TIME-ON-TASK DECREMENTS
Time-on-Task Decrements in “Steer Clear” Performance of
Patients with Sleep Apnea and Narcolepsy
Larry J. Findley, MD,1 Paul M. Suratt, MD,2 David F. Dinges, PhD3
1Sleep
Disorders Center of Northern Colorado, Loveland, Colorado, U.S.A.; 2University of Virginia School of Medicine,
Charlottesville, Virginia, U.S.A.; 3Unit for Experimental Psychiatry, Division of Sleep and Chronobiology, Department of
Psychiatry, and Center for Sleep and Respiratory Neurobiology, University of Pennsylvania School of Medicine,
Philadelphia, Pennsylvania, U.S.A.
Summary: Loss of attention with time-on-task reflects the increasing instability of the waking state during performance in experimentally induced sleepiness. To determine whether patients with disorders of excessive sleepiness also displayed time-on-task
decrements indicative of wake state instability, visual sustained attention performance on "Steer Clear," a computerized simple RT
driving simulation task, was compared among 31 patients with untreated sleep apnea, 16 patients with narcolepsy, and 14 healthy
control subjects. Vigilance decrement functions were generated by analyzing the number of collisions in each of six four-minute periods of Steer Clear task performance in a mixed-model analysis of variance and linear regression equations. As expected, patients
had more Steer Clear collisions than control subjects (p=0.006). However, the inter-subject variability in errors among the narcoleptic
patients was four-fold that of the apnea patients, and 100-fold that of the controls volunteers; the variance in errors among untreated apnea patients was 27-times that of controls. The results of transformed collision data revealed main effects for group (p=0.006),
time-on-task (p=0.001), and a significant interaction (p=0.022). Control subjects showed no clear evidence of increasing collision
errors with time-on-task (adjusted R2=0.22), while apnea patients showed a trend toward vigilance decrement (adjusted R2=0.42,
p=0.097), and narcolepsy patients evidenced a robust linear vigilance decrement (adjusted R2=0.87, p=0.004). The association of
disorders of excessive somnolence with escalating time-on-task decrements makes it imperative that when assessment of neurobehavioral performance is conducted in patients, it involves task durations and analyses that will evaluate the underlying vulnerability of potentially sleepy patients to decrements over time in tasks that require sustained attention and timely responses, both of
which are key components in safe driving performance.
Key words: Steer clear performance; vigilance; DOES; OSAS; narcolepsy
INTRODUCTION
reflect the increasing instability of the waking state in the
face of a continuing requirement for sustained atention.3,4,12
Measures of time-on-task performance requiring sustained attention and rapid reaction times are especially relevant to determining the degree of impairment that might
be present relative to motor vehicle operation. Surprisingly,
time-on-task decrements rarely have been evaluated in
patients with disorders of excessive somnolence, where
concerns about driving safety can be paramount. Both
untreated patients with obstructive sleep apnea (OSA),13,14
and patients with narcolepsy15 have impaired performance
on tasks requiring sustained attention and rapid reaction
times. Accelerated time-on-task decrement functions in
PVT performance4 have been observed in patients with
OSA, both before treatment with nasal continuous positive
airway pressure (CPAP) and after treatment withdrawal.3,16
Concern that the excessive daytime sleepiness of
patients could increase the likelihood of motor vehicle
crashes led to the development of “Steer Clear,” a 30-
Time-on-task decrements refer to decreases in performance as a function of increasing time performing. Such
decrements are especially evident on vigilance tasks and
performance requiring sustained attention. 1-3 When
observed during simple psychomotor vigilance task (PVT)
performance4 time-on-task decrements reflect the increasing difficulty of subjects to attend and sustain timely
responses to salient signals.5 The developing loss of sustained attention and of timely responses as task duration
increases have been demonstrated to be hallmarks of experimentally induced sleepiness.2,3,5,6-11 In sleep-deprived subjects, the loss of attention with time-on-task appears to
Accepted for publication June 1999
Corresponding Author: David F. Dinges, PhD, Unit for Experimental Psychiatry,
Division of Sleep and Chronobiology, University of Pennsylvania School of
Medicine, 1013 Blockley Hall, 423 Guardian Drive, Philadelphia, PA 19104,
U.S.A., Phone: 215-898-9949, Fax: 215-573-6410
SLEEP, Vol. 22, No. 6, 1999
804
Time-on-Task Decrements—Findley et al.
minute, computer-generated, visual vigilance, driving simulation task requiring sustained attention and simple reaction times. 17 The total number of Steer Clear performance
failures, which are “collisions” with computer-generated
obstacles (i.e., steers) appearing on a simulated two-lane
roadway, have been shown to be increased in both untreated patients with sleep apnea and those with narcolepsy.17,18
However, time-on-task decrements have not been systematically assessed for performance on Steer Clear. Moreover,
there have been no comparisons of time-on-task decrements in patients with untreated sleep apnea and those with
narcolepsy, relative to healthy control subjects.
