endo-tracheal intubation. Infec
Immun 1980; 27:614-1934 Niederman MS. Mantovani R, Schoch P. Pappas J, Fein AM.
Patterns and routes of tracheobronchial colonization in mechan-ically ventilated patients: The role of nutritional status in
colonization ofthe lower airway by Pseudomonas species. Chest1989; 95:155-61
35 Beachey EH. Bacterial adherence: Adhesin-receptor interac-tions mediating the attachment of bacteria to mucosal surfaces.
J Infect Dis 1981, 143:325-45
© 1990 American College of Chest Physicians by guest on July 10, 2011chestjournal.chestpubs.orgDownloaded from
180 Pneumonia in Critically Iii Hospitakzed Patient (Nlede,man eta!)
36 JohansonWG, HiguchiJH, ChaudhuriTR, Woods DE. Bacterial
adherence to epithelial cells in bacillary colonization of the
respiratory tract. Am Rev Respir Dis 1980; 121:55-63
37 HiguchiJH, Johanson WG. The relationship between adherence
of Pseudomonas aeruginosa to upper respiratory cells in vitro
and susceptibility to colonization in vivo. J Lab Clin Med 1980;
95:698-705
38 Woods DE, Straus DC, Johanson WG, Bass JA. Role of salivary
protease activity in adherence of gram-negative bacilli to mam-
malian buccal epithelial cells in vivo. J Clin Invest 1981; 68:1435-40
39 Palmer LB. Niederman MS. Matthay BA. Nosocomial pneu-
monia: A threat to your patient? J Respir Dis 1984; 5:80-9840 Palmer LB, Merrill WW, Niederman MS. Ferranti RD. Rey-
nolds HY. Bacterial adherence to respiratory tract cells: Rela-
tionships between in vivo and in vitro pH and bacterial
attachment. Am Rev Respir Dis 1986; 133:754-58
41 Niederman MS. Rafferty TD, Sasaki Cr, Merrill WW, Matthay
BA, Reynolds HY. Comparison ofbacterial adherence to ciliated
and squamous epithelial cells obtained from the human respi-
ratory tract. Am Rev Respir Dis 1983; 127:85-90
42 Niederman MS. Ferranti RD. Ziegler A, Merrill WW Reynolds
HY. Respiratory infection complicating long-term tracheostomy:
The implication of persistent gram-negative tracheobronchial
colonization. Chest 1984; 85:39-44
43 Schwartz SN, DowlingJN, Benkovic C, De Quittner-Buchanan
M, ProstkoT, Yee RB: Sources ofgram-negative bacilli colonizing
the tracheae of intubated patients. J Infect Dis 1978; 138:227-
31
44 Pingleton SK, Hinthorn DR. Liu C. Enteral nutrition in patients
receiving mechanical ventilation: Multiple sources of tracheal
colonization include the stomach. Am J Med 1986; 80:827-32
45 Niederman MS. Bacterial adherence as a mechanism of airway
colonization. Europ J Clin Microbiol Infect Dis 1989; 8:15-20
46 Franklin AL, Todd T, Gurman G, Black D, Mankinen-Irvin PM,
Irvin lit Adherence of Pseudomonas aeruginosa to cilia of
human tracheal epithelial cells. Infec Immun 1987; 55:1523-25
47 Woods DE, Straus DC, Johanson WG, Berry VK, Bass JA. Role
of pili in adherence of Pseudomonas aeruginosa to mammalian
buceal epithelial cells. Infec Immun 1980; 29:1146-51
48 Doig P, Smith NR, Todd T, Irvin RT Characterization of the
binding ofPseudomonas aensginosa alginate to human epithelial
cells. Infect Immun 1987; 55:1517-22
49 Grant MM, Niederman MS. Poehlman MA, Fein AM. Mucins
in human sputum affect the binding ofPseudomonas aeruginosa
to cultured tracheal epithelial cells. Am Rev Respir Dis 1989;
139:A3350 Horan TC, White JW, Jarvis WR. Nosocomial infection surveil-
lance, 1984. Morbid Mortal Weekly Rep 1986; 35:17SS-29SS
51 Gross PA, Neu HC, Aswapokee P, Van Antwerpen C, Aswapokee
N. Deaths from nosocomial infection: Experience in a university
hospital and a community hospital. Am J Med 1980; 68:219-223
52 Haley RW Hooton TM, Culver DH, Stanley RC, Emori TG,
Hardison CD, et al. Nosocomial infections in U.S. hospitals,
1975-1976: Estimated frequency by selected characteristics of
patients. Am J Med 1981; 70:947-59
53 Cells R, Tories A, Gatell JM, Almela M, Rodriguez-Roisin R,
Augusti-Vidal, A. Nosocomial pneumonia: a multi-variate anal-
ysis ofrisk and prognosis. Chest 1988; 93:318-24
54 Stevens RM, Teres D, Skillman JJ, Feingold DS. Pneumonia in
an intensive care unit: A thirty-month experience. Arch Intern
Med 1974; 134:106-111
55 du Moulin GC, Paterson DG, Hedley-Whyte J, Lisbon A.
