Post on 25-Dec-2015
transcript
Implementing Evidence-Based Resident Work Schedules: Work
to Date and Future Directions
Children’s Hospital of PhiladelphiaNovember 2010
Christopher P. Landrigan, MD, MPHResearch Director, Children’s Hospital Boston Inpatient Pediatrics Service
Director, Sleep and Patient Safety Program, Brigham and Women’s HospitalAssistant Professor of Pediatrics and Medicine, Harvard Medical School
Disclosures
1) In the past 12 months, Dr. Landrigan has served as a paid consultant to: AXDev (non-profit organization), to assist in the development of a study of Shift Work Disorder; this work was supported by an unrestricted research grant from Cephalon, Inc. (commercial entity) to AXDev.
2) In addition, Dr. Landrigan has received monetary awards, honoraria, and travel reimbursement from multiple academic and professional organizations for delivering lectures on sleep deprivation and safety.
Patient Safety
• To Err is Human– Institute of Medicine Report, 1999
– estimated 44,000 to 98,000 deaths annually from adverse events
• Report notably silent on issue of sleep deprivation
Courtesy of D. Weaver, Univ Massachusetts Medical School, Worcester, MA
Human Circadian Pacemaker in Suprachiasmatic Nucleus (SCN) of Hypothalamus
SCN
Core Body Temperature
Plasma Melatonin
Eye BlinkRate
Slow EyeMovements
Stage 1 Sleep
Figure 1
Relative Clock Time (h)8 16 24 8 16
°C
36.5
37.0
37.5
pm
ol/L
0
100
200
% o
f 5-m
in e
pochs
0
8
16
# p
er 30-s
epoch
5
10
15
Elapsed Time Awake (h)
0 8 16 24 32
0
1
2
3
Relative Clock Time (h)8 16 24 8 16
Karolinska Sleepiness Scale
3
5
7
Cognitive Throughput
30
45
60
Elapsed Time Awake (h)
0 8 16 24 32
Probe Recall Memory Test
0
2
4
ms
# o
f A
ttem
pts
# o
f corr
ect P
air
s
200
500
1000
6000
Psychomotor Vigilance Performance
more
tir
ed
mean reaction timemedian reaction time
10% slowest reaction time
10% fastest reaction time
Cajochen et al. Am J Physiol 1999 National Transportation Safety Board Study (SS1995/01)
02468
101214161820
1:00AM
3:00AM
5:00AM
7:00AM
9:00AM
11:00AM
1:00PM
3:00PM
5:00PM
7:00PM
9:00PM
11:00PM
Num
ber o
f Acc
iden
ts
Temporal distribution of fatigue-relatedsingle vehicle truck accidents
• Biological Time of Day (circadian rhythms)
• Consecutive Waking Hours
• Night Sleep Duration
• Sleep Inertia
DETERMINANTS OF ALERTNESS AND PERFORMANCE
0 10 20 30 40 50
-30
-20
-10
0
10
20
Hours of Wakefulness
Su
bje
cti
ve A
lert
ness
Co
gn
itiv
e T
hro
ug
hp
ut
Hours of Wakefulness
Devi
atio
n fro
m M
ean (
mm
)
-30
-20
-10
0
10
0 10 20 30 40 50
Devi
atio
n fro
m M
ean (
# a
ttem
pte
d)
Jewett et al., 2000
Hours of Driving
Acute Sleep Deprivation and Performance
“...after [19] hours of sustained wakefulness (at 3 am) cognitive psychomotor performance decreased to a level equivalent to the performance impairment observed at a blood alcohol concentration of 0.05 %. ... After 24 hours of sustained wakefulness (at 8 am) cognitive psychomotor performance decreased to a level equivalent to the performance deficit observed at a blood alcohol concentration of roughly 0.10 %.” - D. Dawson and K. Reid, Nature 388: 235,
1997.
