5/1/2015

1

Quite Alarming! Implementation of Alarm

Management Strategies to

Reduce the Incidence

of False AlarmsCLASS A75M451

UnityPoint Health Methodist/Proctor, Peoria,ILHOI AACN President cherrmann@frontier.com

www.Cherylherrmann.com

Objectives

� Describe the impact of alarm fatigue on patient outcomes

� Implement evidenced based practice alarm management strategies to improve patient safety

It Takes a Team…

• Sasha Carr, RN, BSN, MSN, PCCN• Graduate student

• Seth Weber, RN, BSN• BSN student

• Cardiac Progressive Care Unit PI team• Cardiac Service Line PI team• Biomedical Engineers• Quality and Analytics Department

Team Work

with RN(s) next to you

� Quickly introduce yourself, your role, and where from.

� Next, together count how many alarms you can think of in the hospital.

� Who came up with the most??

Alarm Definitions

� An anxious awareness of danger

� A sudden fear or distressing suspense caused by anawareness of danger; apprehension; fright

� Any sound, outcry, or information intended to warnof approaching danger

� An automatic device that serves to call attention, torouse from sleep or to warn

� A warning sound; signal for attention

� A warning of existing or approaching danger

Alarm Definitions Themes

� An anxious awareness of danger

� A sudden fear or distressing suspense caused by anawareness of danger; apprehension; fright.

� Any sound, outcry, or information intended to warnof approaching danger

� An automatic device that serves to call attention, torouse from sleep or to warn

� A warning sound; signal for attention

� A warning of existing or approaching danger

5/1/2015

2

”

“How many times do the

alarms on your list

cause an anxious

awareness of danger?

Medical Alarm Statistics

� In 1983, there were an average of 6 different medical alarm types

� As of 2011, the average number increased to 40

� Estimated 80-99% of alarms heard today do not require clinical

intervention

� Alarm conditions set too tight

� Default settings are not adjusted for the individual patient

� EKG electrodes are dry or mispositioned

Medical Alarm Statistics

� From January 2009 to June 2012, the Joint Commission logged 98 sentinel alarm

events

� 80 deaths

� 13 permanent loss of function

� 5 unexpected additional care

or extended stay

What is alarm fatigue?

� Alarm fatigue is sensory overload caused by exposure to an excessive number of alarms

� This leads to the nurse or monitor tech becoming desensitized to alarms causing

� delayed response time

� Ignored alarm assuming it’s false

Joint Commission published Sentinel

Event Alert #50 in April 2013

5/1/2015

3

Joint Commission 2014 NPSG

� Manage clinical alarm systems that have the most direct relationship to patient safety

� Phased –in between 2014 and 2016

� 2015: Reduce the harm associated with clinical alarm systems

� By 2016 all organizations to have clear-cut guidelines for managing alarms

� Clinically appropriate settings for alarms

� When alarm can be disabled

� When parameters can be changed

� Who has the authority to set, change or turn off alarm parameters

� Monitoring and responding to alarms

� Checking individual for accurate settings, proper operation, and

detectability

Purpose of Project

The purpose of the project was to improve patient safety and

reduce alarm fatigue by decreasing the incidence of false

alarms for a central telemetry monitor station.

5/1/2015

4

Current Situation

CARDIAC UNITS

UnityPoint Health - Methodist

Study

� Data collected June 2013

� Total of 79 patients (5C = 30, non-cardiac = 27 and CVICU = 22)

� Alarm data was collected for 24 hours on each patient.

� Alarm data includes

� Total number of alarms by unit

� True and false alarms

� Causes of alarms

True versus False Alarms

� A false alarm is one that an alarm sounds in response to inappropriate stimuli.

� A true alarm is one that an alarm sounds in response to an event that needs an action.

Excessive alarms,

especially false ones, can desensitize

medical personnel causing delay or no

response!

Specific Details for

Implementation Purposes

� We manually collected alarm data by going into the alarm events that were recorded for each patient during the last 24 hours.

� We clicked on each alarm event which enabled us to view the EKG strip of the event.