Establishing the magnitude of time-on-task vigilance
decrement functions on Steer Clear would allow a better
quantification of the severity of deficits in sustained attention experienced in different disorders of excessive somnolence, and confirm that Steer Clear assesses some of the
same functional deficits as seen in PVT performance in
sleep-deprived subjects.3 This study provides the first systematic evaluation of vigilance decrement functions in
Steer Clear performance, through examination of response
variability as a function of time-on-task, in patients with
sleep apnea, narcoleptics, and normal controls.
had at least two MSLT naps with REM sleep). All narcolepsy patients had less than 5 apneas + hypopneas per
hour of sleep during nocturnal sleep. Both sleep apnea and
narcolepsy patients were studied at the University of
Virginia Sleep Disorders Clinic or the Sleep Disorders
Center of Northern Colorado.
The third subgroup consisted of 14 volunteer subjects
(11 men and 3 women) without sleep apnea or narcolepsy,
who were in the same age range as the patient subgroups
(mean age=43±15 years). These 14 volunteers were
employees at the University of Virginia who had no symptoms of either sleep disturbance or excessive daytime
sleepiness.
Vigilance Performance Testing on Steer Clear
Visual vigilance performance was assessed using the
computer program “Steer Clear.”17 The program displays
an automobile moving on a two-lane highway. During the
30 minutes of the task, a total of 780 obstacles (steers)
intermittently appeared in the automobile's lane with the
average inter-stimulus interval. To avoid hitting an obstacle, the subject had to “move” the automobile to the other
lane by pressing the space bar on the keyboard (simple
reaction time response).
One 30-minute trial of Steer Clear was performed by
each patient with sleep apnea or narcolepsy before (i.e.,
7:00 P.M. to 9:00 P.M. ) the diagnostic nocturnal sleep study.
Normal volunteers were studied earlier in the day (i.e., 4:00
P.M. to 6:00 P.M.). Prior to the task, subjects were instructed
by a technician on how to perform the task and that they
should do their best to avoid hitting the steer obstacles.
They were then given a one-minute practice test, and then
left alone in a quiet room to perform the test. When the test
was completed, the program calculated the number of
errors (i.e., proportion of collisions with obstacles).
Vigilance decrement functions were generated by taking
the number of collisions in each of six four-minute periods
of task performance (the Steer Clear task had three twominute rest breaks, occurring approximately at three to six
minute and 14-17 minute into the task). The protocol was
approved by the Human Subject's Committee of the
University of Virginia.
METHODS
Subjects
Sleep and subsequent waking performance on Steer
Clear were studied in a total of 61 subjects. Thirty-one subjects were polysomnographically documented to be suffering from untreated obstructive sleep apnea. The number of
apneas and hypopneas (i.e., decrease in tidal breathing
associated with a drop in baseline oxyhemoglobin saturation of three percentage points) were calculated for each
subject, and sleep apnea was defined by at least 10 obstructive apneas plus hypopneas per hour of sleep (AHI mean=
46±30 [SD] per hour of sleep; range 10/hr - 99/hr). The
sleep apnea group was comprised of 27 men and 4 women
(mean age=45±5.5 years). All of these patients complained
of loud snoring and excessive daytime sleepiness. Sleep
apnea patients with a comorbid history of narcolepsy,
seizure disorder, chronic lung disease, chronic sedative
intake, or alcohol abuse were excluded.
A second group of 16 patients were diagnosed
polysomnographically as having untreated narcolepsy. This
group was comprised of nine men and seven women (mean
age = 38±19 years), and all had at least one of the following symptoms: sleep attacks, cataplexy, or sleep paralysis.