• Biological Time of Day (circadian rhythms)
• Consecutive Waking Hours
• Night Sleep Duration
• Sleep Inertia
DETERMINANTS OF ALERTNESS AND PERFORMANCE
Cumulative impact of daily sleep curtailmenton risk of vigilance lapses & subjective alertness
Van Dongen et al. Sleep 2003
Objective performance (PVT) Subjective awareness (KSS)
Pediatric Resident Chronic Sleep Deprivation and Performance
• “Heavy Call” (q5 schedule, ACGME compliant, average of 3h sleep on all) vs. Blood EtOH 0.04-0.05
012345678
mean lanevariability (feet)
mean speedvariability (mph)
Light callLight call w / alcoholHeavy call w / placebo
p=0.06
p=0.01
Arnedt et al. JAMA 2005
• Biological Time of Day (circadian rhythms)
• Consecutive Waking Hours
• Night Sleep Duration
• Sleep Inertia
DETERMINANTS OF ALERTNESS AND PERFORMANCE
Jewett et al., J. Sleep Res., 1999
Time Course of Deficits from Sleep Inertia
Fighters
All aircraft
1
2
3
4
Hours since 6 am wake time1 2 3 4 5 6 7
Aircraft Accident Data (Israeli Airforce)
Ribak et al., Aviat. Space Environ. Med., 1983
Sleep Inertia
Physiological Consequences of Healthcare Provider Schedules
Misalignment of circadian phase
Acute total sleep deprivation scheduled frequently
Chronic partial sleep deprivation resulting in cumulative sleep debt
Performance often required within minutes of awakening
Biological Time of Day (circadian phase)
Number of Hours Awake
Nightly Sleep Duration
Sleep Inertia
Sleep Deprivation and Errors in Detection of Cardiac Arrhythmias on ECG
9.64 ± 1.41
1.8 h
(0-3.8 h)
Sleep Deprived
p < 0.001
p < 0.001
Errors on ECG
sustained attention task
Sleep in prior 32 h
Medical Interns
5.21 ± 0.93
7.0 h
(5.5-8.5 h)
Rested
Friedman et al., N Engl J Med 285: 201, 1971
Grantcharov TP, Bardram L, Funch-Jensen P, Rosenberg J. BMJ 2001;323:1222-1223
Impaired speed and errors in
performance: laparoscopic
surgical simulator
•17-hour overnight call duty in a surgical department
•median reported sleep time 1.5 h (range 0-3 h)
Resident Performance and FatiguePhilibert I. Sleep 2005; 28: 1392-1402.
• Meta-analysis 60 studies (959 MDs, 1028 non-MDs) – For MDs, 24 hours with
no sleep leads to major performance drops to:
• 15th percentile of rested MD performance level
• 7th percentile on clinical tasks -4 -3 -2 -1 0 1 2 3 4
Standard Deviations
•Effect of Sleep Deprivation on Physicians’ Mean Clinical Performance: Results of 14 Studies
Harvard Work Hours, Health, and Safety Study
• National survey: To objectively quantify the work schedules experienced by house staff, and determine if increased hours are associated with increased risk of house staff injury– Study of a national sample of house staff
• 1,417 person-years monthly survey data collected from 2,737 interns nationwide in 2002-2003
– Monthly surveys– Work hours, crashes, and injuries– Correlation of work hours and motor vehicle crashes
Barger, L. K. et al. N Engl J Med 2005; 352:125-134
0
5
10
15
20
25
30
0 >0-1 >1-2 >2-3 >3-4 >4-5 >5-6 >6-7 >7-8 >8
Reported hours of sleep obtained during extended duration work shifts
Per
cen
tag
e o
f m
on
thly
rep
ort
s
Harvard Work Hours, Health, and Safety Study: Results
0
0.2
0.4
0.6
0.8
1
1.2
Crashes per 1000 commutes home
OR: 2.3 (95% CI, 1.6-3.3)
Barger LK et al. NEJM 2005; 352:125-134
0
0.2
0.4
0.6
0.8
1
1.2
1.4
Injuries per 1000 opportunities
Motor Vehicle Crashes Percutaneous InjuriesOR: 1.6 (95%CI, 1.5-1.8)
Ayas, et al. JAMA 2006; 296:1055-1062
