-
The interaction of OSA, Fatigue, & Sleep:
A Performance Model
November, 2011
Steven R. Hursh, Ph.D
-
Operational Definition
DOT Human Factors Coordinating Committee, 1998
● Fatigue is more than sleepiness and its effects are more than falling asleep.
● Fatigue is a complex state characterized by a lack of alertness and reduced mental and physical performance, often accompanied by drowsiness.
2
-
Symptoms versus Root Causes
Symptoms
Operational Consequences
● Measurable Changes in Performance
● Lapses in attention and vigilance
● Delayed reactions
● Impaired logical reasoning and decision-making
● Reduced “situational awareness”
● Low motivation to perform “optional” activities
● Poor assessment of risk or failure to appreciate consequences of action
● Operator inefficiencies
● Fatigue is one
potential root cause.
● No direct measure,
physiological marker,
or “blood test” for
fatigue.
3
Root Cause Analysis
-
Modeling Provides an Objective
Metric for Fatigue ● The conditions that lead to fatigue are well
known.
● A fatigue model simulates the specific
conditions and determines if fatigue could be
present.
● The model can estimate the level of
degradation in performance and provide an
estimate of fatigue risk.
4
-
5
ALERTNESS & COGNITIVE PERFORMANCE
CIRCADIAN
RHYTHM CUMULATIVE
SLEEP DEBT
ALERTNESS &
COGNITIVE
PERFORMANCE
Time of Day Sleep History and Time on Duty
Daily Variations in Effectiveness
-
SAFTE
● The Sleep, Activity, Fatigue, and Task Effectiveness (SAFTE) Model is based on 17 years of fatigue modeling experience.
● Validated against laboratory and simulator measures of fatigue.
● Validated and calibrated to predict accident risk by the Department of Transportation.
● Peer reviewed and found to have the least error of any available fatigue model.
● Accepted by the US DOD (Air Force, Army, Navy, Marines) as the common warfighter fatigue model.
6
-
SAFTE Model Components
7
-
Practical Software for Fatigue
Assessment and Management
● Fatigue Avoidance Scheduling Tool (FAST)
● FAST is a fatigue assessment tool using the SAFTE model
● Developed for the US Air Force and the US Army.
● Department of Transportation has sponsored work leading to
enhancements for transportation and industrial applications.
● DOT validated and calibrated.
8
Now available as a tool for commercial applications.
-
9
Walter Reed Restricted Sleep Study PVT Speed
Chronic Restriction Adaptation
50
65
80
95
110
0 T1 T2 B E1 E2 E3 E4 E5 E6 E7 R1 R2 R3Day
Mea
n S
pe
ed
(as a
% o
f B
ase
line
)
9 Hr
7 Hr
5 Hr
3 Hr
PVT Speed
Chronic Restriction Adaptation
50
65
80
95
110
0 T1 T2 B E1 E2 E3 E4 E5 E6 E7 R1 R2 R3Day
Me
an
Sp
ee
d
(as a
% o
f B
ase
line
)
9 Hr
7 Hr
5 Hr
3 Hr
SAFTE/FAST R2 = 0.94
Restriction Recovery Baseline
Belenky, G, Wesensten, NJ, Thorne, DR, Thomas, ML, Sing, HC, Redmond, DP, Russo, MB, and Balkin, TJ (2003). Patterns of performance
degradation and restoration during sleep restriction and subsequent recovery: a sleep dose-response study. Journal of Sleep Research, 12, 1-12.
-
Two-Step Sleep/Performance Model
Sleep Input
Actigraph or Logbook
Physiological Model Sleep
Estimator
Work Schedule 10
OSA
-
Questions
1. Can actigraphy detect sleep disturbances
produced by sleep apnea?
2. Can actigraphy detect differences in
severity of sleep apnea?
3. Do these differences translate into
meaningful differences in estimated sleep
and predicted performance?
4. Will exposure to feedback based on
actigraphy and predicted performance
reinforce greater CPAP use and adherence
to treatment protocols?
-
Single Day of Activity, Example
Noon, March 24 to noon, March 25
Major Sleep Period – Time in Bed
-
Complete 24 hr Records for the Entire Month
Before CPAP
After CPAP
-
Focus on Sleep Periods Only
Before CPAP
After CPAP
-
Pre- and Post-CPAP Activity Patterns
Post-CPAP: Noon, April 12 to noon, April 14
Major Sleep Period – Time in Bed
Pre-CPAP: Noon, March 24 to noon, March 25
Major Sleep Period – Time in Bed
-
Sleep Quantity per Day
Sleep Hours Per Day (Two Days Means)
0
2
4
6
8
10
12
3/18/2006 3/23/2006 3/28/2006 4/2/2006 4/7/2006 4/12/2006 4/17/2006
Days
Ho
urs
Untreated
Treated
CPAP TreatmentUntreated Sleep Apnea
-
Sleep Fragmentation Before and After
Major Sleep Episodes per Day (Separated by > 2.5 min Awake)
0
2
4
6
8
10
12
14
16
3/18/2006 3/23/2006 3/28/2006 4/2/2006 4/7/2006 4/12/2006 4/17/2006
Days
Ep
iso
de
s
Untreated
Treated
CPAP TreatmentUntreated Sleep Apnea
-
How does this translate to better
brain functioning and Performance?
