surveillance+monitoring++ at...

Surveillance Monitoring at

Dartmouth-‐Hitchcock

George Blike, MD Susan P. McGrath, PhD

Disclosures

•  Sue McGrath: No disclosures. •  George Blike: No disclosures. •  This presenta:on includes data from a before-‐and-‐a?er study at Dartmouth-‐Hitchcock Medical Center, which in December 2007 installed Masimo’s Pa:ent SafetyNet system. Masimo has supported the Hitchcock Founda:on, which helped to fund this research.

Value-‐Based Care: •  Safe Care-‐Elimina.ng iatrogenic pa.ent harm •  Appropriate & Effec.ve Care-‐Achieving the best

outcomes that pa.ents want and need •  Excellent Experience-‐Culture of caring and empathy •  Cost Effec.ve-‐Efficient with no waste © 2015 Dartmouth-‐Hitchcock

All Rights Reserved 3

Failure to Rescue—Conceptual Model

4

Hospitalization

Early signs and symptoms of physiological deterioration

Complication Arrest Death

Cascading “domino” physiological derangement

Terminal processes of multisystem organ failure and death

Increasing Physiological Instability t0# tdeath#

Ac#ve&Late*Stage&Research&Limited&Research&Ac#ve&Preven#on&Research&

© 2015 Dartmouth-‐Hitchcock All Rights Reserved 4

Failure to Rescue (FTR)

•  Failure to rescue –failure to detect and manage a clinically important deteriora:on, such as death or permanent disability from a complica:on of an underlying illness (e.g., cardiac arrest in a pa:ent with acute myocardial infarc:on) or a complica:on of medical care (e.g., major hemorrhage a?er thrombolysis for acute myocardial infarc:on).

•  FTR Rate = Death/1000 Complica:ons •  9 complica:ons considered

Source: AHRQ defini:on @ h^p://psnet.ahrq.gov/popup_glossary.aspx?name=failuretorescue


FTR drives hospital mortality rates

6

Ghaferi, Birkmeyer et. al.; Annals of Surgery • Volume 250, Number 6, December 2009


Environment at DHMC in 2006 Recogni:on that: •  Risks associated with anesthesia in post-‐op

environment were under-‐emphasized in rela:on to poten:al impact

•  Presence of technology was increasing, increasing complexity of the environment (PCA installa:on)

•  Nurse-‐to-‐Pa:ent ra:os were changing, and more tasks were being added to individual nurse task load

•  Preven:on of circumstances leading to deteriora:on is not possible

•  Rapid Response Teams were only part of the solu:on


.

Prior Work: ARTEMIS Field Triage System

Pulse Oximeter GPS Headset

Processor Unit (Gums:x or PDA)

Wireless Memory

Physiologic monitoring, state determina:on via triage algorithm, state communica:on

Assessment, interac:on, interven:on decision

Soldier or Emergency Responder Medic or EMT


R. Carella and S. McGrath, “ARTEMIS: Personal Area Networks for Emergency Remote Triage and Informa:on Management”, ISCRAM Conference, Newark, NJ, May 2006.

DHMC Lebanon, NH 25,000 Inpa:ent Discharges

90,000 Inpa:ent Days 8,500 Intermediate Care Days

24,000 ICU Days 30,000 ED visits

9 © 2015 Dartmouth-‐Hitchcock All Rights Reserved

1.7500&1.8000&1.8500&1.9000&1.9500&2.0000&2.0500&2.1000&2.1500&2.2000&

07/2012&

09/2012&

11/2012&

01/2013&

03/2013&

05/2013&

07/2013&

09/2013&

11/2013&

01/2014&

03/2014&

05/2014&

07/2014&

09/2014&

11/2014&

01/2015&

03/2015&

05/2015&

Case%Mix%Index,Inpa.ent%Only%

Case&Mix&Index&

InpaRent Surveillance Model

System Controller

Measurement System

Reference Signal

System Output

Measured Output

System Input

Measured Error

•  Interven:ons •  Devices

•  Mental model •  Knowledge •  Capacity

•  Measurement devices •  Observa:ons •  Algorithms


•  Data-‐based Change Management

•  Component selec:on •  Pulse oximetry •  Con:nuous monitoring •  All inpa:ents

•  Workflow analysis & integra:on

•  Educa:on •  Rounding

PaRent Surveillance System ImplementaRon


Alarm Design and Management

•  Leveraged design approach from other industries •  Wide Default thresholds (airbag sesngs) •  SpO2 < 80% •  PR < 50 BPM •  PR > 140 BPM

•  Averaging, directed no:fica:on and delays (e.g., SpO2) •  15 sec audible alarm at the bedside •  + 15 sec for nurse pager annuncia:on •  + 1 min sec for buddy/charge nurse pager escala:on

•  Threshold adjustments •  +/-‐ 10% by nurse •  Provider order (condi:on monitoring)


Alarms

CondiRonal Monitoring

Surveillance Monitoring

Alarms 15.62 3.7

0

2

4

6

8

10

12

14

16 Alarm Pages per Hour

Post

Pre

•  Typical alarm rates between two and four per pa:ent per 12 hour shi?.

•  Sample analysis of pager no:fica:ons indicates that over 85% of all alarm condi:ons are resolved within 30 seconds and over 99% before escala:on is triggered.

•  Observa:ons and alarm dura:on analysis suggest that the vast majority of audible-‐only alarms occur with the nurse at the bedside during system set up and sensor applica:on.

•  Units where leadership ac:vely encourages staff to incorporate trend analysis and integra:on of con:nuous vital signs monitoring into the pa:ent assessment ac:vi:es tend to have lower overall alarm and pager no:fica:on rates.