Aspiration of gastric bacteria in antacid-treated patients: A
frequent cause of post-operative colonization of the airway.
Lancet 1982; 1:242-45
56 Craven DE, Kunches LM, Kilinsky V, Lichtenberg DA, Make
BJ, McCabe WR. Risk factors for pneumonia and fatality in
patients receiving continuous mechanical ventilation. Am Rev
Resp Dis 1986; 133:792-9657 Daschner F, Kappstein I, Engles I, Reuschenbach K, Pfisterer
J, Krieg N, et al. Stress ulcer prophylaxis and ventilation
pneumonia Prevention by antibacterial cytoprotective agents.
Infect Control Hosp Epidemiol 1988; 9:59-65
58 Garibaldi HA, Britt MR. Coleman ML, Reading JC, Pace NL.
Risk factors for post-operative pneumonia. Am J Med 1981;
70:677-680
59 Huxley EJ, Viroslav J, Gray WR, Pierce AK. Pharyngeal
aspiration in normal adults and patients with depressed con-
sciousness. Am J Med 1978; 64:564-68
60 Pierce AK, Sanford JP, Thomas GD, Leonard JS. Long-term
evaluation ofdecontamination of inhalation-therapy equipment
and the occurrence of necrotizing pneumonia. N Engl J Med
1970; 282:528-31
61 Reinarz JA, Pierce AK, Mays BB, Sanford JP The potential role
of inhalation therapy equipment in nosocomial pulmonary
infection. J Clin Invest 1965; 44:831-3962 Craven DE, Goularte TA, Make BJ. Contaminated condensate
in mechanical ventilator circuits: A risk factor for nosocomial
pneumonia? Am Bev Respir Dis 1984; 129:625-28
63 Maclntyre NR, Anderson HR. Silver RM, Schuler FR, Coleman
RE. Pulmonary function in mechanically-ventilated patients
using 24-ho�sr use of a hydroscopic condensor humidifier. Chest
1983; M:5�0-64
64 Make BJ, Craven DE, O’Donnell C, Manning M, Gilmartin
ME, Miller G. Clinical and bacteriologic comparison of hydro-
scopic and cascade humidifier in ventilated patient [abstract].
Am Rev Resp Dis 1987; 135:A12.
65 Goularte TA, Craven DE. Bacterial colonization of cascade
humidifier reservoirs after 24 and 48 hours of continuous
mechanical ventilation. Infec Control 1987; 8:200-03
66 Schuman RB, Schuster DT, Zuckerman GR. Prophylactic treat-
ment for stress ulcer bleeding: A reapprisal. Ann Intern Med
1987; 106:562-6767 Hots J. Antacids for the prevention of gastrointestinal stress
bleeding. In Tryba M, ed: Prevention of stress bleeding in
critically ill patients-a new concept. New York: Thieme, 1988:
19-21
68 Huchzermeyer H, Tryba M. Efficacy ofH2-receptor antagonists
for the prevention ofstress bleeding. In Tryba M, ed: Prevention
of stress bleeding in critically ill patients-a new concept. New
York: Thieme, 1988: 22-7
69 Tryba M. Studies on the prevention of stress bleeding with
sucralfate. In: (Tryba M, ed). Prevention of stress bleeding in
critically ill patients-a new concept. New York: Thieme, 1988:
1-162
70 Cannon LA, Heiselman D, Gardner W, Jones J. Prophylaxis of
upper gastrointestinal tract bleeding in mechanically ventilated
patients: A randomized study comparing the efficacy of sucral-
fate, cimetidine, and antacids. Arch Intern Med 1987; 147:2101-
06
71 Schneider JS, Ouellette MS. Sucralfate administration via
nasogastric tube [letter]. N EngJ Med 1984; 310-990
72 Garrod LP. A study of the bacterial power of hydrochloric acid
and ofgastricjuice. St Barth Hosp Rep 1939; 72:145-67
73 Arnold I. The bacterial flora within the stomach and small
intestine-the effect of experimental alterations of acid-base
balance and the age ofthe subject. Am J Med Sci 1933; 186:471-
81
74 Gracey M, Suharjano 5, Stone DE. Microbial contamination of
the gut: Another feature of malnutrition. Am J Clin Nutr 1973;
26:1170-74
75 Draser BS, Shiner M, McLeod GM. Studies on the intestinal
flora. The bacterial flora of the gastrointestinal tract in healthy
© 1990 American College of Chest Physicians by guest on July 10, 2011chestjournal.chestpubs.orgDownloaded from
Plan to AMend ACCP’s
56th Annual Scientific Assembly
Toronto, Ontario, Canada