Each narloleptic subject underwent a five-nap multiple
sleep latency test (MSLT) performed using standard techniques (Mitler, 1982), and each had a mean sleep latency of
less than 10 minutes (group mean=6.1±4.7 minutes) and at
least one MSLT nap with rapid eye movement sleep (mean
= 2.9±1.2 REM naps per subject; 94% of the 16 patients
SLEEP, Vol. 22, No. 6, 1999
Statistical Analysis
The average and inter-subject variability in the number
of collisions (errors) during Steer Clear were evaluated
among the three groups. For statistical analyses, data were
transformed (√x + √[x + 1]) to adjust for heterogeneity of
variance (i.e., proportionality between mean and variance
typically seen in the performance of sleepy subjects).2 A
one-way analysis of variance (ANOVA) and post-hoc
Tukey tests were used to compare the total number of task
errors among the three groups. A mixed-model ANOVA
805
Time-on-Task Decrements—Findley et al.
involving the three groups (between subjects) evaluated
across six consecutive Steer Clear task four-minute time
bins (within subjects) was used to test performance differences as a function of time-on-task. Significance levels
were corrected for sphericity by Greenhouse-Geisser
epsilon (GGε), and significant F-ratios were further analyzed by single-degree-of-freedom polynomial contrasts,
for linear, quadratic and cubic trends across time-on-task
bins. To further clarify the effects of group on time-on-task
decrements, comparisons were also made between control
subjects and patients with apnea, between control subjects
and patients with narcolepsy, and between apnea patients
and narcolepsy patients. Linear regression equations were
fit to mean errors across consecutive four-minute time bins
within each group to evaluate vigilance decrements with
time-on-task. To identify factors most likely to be associated with performance decrements, Spearman correlations
were calculated between task errors on Steer Clear performance and clinical variables for the 31 patients with apnea
(i.e., apnea + hypopnea index; O 2 nadir in NREM sleep; O 2
nadir in REM sleep; number of arousals during sleep; percentage of stage 1, stage 2, stage SWS, and stage REM
sleep; and n = 15 patient ratings on the Stanford Sleepiness
Scale [SSS] and Epworth Sleepiness Scale [ESS]); and for
the n=16 patients with narcolepsy (i.e., MSLT sleep latency and number of naps with REM sleep).
Figure 1—Means and standard deviations for the total number of errors
(obstacles collided with) during a 30-minute trial of performance on Steer
Clear for 14 healthy control subjects, 31 patients with untreated sleep apnea,
and 16 patients with untreated narcolepsy. See text for discussion of the differences among groups in mean errors and inter-subject variability.
minute time bins within each Steer Clear trial) are shown in
Figure 2. The ANOVA on transformed time-on-task data
yielded a statistically significant main effect for group
(F2,58 = 5.63, p = 0.006), a significant main effect for timeon-task (F 5,290 = 5.226, p = 0.001; GGε = 0.718), and a significant interaction (F 10,290 = 2.39, p = 0.022). Polynomial
contrasts revealed only a significant linear trend for timeon-task (F1,58 = 12.92, p = 0.001). Linear functions fit to the
time-on-task data within each group are displayed in Figure
2. An increasing slope indicates a more severe vigilance
decrement function. While control subjects showed no
clear evidence of increasing collision errors with time-ontask (adjusted R2 = 0.22, F1,4 < 1), a trend toward a group
vigilance decrement was evident for apnea patients (adjusted R2 = 0.42, F1,4 = 4.65, p = 0.097). In contrast, narcolepsy patients evidenced a clear decrement with time-on-task
(adjusted R2 = 0.87, F1,4 = 34.63, p = 0.004).
Separate two-group ANOVAs were performed comparing time-on-task decrements (transformed collision errors)
between groups. For the ANOVA involving control subjects and patients with apnea, there was a trend for a main
effect for group (F1,43 = 3.95, p = 0.053) and a significant
group-by-time-on-task interaction (F5,215 =3.05, p=0.020;
GGε=0.767). These results indicate that apnea patients had
vigilance decrements with time-on-task, but control subjects did not (Figure 2). Not surprisingly, the ANOVA comparing control subjects and patients with narcolepsy yielded a significant main effect for group (F1,28=8.48, p=
0.007), a significant main effect for time-on-task (F 5,140 =
RESULTS
The 14 volunteers collided with an average of 1.1%±
0.2% of obstacles; the 31 patients with sleep apnea collided with an average of 3.5%±0.9% of obstacles; and the 16
patients with narcolepsy collided with an average of 6.9%
±2.4% of obstacles. There were significant differences
among the three groups in the total number of obstacles
collided with during the task (F2,58=4.70, p=0.013). Figure
1 displays these results. As suggested by the widely disparate standard deviations in Figure 1, the inter-subject
variability in errors among the narcoleptic patients was 4fold greater than that of the apnea patients, and 100-fold
greater than that of the controls volunteers; the variance in
errors among apnea patients was 27-times that of controls.