Alarm Management

We do not focus on sensi:vity, specificity and predic:ve value calcula:ons •  Time intensive data collec:on •  True/False Posi:ve/Nega:ve characteriza:ons are highly

subjec:ve due to complexity of tasks, feedback loops, nurse experience, percep:on of alarms, and many other factors

•  Context is cri:cal-‐ in a review of reported safety events (2012-‐2014) there were twice as many reports indica:ng that the “sensor off” alarm provided helpful pa:ent state informa:on as for any other category of report related to the surveillance system.


Alarm Management

Instead we focus on understanding the benefit/burden considera:ons and focus on using our limited resources on broad tac:cs such as: •  Ensuring standardiza:on of configura:ons and prac:ces via

policies and procedures •  Providing bedside educa:on around device opera:on and

troubleshoo:ng •  Encouraging educa:on around appropriate integra:on of

monitors for pa:ent state assessment


2006 Rapid Response Team

Jan 2007 Ins:tu:onal Outcome Data

Dec 2007 PSS on Orthopedics (Pilot unit)

Feb 2009 PSS on all Surgical Units

Apr 2010 PSS on all Medical Units

Feb 2012 PSS on Pediatrics

Mar-‐Apr 2012 PSS with acous:c respiratory sensors

FTR System Development Timeline

PSS: Pa:ent Surveillance System – the Dartmouth implementa:on of Pa:ent Safey Net ™


Pilot Performance

From: Taenzer AH, Pyke JB, McGrath SP, Blike GT. Impact of Pulse Oximetry Surveillance on Rescue Events and Intensive Care Unit Transfers. A Before-‐and-‐A?er Concurrence Study. Anesthesiology 2010, 112: 282-‐7

In a comparison of 11 months prior to and 10 months a?er implementa:on in a 36-‐bed orthopedic unit: •  50% reduc:on in transfers to higher levels of care •  60% reduc:on in rescue events There was no significant difference in these metrics in either of 2 comparison units during the same :me period.


Pediatric ImplementaRon

CondiRonal Monitoring

Surveillance Monitoring

Alarms 15.62 3.7

0

2

4

6

8

10

12

14

16 Alarm Pages per Hour

Post

Pre

56%

44%

The new Masimo devices have given me greater confidence in providing safe paRent care.

Yes No No Different

56%

44%


Yes No No Different

56%

44%


Yes No No Different


From: Mudge, B. et al. Reducing Monitor Alarms While Enhancing Pa:ent Safety. Children’s Hospital Network 2014 Organiza.onal Transforma.on Conference (2014)

DHMC

0.85

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

HE

RT

HERT Activations per 1,000 patient days

2.7

DHMC Average

0

1

2

3

4

5

6

Tran

sfer

s

Transfers to Critical Care

Pilot Unit 8 Years Later


DHMC

0.14

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Cod

es

Code Activations per 1,000 patient days

DHMC

11.50

0.0

5.0

10.0

15.0

20.0

25.0

Con

sults

Life Safety Consults per 1,000 patient days

0

20

40

60

80

100

120

2009 2010 2011 2012 2013 2014 2015

COUNT OF RE

VERS

ALS

YEAR

Life Safety Consults and Rescue Team Calls Associated with Opioid Reversals in Surveillance Unts, 2009-‐2015 YTD

No Call

Life Safety

HERT

Surveillance System Data


Hospital-‐wide Code DistribuRons and Trends

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

2010 2011 2012 2013 2014 2015 YTD

% Arrests by Type Not Assigned

hemorrhaged leading to hypotension and arrest Not On List -‐ Other

Respiratory

Brady

VF

VT

Asystole/PEA


0

2

4

6

8

10

0

50

100

150

200

2010 2011 2012 2013 2014 2015 Ra

te

Code

s

Codes

DHMC Codes

DHMC Rate Per 1000 Pa:ent Days

0.37/1000 pt days medicine units 0.11/1000 pt days surgical units

Pilot Unit 8 Years Later

86% 79%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Feb-

14

Apr

-14

Jun-

14

Aug

-14

Oct

-14

Dec

-14

Feb-

15

Apr

-15

Jun-

15

Aug

-15

% E

vent

s

% of Events With Surveillance Data 6 Month Rolling Average

Unit Mean DHMC Mean

0:40 0:51

00:00 00:30 01:00 01:30 02:00 02:30 03:00 03:30 04:00 04:30 05:00

Del

ay (h

h:m

m)

Average Delay to Activate 6 Month Rolling Average

Unit Mean DHMC Mean


0

10

20

30

40

50

60 CO

UNT OF RE

SCUE EV

ENTS

UNIT

Post Event PaRent DisposiRon

Step Down

Cri:ca Care

Stayed in Room

Not On List -‐ Other/Blank Died

© 2015 Dartmouth-‐Hitchcock All Rights Reserved

23

Risk Assessment and PopulaRon Differences

FTR Learning Lab—Rapid Learning/InnovaRon

25

Technology

People

Workflows

Culture

Candidate Solu:ons •  Tech-‐Early

Detec.on •  People-‐Ac.on

Teams •  Integrated System-‐

Resilience Engineering Principles

1—IniRal evaluaRon of soluRons

2—Prototyping and sim-‐based

tesRng

3—Element ready for integraRon and

field tesRng

2

3

Ideal Rescue System

1


Summary

•  Con:nuous monitoring with pulse oximetry in general care sesngs has a posi:ve impact on pa:ent safety and is not cost prohibi:ve

•  Benefit-‐burden analysis with respect to nursing tasks and cogni:ve impact remains posi:ve

•  Standardiza:on and con:nuing educa:on are key elements to maintaining results

•  More work is needed to increase understanding of the socio-‐technical aspects of pa:ent state assessment and response systems


surveillance+monitoring++ at...

Documents