Extended shifts
Non-extended shifts
Dose-Response Relationship
• For each additional extended duration work shift scheduled per month, interns had:– 8.8% increased monthly risk of any
Motor Vehicle Crash; – 16.0% increased monthly risk of a Motor
Vehicle Crash on the commute home from work
• Therefore, q3 schedule (10 overnights per month) = 160% increase over baseline risk
Harvard Work Hours, Health, and Safety Study: Part 4
• Objective: To determine if interns report making more harmful medical errors when working 24-hour shifts– Odds of reporting a harmful fatigue-related error
was 7-fold higher when working five or more 24h shifts in a month (compared with self when working no 24h shifts)
– Odds of a fatal error due to fatigue 4-fold higher
Barger, L.K. et al. PLoS Medicine 2006;3:e487
02468
1 01 21 41 61 8
E r r o r w / A d v e r s eo u t c o m e
F a t a l e r r o r
0 2 4 h s h i f t s
1 - 4 2 4 h s h i f t s
> = 5 2 4 h s h i f t s
OR 7.0 (4.3-11) OR 4.1 (1.4-12)
Errors/ 1000 person-months
Harvard Work Hours, Health, and Safety Study
• 1 of every 5 interns admitted making a fatigue-related mistake that injured a patient (700% when interns worked >24 hours in a row)
• 1 of every 20 interns admitted making a fatigue-related mistake that resulted in a patient’s death (300% in months interns worked five >24 hour shifts)
Randomized Trial with the following EXPERIMENTAL QUESTION:
Would ICU patients fare better when the physicians caring for them consisted of:
1. Current standard TRADITIONAL team of 3 residents working on a Q3 schedule which minimized handoffs by relying on repetitive 30-hour scheduled work shifts; or
2. An INTERVENTION team of 4 residents working on a schedule which increased handoffs in order to limit scheduled work shifts to no greater than 16 hours
Intern Sleep and Patient Safety Study
Intern A
Intern B
Intern C
Clock time (h)
TueMonSunSatFriThuWed
Intern 1
Intern 2
Intern 3
Intern 4
B
A
0 12 0 12 0 12 0 120 12 0 12 0 12 0
0 12 0 12 0 12 0 120 12 0 12 0 12 0
Traditional (A) vs. Intervention (B) Intern Schedule
Landrigan, C. P. et al. N Engl J Med 2004;351:1838-1848
Results: Sleep and Work Duration
Lockley, S. W. et al. N Engl J Med 2004;351:1829-1837
Attentional Failures at Night: 11pm-7am
•0.69 (traditional) vs. 0.33 (intervention) attentional failures per hour, p=0.02
•Non-significant trend toward decreased day / evening attentional failures as well
0
20
40
60
80
100
120
140
160
SeriousMedical Errors
- Total
SeriousMedication
Error
SeriousDiagnostic
Error
Traditional "q3" 24-30hour shifts
Intervention Schedule- <16 hour scheduledshifts
Serious Medical Errors•Interns made 36% more serious errors on traditional schedule, including 5 times as many serious diagnostic errors
Landrigan, C. P. et al. N Engl J Med 2004;351:1838-1848
p<0.001
p<0.001
p=0.03
Landrigan, C.P. et al. N Engl J Med 2004;351:1838-1848
Err
ors
per
10
00 p
t da
ys
Effects of Reducing or Eliminating Shifts >16h on Patient Safety and Quality of Care: a Systematic Review
Levine AC, Adusumilli J, Landrigan CP. Sleep 2010; 33: 1043-1053Source Setting Intervention Outcomes
Afessa et al. Medical ICU Transition from Q3 to a 14 hour work shift
model
No change in mortality or LOS
Bhavsar et al. Cardiology service
Elimination of extended shifts for senior
residents (no change for interns)
Improved guideline adherence, decreased
LOS, and improvement in 6-month mortality
de Virgilio et al.