When these disparities were adjusted for by transformation,
significant differences remained among groups (F 2,58=
5.54, p=0.006). Tukey HSD post-hoc tests (using a mean
square error of 30.13, df=58) on transformed total errors
and revealed that narcoleptic patients had significantly
more collision errors than healthy volunteers (HSD=6.57, p
=0.005), and tended to have more collisions than patients
with sleep apnea (HSD=3.92, p=0.06). Patients with apnea
did not have more errors than control subjects (HSD = 2.64,
p = 0.30).
Analyses of differences among groups in collision errors
as a function of time-on-task (i.e., across consecutive 4SLEEP, Vol. 22, No. 6, 1999
806
Time-on-Task Decrements—Findley et al.
patients were for correlation coefficients between Steer
Clear collisions and the number of arousals during sleep
(rho=0.25), the percentage of stage 2 sleep (rho=- 0.34) and
REM sleep percentage (rho=0.39). Neither SSS ratings (n
=15, rho=- 0.02), nor ESS scores (n=15, rho=0.04) were
correlated with performance errors within apnea patients.
The highest coefficient resulting from analyses within the n
= 16 narlocepsy patients was for the correlation between
Steer Clear collisions and MSLT sleep latency (rho=- 0.33).
Too few patients with narcolepsy (n=5 of 16) had SSS or
ESS scores for correlational analyses. Among normal control subjects, age was significantly correlated with Steer
Clear collisions (n=14, rho=0.714).
DISCUSSION
Persons with untreated sleep apnea or narcolepsy had
poorer performance overall and greater time-on-task decrements on a 30-minute “Steer Clear” test than did comparably-aged control subjects. The differences were especially
marked in patients with narcolepsy. Performance by narcoleptics was significantly worse than that of patients with
sleep apnea, and narcoleptic patients averaged the steepest
time-on-task decrement function. Although complaints of
excessive daytime sleepiness are well-documented among
untreated patients with sleep apnea and those with narcolepsy,19 the data obtained here suggest that narcolepsy
patients not only suffer a greater degree of performance
impairment than do sleep apnea patients, but also that they
have far greater difficulty sustaining performance as task
duration increases. The correlation coefficient (rho=0.33)
between Steer Clear collisions and MSLT latency suggested that narcolepsy patients with shorter latencies had more
performance errors.
Although superior to that of untreated narcoleptic
patients, the Steer Clear performance of the patients with
untreated sleep apnea tended to be inferior to that of control subjects. Patients with sleep apnea performed comparable to control subjects during the first four-min. epoch of
Steer Clear, but their performance subsequently deteriorated relative to controls, indicative of a reduced ability to
sustain Steer Clear performance with time-on-task. Such
increased vigilance decrement functions in untreated
OSAS patients have been observed for performance on the
psychomotor vigilance task, and their reduction following
treatment with nasal CPAP suggests that they are associated with impairment induced by OSAS. 3 However, neither
indices of OSAS severity (AHI, O2 nadir, and number of
arousals) nor of subjective sleepiness (SSS and ESS) were
statistically significantly correlated with Steer Clear collisions among untreated OSAS patients in the current study.
This result is consistent with other studies in untreated
OSAS patients,21 and it suggests that patients may differ in
their vulnerability to neurobehavioral impairment from
sleep-disordered breathing via mechanisms that are only
Figure 2—Analyses of differences among groups in collision errors as a function of time-on-task. Each data point represents the average number of errors
for each consecutive 4-minute time bin during the 30-minute Steer Clear task.
Mixed model analysis of variance yielded a main effect for group (p = 0.006),
a main effect for time-on-task (p = 0.001), and a group by time-on-task interaction (p = 0.022). Polynomial contrasts revealed only a significant linear
trends for time-on-task (p = 0.001). Least squares linear functions were fit to
the time-on-task data within each group. An increasing slope indicates a
greater vigilance decrement function. See text for discussion of specific differences among groups.