● The principle effect of sleep loss is reduced
cognitive ability and performance lapses.
● Difficult to measure directly.
● Fatigue model can predict how the “average”
person’s performance would improve with
increased of sleep.
● Performance feedback could be provided
daily to reinforce adherence.
● Does this additional information matter?
-
0
200
400
600
800
1000
1200
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
mse
c
PVT trial
Lapses (RT > 500ms) Mean = 6.07 SD = 6.61
(Note: the ‘down’ periods (green) were calculated with the automatic option – to illustrate how this calculation fails for this data set)
-
Hypothetical “Normal” Sleeper Performance while awake (black line) is always above 90%
-
Before CPAP After CPAP
12 Hr Sleep
Performance Estimate Based on Sleep Recording Performance gradually improves to near ideal after the CPAP
Performance Effectiveness
Lapses
-
Daily Performance Before and After CPAP
0
10
20
30
40
50
60
70
80
90
100
110
3/2
0/2
00
6
3/2
1/2
00
6
3/2
2/2
00
6
3/2
3/2
00
6
3/2
4/2
00
6
3/2
5/2
00
6
3/2
6/2
00
6
3/2
7/2
00
6
3/2
8/2
00
6
3/2
9/2
00
6
3/3
0/2
00
6
3/3
1/2
00
6
4/1
/20
06
4/2
/20
06
4/3
/20
06
4/4
/20
06
4/5
/20
06
4/6
/20
06
4/7
/20
06
4/8
/20
06
4/9
/20
06
4/1
0/2
00
6
4/1
1/2
00
6
4/1
2/2
00
6
4/1
3/2
00
6
4/1
4/2
00
6
4/1
5/2
00
6
4/1
6/2
00
6
Effectiveness
Percent Below 77Pre-CPAP Post-CPAP
Fully Rested Range
.08 BAC
.05 BAC
Percent of Awake Time Below 77 (0.05 BAC)
Average Effectiveness
-
0
2
4
6
Pre-CPAP Post-CPAP
Fully Rested Range
Lapse Likelihood
Daily Tendency to Have Lapses
-
Proposed Studies
● Study 1: Descriptive analysis of actigraph
recordings of patients with sleep apnea
before and after CPAP treatment.
● Study 2: Controlled study of the benefits of
sleep and performance feedback.
● Protocols have been developed and study is
planned for the near future.
-
Questions?
Steven R. Hursh
-
Experiment 1: Actigraphy Measures of
Sleep Apnea and CPAP Therapy Technological Proof-of-Concept/Validation
● Enroll equal numbers of patients diagnosed as Mild, Moderate, &
Severe OSA. Case-control so that all groups are equivalent in as many
parameters as possible prior to study. Ideally, we'd have a similarly
case-controlled group of non-sleep-disordered subjects, but not
necessary.
● Outfit each subject with Actigraph for continuous monitoring and
provide a laptop or netbook computer for data upload:
1-2 weeks of no CPAP baseline followed by 2-4 weeks of CPAP usage (realistically,
one week of baseline + 2 weeks CPAP)
No feedback, simply record and upload the data
Analyze sleep parameters as a function of OSA severity with covariate of compliance
rate.
● Answer the simple question: Does actigraphy detect sleep benefits of
CPAP in OSA patients? If it does not, then feedback would be arbitrary
and irrelevant. If it does, then proceed to Experiment 2.
-
0
200
400
600
800
1000
1200
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
mse
c
PVT trial
Lapses (RT > 500ms) Mean = 1.28 SD = 1.44
-
Experiment 2: Effects of Actigraphy
Feedback on CPAP Compliance
● Enroll equal numbers of Mild, Moderate, & Severe OSA patients. Case-
control and similar to Experiment 1.
● Outfit each subject with Actigraph for continuous monitoring and
provide a laptop or netbook computer for data upload.
● Randomly assign half to Feedback or No Feedback conditions.
The No Feedback condition is identical to Exp 1.
In the Feedback condition, upon daily data upload, each subject receives a computer-
generated report on their laptop with the following individualized data: Line chart of total sleep, sleep efficiency, and wake episodes for each day of the study (accumulates as study period progresses)
SAFTE/FAST chart of predicted performance effectiveness (accumulates as study period progresses)
Performance effectiveness: % change from pre-CPAP baseline and % change from previous day
● If we want to save time and money, and we're not worried about time-of-
year or other cohort effects, we could simply compare the feedback
groups in Exp 2 to the subjects in Exp 1 who received no feedback, so
long as they are equally case-controlled.
● Answer the Question: Does sleep and performance feedback
increase compliance with CPAP therapy.