Trauma service
Decrease from average of Q4.8 call to Q6.4
No change in complication rate or
mortality rateGoldstein et
al.Surgical inpatient service
Transition from Q4 to night float system with
12-14 hour shifts
Improvement in patient and nurse ratings of care
qualityGottlieb et al. VA medicine
serviceTransition from Q4 to
schedule with maximum 16h shifts
Decrease in LOS and medication errors
Horwitz et al. Medicine service
Elimination of extended shifts for residents (no
change for interns)
Decrease in ICU admissions and pharmacy
interventions to prevent med errors
Effects of Reducing or Eliminating Shifts >16h on Patient Safety and Quality of Care: a Systematic Review (continued)
Levine AC, Adusumilli J, Landrigan CP. Sleep 2010; 33: 1043-1053Source Setting Intervention Outcomes
Hutter et al. Surgical service
Reduction from Q3 to Q4 call
No change in complication or mortality
ratesLandrigan et
al. Medical and
Cardiac ICUsComparison of Q3 and 16 hour shift systems
Decrease in serious medical errors
Malangoni et al.
Surgical service
50% reduction in call shifts per month per
resident
Decrease in mortality
Mann et al. Radiology Department
Elimination of extended shifts without sleep for
radiology residents
Decrease in errors reading films requiring patient call back to ER
Sawyer et al. Surgical service
Comparison of interns on Q2, Q3, and Q4
No difference in error rates
n Significant improvement
No change Significant decrement
Summary 11 7 4 0
Reduction vs. Elimination of Shifts >16 hours
• Of programs that reduced the frequency of shifts >16 hours but did not eliminate them (e.g., transition from Q3 to Q4 schedule), 1/4 (25%) saw improvements in safety or quality
• Of programs that eliminated 24h shifts, 6/7 (86%) saw improvements in safety or quality
Did the 2003 ACGME duty hour standards solve the
problem?
• <80 hours per week, averaged over 4 weeks
• <30 hours in a row, including time for hand-offs of care and education
• 1 day off in 7, averaged over 4 weeks
Harvard Work Hours, Health, and Safety Study: Compliance with Duty Hour Standards
Work and Sleep, Before and After
‡ p<0.001‡ p<0.001
Landrigan C.P., et al. JAMA 2006;296:1063-1070
Compliance with ACGME Duty Hour Standards
• 83.6% of interns in violation of standards– 85.4% of programs; 90.8% of hospitals
• 44.0 % of all intern-months in violation – 61.5% of inpatient intern-months in violation
‡ p<0.001
Landrigan C.P., et al. JAMA 2006;296:1063-1070
Pediatric Duty Hours StudyLandrigan et al., Pediatrics 2008;122;250-258
• Specific Aim: To comprehensively measure the effects of the ACGME Duty Hour Standards in 3 major pediatric programs on:– house officer work and sleep hours– patient safety– house officer health and safety– medical education– quality of life
Hours of Work and Sleep
0
10
20
30
40
50
60
70
80
Mean Weekly WorkHours
0123456789
101112
Mean Nightly Hours of Sleep
Pre-
Post-
P=0.49
P=0.94
Safety-sensitive US industries
• Truckers : maximum 11 hours in a row
• Pilots : maximum 8h per 24 (domestic routes)
• Nuclear Power : maximum 12 hours
• Train engineers: maximum 12 hours
What are the Personal and Legal Implications of Residents Working >24
Hours?