3.01, p=0.035;GGε=0.596), and a significant interaction
(F5,140=3.34, p=0.023), indicating that narcoleptics had
worse performance than controls at all times of the task,
and markedly worse vigilance decrements as task duration
continued (Figure 2). A final ANOVA comparing performance of apnea patients with that of narcoleptics yielded a
significant main effect for group (F1,45=4.48, p=0.040), and
a significant main effect for time-on-task (F5,225= 8.20,
p=0.0001; GGε=0.665), with no interaction (F5,225 = 1.26,
p=0.288). Performance of both groups declined with timeon-task, but overall narcoleptics had more errors than apneics and steeper vigilance decrement functions (Figure 2).
Spearman correlations between task errors on Steer
Clear performance and clinical variables were generally
near zero and not statistically significant within either
apnea patients or narcolepsy patients. The highest coefficients resulting from analyses within the n=31 apnea
SLEEP, Vol. 22, No. 6, 1999
807
Time-on-Task Decrements—Findley et al.
loosely associated with objective indices of disease severity or subjective measures of patients' sleepiness.
To the extent that Steer Clear makes continuous
demands on sustained attention, the time-on-task deficits
observed in patients with sleep apnea and with narcolepsy
are signs of limited vigilance capability. The presence of
such a "decrement function"20 or "vigilance decrement"1
across time-on-task in patients with disorders on excessive
somnolence is consistent with the findings of accelerated
vigilance decrement functions in the PVT performance of
healthy, sleep-deprived subjects.2,22 The processes involved
in these escalating vigilance decrements are not entirely
understood, but likely include increased instability of the
waking state when sleepiness occurs that is manifested in a
growing effortfulness to remaining awake, and a heightened vulnerability in the presence of sleepiness to factors
that modulate response capacity. 12 For example, when
attempting to perform while sleepy, alert sustained attention requires effort. A deficit in vigilance with time-on-task
(i.e., a vigilance decrement function) can be exposed even
in moderately sleepy persons if the task has a relatively
high signal load and specifically demands a timely
response to each signal–Steer Clear and the PVT have these
characteristics but the more classic vigilance tasks used in
early sleep deprivation lacked the high signal rate and
failed to require or carefully measure response speed. 2,3
Increased lapsing (and therefore its presumed physiological correlate, microsleeps) has been found to contribute
to vigilance decrement with time-on-task.3 However, it
appears that not only the frequency and duration of performance lapsing increase with time-on-task, but also that
there is an accelerated rate of decline in response speed
independent of lapsing.2,3,22 The two processes probably
combine to produce a worsening time-on-task decrement as
subjects become progressively sleepy. The fact that untreated patients with sleep apnea, patients with narcolepsy, and
experimentally sleep-deprived subjects all show clear evidence of time-on-task decrements suggests that relative to
treated patients and alert control, it is the excessive somnolence of these former conditions that produces the vigilance
decrements with time-on-task.
Since maintenance of sustained attention and timely
responses are the cornerstone cognitive requirements of
motor vehicle operation, it is understandable why complaints of sleepiness interfering with the ability to drive,
especially for long distances, are the most common functional deficits found among patients with untreated sleep
apnea. 23 When untreated, such patients not only make more
mistakes of omission on Steer Clear, 17,24 and other lowfidelity driving simulator video tasks,25 but they also
appear to have an increased risk of being involved in automobile accidents. Accident rates for untreated patients with
apnea have been reported to be two to twelve times greater
than those of non-apneic controls, and the risk of a crash
SLEEP, Vol. 22, No. 6, 1999
increased with the severity of sleep apnea. 24,26-29 Although
these studies do not establish whether the patients studied
had a driving exposure comparable to the various control
groups used, they strongly suggest that untreated sleep
apnea can result in an elevated crash risk. Elevated risks for
motor vehicle crashes due to sleepiness and cataplexy also
have been reported for persons with untreated narcolepsy.29-32 The association of disorders of excessive somnolence with escalating time-on-task decrements makes it
imperative that when assessment of neurobehavioral performance is conducted in patients, it involves task durations and analyses that will evaluate the underlying vulnerability of potentially sleepy patients to decrements over
time in tasks that require sustained attention and timely
responses—key components in safe driving performance.
ACKNOWLEDGMENTS
Support for this research was provided in part by NIH
grants HL42236 and NR04281, in part by NASA NCC-2599, and in part by the Institute for Experimental
Psychiatry Research Foundation.
REFERENCES
1. Parasuraman R. Sustained attention in detection and discrimination.
In: Parasuraman R & Davies DR, eds. Varieties of attention. New York:
Academic Press, Inc., 1984;243-71.