� If it was a true arrhythmia it counted as true alarm. For

example: VT event and EKG showed a run of VT – this is

a true alarm

� If it alarmed VT and the EKG strip showed artifact, it was

counted as a false alarm.

Sample of Data Collection Tool

� We used tally marks when collecting the data then typed them into an excel file to analyze the data.

� Collected alarm events for 24 hours per patient

Types of false alarm

Room True Alarm False Alarm

% of false alarms VT/VF Brady Tachy couplets asystole ST pauses

C501 13 16 55 0 4 3 9

C504 0 8 100 4 4

C505 0 5 100 2 2 1

C507 35 1 3 1

C509 0 6 100 4 2

C512 0 3 100 2 10 1

C518 28 72 72 23 47

C520 0 4 100 4

Other Tips:

� Our data does not include blood pressure, Sp02, leads off alarms or any advisory alarms

� These alarms do not record on the monitor. To get these alarms, someone would need to sit at the monitors and count them as they hear them – very

time consuming!

5/1/2015

5

Percentage of

False Alarms by Unit

5C had and extremely high percentage of false alarms at 67%

The non-cardiac units had just over half of their alarms being false

CVICU was the only unit to have less than 50% of their alarms at 30%

The goal is to lower each units percentage of false alarms

True vs. False

Alarms per Patient

5C again has the highest

number of false alarms per

patient at 11 compared to only 5.4 true alarms

Non-Cardiac patients had

roughly the same amount of true and false alarms per day

The CVICU was again the only unit to have more true alarms at

8 per patient compared to only 5 false alarms

Note– CVICU had alarm

Total Number of

Alarms by Unit

5C had the highest

total number of alarms with 490 in two

days

� Fall 2010 – Winter

2011 CVICU had done some alarm management

based on this article.

Demographics of Project

Cardiac Progressive-Telemetry Unit

� Data was collected from 57 patients from both cardiac and non-cardiac patient floors

� Central Telemetry Monitoring Station

� Station is able to watch up to 62 patients at any given time

� Time frame of 48 hours for data collection; 24 hours of data per patient

Oh! So Many Alarms!

Yes!

787 Alarms!

In just 48 hours!

Pre Data

Excludes BP, SpO2, leads off, and advisory alarms

5/1/2015

6

Central Monitor station

One alarm at least every 2.5

minutes

� Excludes BP, SpO2, leads off, and advisory alarms

ST False Alarms for

5C

Missing ST alarm data was collected on July 24, 2013

Pre data was collected on 31 patients over a 24 hour period

There were a total of 158 ST false alarms in that time

How Can We Decrease the

Incidence of False Alarms?

Over 50% of Alarms in both groups were false!

What Were

the False

Alarms?

Biggest Culprits of False Alarms were couplets

and ST Segment.

All ST Segment alarms were false!

100

1330

282

20

178

24

0

50

100

150

200

250

300

Nu

mb

er o

f Fa

lse

Ala

rms

Pre data Causes of False Alarms

VT/VF Brady Tachy Couplets Asystole ST Pauses Other

”

“ How do We

Decrease the

Number of False

Alarms?AACN There are EBP guides for alarm

management that we will implement

5/1/2015

7

Interventions

� Adjust Alarms (this was implemented first)

� Changed couplet and ST segment alarms to message

� Adjusted alarms to specific patients

E-Learning Education Module

Change Alarms to Messages

� Studies have shown that changing alarm default settings and customizing alarms can decrease the number of false alarms

� Customize alarms to meet the needs of individual patients

� Change alarms within 1 hour of patient admission

� Changing alarms to message may decrease the number of false alarms by displaying a message rather than sounding a tone

Parameter Adjustment Guideline

Alarm Hi & Low Limits

� We do not want to turn off alarms, however inappropriate alarms may desensitize us and lead to delayed response or poor outcomes

� Alarms that we are not treating should be adjusted.

� For example

� CVICU had 36 true alarms in two hours for Atrial Fib rate > 150. Alarm rate should have been adjusted to 170 bpm until the HR decreased as a response to the medicine.

� HR drops to occasionally to frequently drops to 48 from 55. MD order to call for HR < 45. Adjust the rate to 45

� CVICU and MT should review and adjust the alarm limits at the beginning of the shift and prn.