Resident Car Crashes
• Heather Brewster: Student permanently brain damaged after being rear-ended by a resident from Rush-Presbyterian in Chicago who fell asleep at the wheel after having been awake for 34 of the preceding 36 hours
• Dr. Valentin Barbulescu: Senior resident from New York who died in a one-car crash after falling asleep at the wheel post-call from the CCU
Potential Legal Implications for House Staff
• In New Jersey, “driving after having been without sleep for a
period in excess of 24 consecutive hours” now explicitly considered reckless
– Similar laws pending in several other states
• Brewster case - Illinois court system – appellate court found hospital not liable for resident driving home from work after 36-hour shift, but appeal to Supreme Court of Illinois may be forthcoming
Potential Legal Implications for Hospitals
• Courts in two states have ruled that an employers may be held responsibility for fatigue-related crashes even after workers have left– similar to concept of restaurants / bars being
potentially liable for alcohol-related crashes if they served alcohol to a driver
• Employers have been sued and settled for as much as $10 million in a similar case
What about the risks of lost continuity of care?
Problems in Care Continuity• Night float admission patients had longer LOS
and more tests ordered Lofgren et al. J Gen Intern Med. 1990
• Work hour limits and presumed resulting discontinuities associated with increased hospital complications and test ordering Laine et al. JAMA 1993
• Cross coverage associated with an increased risk of errors (OR 6.0) Petersen et al., Ann. Int. Med 1994
– Signout errors an be improved substantially with structured sign out Petersen et al., Joint Comm J on QI 1998
Communication During Post-operative Patient Hand Off in the Pediatric Intensive Care Unit
Mistry KP, Landrigan CP, Goldmann DA, Bates DWCritical Care Medicine 2005; 33: A12.
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
medianmean
Nu
mb
er
of T
ran
sfe
rs
Number of Miscommunication Events
-Audio recording and analysis of 150 post-op sign-outs-100% of sign-outs contained at least one error
Signout Process : A survey of residents at BWH
Carty M, Smith C, Schnipper JL, Harvard Education Day 2004, unpublished data
• 37% said that signout occurred in a quiet place most of the time
• 52% provided written and verbal signout on every patient
• Only 55% of night-float residents said that when called about a patient, the relevant information could be found in the sign-out
What about the effects of lost experience on learning?
Does sleep loss have an effect on Learning?
Memory Consolidation During Sleep – Motor Sequence Task
Stickgold R. Nature 2005; 437: 1272-8
Visual Discrimination Learning Requires Visual Discrimination Learning Requires Sleep after Training*Sleep after Training*
0
5
10
15
20
25
30
0 1 2 3 4 7
Imp
rovem
ent
(ms)
*Stickgold et al., Nature Neurosci 3:1237, 2000
SLEEP
SLEEP
SLEEP
NO SLEEP
0
5
10
15
20
25
30
0 1 2 3 4 7
Imp
rovem
ent
(ms)
SLEEP
SLEEP
NO SLEEP
Visual Discrimination Learning Requires Visual Discrimination Learning Requires Sleep after Training*Sleep after Training*
*Stickgold et al., Nature Neurosci 3:1237, 2000
Stages of Sleep and Memory Consolidation
• Different Stages of Sleep are Correlated with Consolidation of Different Types of Memory– Visual discrimination task: Slow Wave
Sleep (Stages 3 and 4) and REM– Motor sequence task: Stage 2– Motor adaptation task: Slow Wave Sleep
Localized Slow Wave Sleep increase in Cortical Areas of Activation
Ghilardi, M. et al. Brain Res. 2000; 871: 127–145
Complex Tasks and Medical Education
• Mathematical and Geometrical Puzzle Solving show same type of sleep dependent improvement (Kuriyama et al. 2004; Learn Mem. 11: 705–713; Wagner et al. Nature 2004; 427: 352–355)
– Thus, complex as well as simple cognitive abilities affected
• No basic / laboratory research on sleep deprivation’s effects on medical learning but a number of field studies have emerged– Residents working longer hours report decreased
satisfaction with learning environment and decreased motivation to learn Baldwin et al 1997
Systematic Review: Effect of Reducing or Eliminating Shifts > 16h on Medical Education
Levine AC, Adusumilli J, Landrigan CP. Sleep 2010; 33: 1043-1053Source Population Intervention Outcomes
Afessa et al. Internal Med Residents
Transition from Q3 to a 14 hour work shift
model
No change in performance on post
rotation examBarden et al. Surgical
Residents Reduction of ICU Call
from Q2 to Q3; elimination of most call
on other rotations
No change in case volume; improved
ABSITE exam scores
Carey et al. OB/Gyn Residents
Decrease from Q3 to Q7 call
Improved CREOG exam scores
Cockerham et al.