October 22-26, 1990
CHEST I 97 I 1 I JANUARY, 1990 181
and achlorhydric persons. Gastroenteroll9fi9; 56:71-9
76 Gray JD, Shiner M. Influence of gastric pH on gastric and
jejeunal flora. Gut 1967; 8:574-81
77 Atherton ST. White DJ. Stomach as source ofbacteria colonizing
respiratory tract duringartificial ventilation. Lancet 1978; ii:968-
69
78 Ruddell WSJ, Axon ATh, Findlay JM, Bartholomew BA, Hill,
MJ. Effect of cimetidine on the gastric bacterial flora. Lancet
1980; i:672-74
79 Goularte TA, Lichtenberg DA, Craven DE. Gastric colonization
in patients receiving antacids and mechanical ventilation: A
mechanism for pharyngeal colonization. Am J Infect Control
1986; 14:88
80 Donowitz GL, Page MC, Mileur BL, Guenthner SH. Alteration
of normal gastric flora in critical care patients receiving antacid
and cimetidine therapy. Infect Control 1986, 7:23-681 Tryba M: Risk of acute stress bleeding and nosocomial pneu-
monia in ventilated intensive care unit patients: Sucralfate
versus antacids. Am J Med 1987; 83:suppl 3B:117-24
82 Driks MR, Craven DE, Celli BR, Manning M, Burke RA,
Garvin GM, et al. Nosocomial pneumonia in intubated patients
given sucralfate as compared with antacids or histamine type 2
blockers. N Eng J Med 1987; 317:1376-82
83 Laggner AN, Lenz K, Druml W, Schneeweib B, Grimm G.
Stressulcusprophylaxe auf einer internen intensivstation. Wien
KIm Wochenschr 1987; 99:12-3
84 Mantey-Stiers F, Tryba M. Antibacterial activity ofsucralfate in
human gastric juice. Am J Med 1987; 87(3B):125-7
85 Stoutenbeek CE van Saene HKF, Miranda DR. Zandstra DF.
The effect of selective decontamination of the digestive tract on
colonization and infection rate in multiple trauma patients.
Intensive Care Med 1984; 10:185-92
86 Unertl K, Ruckdeschel G, Selbmann HK, Jensen U, Forst H,
Lenhart FP, et al. Prevention of colonization and respiratory
infections in long-term ventilated patients by local antimicrobial
prophylaxis. Intensive Care Med 1987; 13:106-13
87 Ledingham IM, Alcock SR. Eastaway AT, McDonald JC, McKay
IC, Bamsay G. Triple regimen of selective decontamination of
the digestive tract, systemic cefo-taxime, and microbiological
surveillance for prevention of acquired infection in intensive
care. Lancet 1988; i:785-9088 FlahertyJ, Nathan C, Kabins SA, Weinstein BA. Nonabsorbable
antibiotics versus sucralfate in preventing colonization andinfection in a cardiac surgery intensive care unit [abstract]. 28th
Interscience Conference on Antimicrobial Agents and Chemo-
therapy, Los Angeles, 1988
© 1990 American College of Chest Physicians by guest on July 10, 2011chestjournal.chestpubs.orgDownloaded from
DOI 10.1378/chest.97.1.170 1990;97; 170-181Chest
M S Niederman, D E Craven, A M Fein and D E SchultzPneumonia in the critically ill hospitalized patient.
July 10, 2011This information is current as of
http://chestjournal.chestpubs.org/content/97/1/170.citationUpdated Information and services can be found at:
Updated Information & Services
http://chestjournal.chestpubs.org/content/97/1/170.citation#related-urlsThis article has been cited by 6 HighWire-hosted articles:
Cited Bys
http://www.chestpubs.org/site/misc/reprints.xhtmlonline at: Information about reproducing this article in parts (figures, tables) or in its entirety can be foundPermissions & Licensing
http://www.chestpubs.org/site/misc/reprints.xhtmlInformation about ordering reprints can be found online:
Reprints
the right of the online article.Receive free e-mail alerts when new articles cite this article. To sign up, select the "Services" link to
Citation Alerts
slide format. See any online figure for directions. articles can be downloaded for teaching purposes in PowerPointCHESTFigures that appear in Images in PowerPoint format
© 1990 American College of Chest Physicians by guest on July 10, 2011chestjournal.chestpubs.orgDownloaded from