Parameter Adjustment Guideline ST analysis

� CVICU

� Individualize the upper limits per the ST Segment blue print

� For example if the ST elevation is 2.5, increase the limit to 3

� When walking patient with telemetry turn ST to message so it does not alarm constantly during the walk.

� Relearn the ST blueprint

� 100% pacing, ST analysis is inappropriate. Turn off

� 5C

� Current process, ST limits cannot be adjusted (default to +2 to – 2)

� Biomed changed the default alarm to message rather than advisory. Thus instead of one beep, no noise will occur – just a message across the screen

� Patients who are a rule out MI, you may want to consider turning it to advisory. – this can be done in the room.

Parameter Adjustment Guideline

Couplets

� CVICU

� No change. Couplet alarm default to message

� 5C

� Biomed changed the default alarm to message rather than advisory. Thus instead of one beep, no

noise will occur – just a message across the screen

� The couplet alarm frequently alarmed due to artifact

� VT > 2 is still warning so we will capture if there is a problem

Post Data Alarm Changes –

WOW!

� Changing Couplets

and ST to message,

we decreased 423

alarms at 5 C

central station in 48

hours

5/1/2015

8

However….

� False VT/VF alarms are now the #1 cause of false alarms

NEXT STEP TO

DECREASE MORE

FALSE

ALARMS…..

AACN There are EBP guides for alarm

management that we will implement

ARE WE

FOLLOWING THE

GUIDELINES?

MONKEY SURVEY

CVICU & 5C RNS

� NOTE – Survey was done at the beginning of the project before education.

5/1/2015

9

WE HAVE

OPPORTUNITIES

FOR

IMPROVEMENT!

Interventions

� Adjust Alarms (this was implemented first)

� Changed couplet and ST segment alarms to message

� Adjusted alarms to specific patients

� Skin Prep for lead placement

� This was only done 4% of the time prior to change

E-Learning Education and House-wide Education

Skin Preparation can help!

� Proper skin preparation before ECG electrodes

are placed decreases skin impedance and signal noise, thereby enhancing conductivity.

� Skin prep techniques

� Wash skin with soap and water

� Remove excess hair

� Roughen skin with abrasive washcloth or

sandpaper

Skin Prep GuidelineSkin Prep Step 1

Many of our false alarms were due to artifact that the monitors picked up as VT/VF or couplets

Process Change

� For direct admits, with admission packet/supplies take two

wash cloths in the room

� Wash the electrode area with soap and water

� Use dry wash cloth to dry the electrode area

� Apply electrode

� Do not use bath wipes due to the oils

� Avoid using alcohol as it can dry out the skin. May use alcohol

for oily skin

� For ED patients – if you are getting a good reading without

artifact, no need to change electrodes

5/1/2015

10

Continued Artifact/False

Alarms Skin Prep Step 2

� When the MT notices continued artifact causing false alarms, the RN or CNA will be notified to do a skin prep and change electrodes.

� Date the new electrodes with a marker.

Continued Artifact/False

AlarmsSkin Prep Step 3

� Oily Skin: use alcohol

� Flaky skin: use sandpaper that is located at the monitor station

� Tear 1 inch piece of sand paper from dispenser

� Use brushing action to gently abrade each electrode area

� Apply electrode to prepped area

Electrodes

� Even though the guideline states to change electrodes daily, we have chosen not to implement this at the present time.

� We want to see if skin prep is all we need and also we are concerned with skin tears.

� However, change electrodes for increased artifact

Misc

� To decrease noise, the EKG monitor alarm volume

in the 5C rooms has been defaulted to 20% rather

than 40%.

� When patient leaves 5C (including cath lab):

� Notify MT patient is leaving the floor

� Pause the monitor in the room so it does not alarm

� When patient returns, call MT so a post procedure

strip can be run.

Summary of changes

�Skin prep prior to electrode placement

�Change couplet alarm to message

�Change ST alarm to message

�Adjust inappropriate alarm limits

Outcomes

� 159 decrease in false alarms by with 1st

intervention of changing ST and couplet alarms to message

� An additional 124 decrease in false alarms with the 2nd intervention of doing skin prep before lead placement

5/1/2015

11

Outcomes Continued

We decreased the biggest culprits to “0” after both interventions!