Surgical Residents
Elimination of call with night float system
Reduction in % time spent on non-patient care,
significance not measuredDe Virgilio et
al.Surgical
Residents25% reduction in call
frequencyNo change in ABSITE scores; improved case
volumeFerguson et
al.Surgical
ResidentsReduction in call
frequency from q3 to q4No change in case
volume
Goldstein et al.
Surgical Residents
Transition from q4 to night float system
No change in case volume
Systematic Review: Effect of Reducing or Eliminating Shifts > 16h on Medical Education (cont)
Levine AC, Adusumilli J, Landrigan CP. Sleep 2010; 33: 1043-1053Source Population Intervention Outcomes
Hutter et al. Surgical Residents
Reduction in call frequency from q3 to q4
No change in ABSITE scores; attendings rated
resident skills lowerJarman et al. Surgical
Residents Transition from q3 and q4 to night float system
No change in percent of resident cases
Kelly et al. OB/Gyn Residents
Transition from q3 and q4 to night float system
No change in volume or CREOG exam scores
Malangoni et al.
Surg PGY 4/5 Residents
50% reduction in calls No change in percent of resident cases
McElearney et al.
Surgical Residents
Elimination of extended shifts with night float
No change in case volume
Sawyer et al. Surgical Interns
Comparison of interns on Q2,Q2,Q4
Improved case volume on Q4
Welling et al. Surgical Residents
Comparison of Q4 to night float system
No change in conference attendance
n Significant improvement
No change Significant decrement
Summary 14 4 9 1
Moving Forward
Resident Duty Hours: Enhancing Sleep, Supervision, and Safety
• Study commissioned by Congress, released by IOM in December 2008
• Concluded that it is unsafe for residents to work over 16 hours in a row without sleep
• Two solutions proposed:–Mandatory 5h nap during a 30h shift
or–16h shift limit
• Also called for improved handoff processes, increased supervision, and ongoing research to test best practices for implementation
Proposed 2011 ACGME Duty Hour Standards
• 16h consecutive work limit for Interns• PGY2s and higher continue to work 28h
shifts• Development of standards for supervision,
workload
European Working Time Directive
- 13 hours in a row maximum
- 48-56 hours per week
New Zealand
- 72 hours per week limit
- 16 hours in a row
- in place for 20 years
How do we move to a 16h Shift Limit for Interns?
• Must design schedules that respect principles of sleep and circadian medicine– Limit nights in a row; sufficient time off after night shift for
recovery; proven rotation patterns• Must not excessively overburden those residents
who remain on duty as others are home• Must improve sign-outs, and infrastructure for sign-
outs• Improve supervision• Education will need substantial redesign
– Changes in timing of didactic education• Creative scheduling• Less “on the fly” teaching; more structure required• ? Increased use of simuation / web
– ? Changes in rounding patterns
Slide courtesy of J Rothschild, MD, MPH
Rates of Complications by Attending Physicians After Performing Nighttime Procedures
Rothschild JM et al., JAMA 2009; 302(14):1565-1572
Remember…..
• Adjusting to a new shift system is not easy• Invariably, the first few steps are difficult and
adjustments to an intervention schedule need to be made– Iterative improvement is the key
• Infrastructure needs to change to reflect new schedules
• Changing lifestyle / sleep hygiene patterns is difficult but can be accomplished
Acknowledgements
• Harvard Work Hours, Health, and Safety Group– Especially Charles Czeisler, Steve Lockley, Laura
Barger, and Najib Ayas
• Center of Excellence in Patient Safety at BWH– Especially Jeffrey Rothschild and David Bates
• Children’s Hospital Boston, Brigham and Women’s Hospital and Harvard Medical School
• AHRQ and NIOSH