Outcomes Continued

• False alarms decreased by an average of 5 per patient in 24 hours at a

central monitor station that can monitor up to 62 patients at a time.

• Thus the potential to have 310 less false alarms in 24 hours!

11

8

6

3

4

5

6

7

8

9

10

11

12

Pre-data (6/13) Post Monitor

Change (8/13)

Post Skin Prep

(12/13)

Average Number of False Alarms per Patient

Avg False Alarms…

Conclusion

� Implementation of two aspects of the AACN Practice Alert on

Alarm Management reduced false alarms by 283 alarms and the average number of false alarms per patient by 5 alarms in

24 hours.

� Data should be collected again to determine next steps to

continue to decrease false alarms to prevent alarm fatigue

and potential harm to patients.

NTI 2014

3M and GE exhibitors viewed

poster

5/1/2015

12

More

Data

Seven Monitor Stations

� CVICU

� ICU

� ICU – Proctor

� 5C – Progressive

� 2P – telemetry

� ED

� ED - Proctor

Total Monitor Alarms both campuses

for one week = 147,447

� 2P – Proctor = 4,506

� 5C – Methodist = 16,057

� CVICU – Methodist = 38,376

� ICU – Methodist = 23,799

� ED – Methodist = 33,507

� ED – Proctor = 9,883

� ICU – Proctor = 21,319

NEXT STEP TO

DECREASE MORE

FALSE

ALARMS…..

5/1/2015

13

o

Recommended

changes – given

approvalParameter Alarms

SPO2 LO ↓ 88 or 89

ST HI

↑ or ↓ to 2.4 or 2.5 from 2.0 or message

Do you know any patients who have died because a CVP or PAD alarm didn’t sound?

SpO2 decreased to 88 - 89% Hospital protocol is to keep > 90%

If 90% okay, why is set to alarm?

ST: ↑ or ↓ to 2.4 or 2.5 from 2.0 or message

Arrhythmia Alarms

If not treating it,

turn to message or off.

Note – we had inconsistent alarms settings between the units

Technical Alarms

Possibly change to disposable SpO2 probe

Use snap electrodes for patient’s who pull

off electrodes

Check the newness of

the electrodes

5/1/2015

14

”

“Implementation

April and May

Post Data

Other interventions to consider…

� AJCC January 2015

Joint Commission and AACN

Practice Alert

� Alarms that are not be treated should be adjusted to prevent

alarm fatigue

Alarm Management Tools

� www.aacn.org

� AACN Practice Alert http://www.aacn.org/dm/practice/actionpaklist.aspx?menu=practice&lastmenu

� AACN Clinical Tool on Strategies of Alarm Fatigue (2013 AACN NTI ActionPak)

http://www.aacn.org/wd/practice/content/practicealerts/alarm-management-practice-alert.pcms?menu=practice

Quite Alarming! Implementation of Alarm

Management Strategies to

Reduce the Incidence

of False AlarmsCLASS A75M451

UnityPoint Health ethodist/Proctor, Peoria,ILHOI AACN President cherrmann@frontier.com

www.Cherylherrmann.com

Cheryl Herrmanncherrmann@frontier.comwww.cherylherrmann.com

References

� AACN Practice Alert: Alarm Management 2013. Retrieved from www.aacn.org

� Frellick, M. (2013, April 24). Joint commission alerts hospitals to “alarm fatigue”. American Medical News. Retrieved from http://www.amednews.com/article/20130424/profession/130429992/8/

� George, What’s that sound? Managing alarm fatigue. NsgIncred Easy Sept/Oct 2014, 6-10.

� Graham, K. (2010) Monitor Alarm Fatigue: Standardizing use of physiological monitors and decreasing nuisance alarms. AJCC:19(1)28-34.

� Sendelbach, S., & Jepsen, S. (2013, April). Alarm management. Retrieved from http://www.aacn.org/wd/practice/content/practicealerts/alarm-management-practice-alert.pcms?menu=practice