2. Dinges DF, Kribbs NB. Performing while sleepy: effects of experimentally-induced sleepiness. In: Monk TH, ed. Sleep, sleepiness and
performance. New York: John Wiley & Sons, 1991;97-128.
3. Kribbs NB, Dinges DF. Vigilance decrement and sleepiness. In:
Harsh J, Ogilvie R, eds. Sleep onset mechanisms. Washington, DC:
American Psychological Association, 1994;133-25.
4. Dinges DF, Powell JW. Microcomputer analyses of performance on a
portable, simple visual RT task during sustained, operations. Behavioral
Research Methods, Instruments, and Computers, 1985;17:652-5.
5. Dinges DF. Probing the limits of functional capability: the effects of
sleep loss on short-duration tasks. In: Broughton RJ, Ogilvie RD, eds.
Sleep, arousal, and performance. Boston: Birkhäuser, 1992;177-88.
6. Wilkinson RT. Lack of sleep and performance. Bulletin of the British
Psychological Society, 1958;34:5A-6A.
7. Wilkinson RT. Sleep deprivation: Performance tests for partial and
selective sleep deprivation. In Abt LE, Riess BF, eds. Progress in clinical psychology: dreams and dreaming (8th ed.). New York: Grune and
Stratton, 1968: 28-43.
8. Williams HL, Lubin A, Goodnow JJ. Impaired performance with
acute sleep loss. Psychological monographs: General and applied,
1959;73:1-26.
9. Williams HL, Kearney OF, Lubin A. Signal uncertainty and sleep
loss. Journal of Experimental Psychology, 1965;69:401-407.
10. Lisper HO, Kjellberg A. Effects of 24-hour sleep deprivation on rate
of decrement in a 10-minute auditory reaction time task. Journal of
Experimental Psychology, 1972;96:287-290.
11. McCarthy ME, Waters WF. Decreased attentional responsivity during sleep deprivation: orienting response latency, amplitude and habituation. Sleep, 1997;20:115-123.
12. Dinges DF. The nature of sleepiness: causes, contexts and consequences. In: Stunkard A, Baum A, eds. Perspectives in behavioral
medicine: eating, sleeping and sex. Hillsdale: Lawrence Erlbaum,
1989;147-179.
13. Bedard M, Montplaisir J, Richer F, Malo J. Sleep disruption and
808
Time-on-Task Decrements—Findley et al.
nocturnal hypoxemia as a determinant of vigilance impairment in sleep
apnea syndrome. Chest, 1991;100:367-371.
14. Findley L, Barth J, Powers D, Wilhoit S, Boyd D, Suratt P.
Cognitive impairment in patients with obstructive sleep apnea and associated hypoxemia. Chest, 1986;90:686-690.
15. Valley V, Broughton R. The physiological (EEG) nature of drowsiness and its relation to performance deficits in narcoleptics.
Electroencephalography and Clinical Neurophysiology, 1983;55:243251.
16. Kribbs N, Pack A, Kline L, et al. Effects of one night without nasal
CPAP treatment on sleep and sleepiness in patients with obstructive
sleep apnea. Am Rev Resp Dis, 1993; 147:1162-1168.
17. Findley L, Fabrizio M, Knight H, Norcross B, LaForte A, Suratt P.
Driving simulator performance in patients with sleep apnea. American
Review of Respiratory Diseases, 1989;140:529-530.
18. Findley L, Unverzagt M, Guchu R, Fabrizio M, Buckner J, Suratt P.
Vigilance and automobile accidents in patients with sleep apnea or narcolepsy. Chest, 1995;108:619-624.
19. Guilleminault C. Clinical features and evaluation of obstructive
sleep apnea. In: Kryger MH, Roth T, Dement WC, eds. Principles and
practice of sleep medicine, Philadelphia, PA: Harcourt Brace
Jovanovich, 1989:552-558.
20. Dember WN, Warm JS. Psychology of perception, (2nd ed). New
York: Holt, Rinehart, Winston, 1979.
21. Kribbs NK, Getsy J, Dinges DF. Investigation and management of
daytime sleepiness in sleep apnea. In: Saunders NA, Sullivan CE, eds.
Sleep and breathing, vol. 2. New York: Marcel Dekker, 1994;575-604.
22. Dinges DF, Powell JW. Sleepiness is more than lapsing. Sleep
Research 1988;17:84.
23. Weaver TE, Laizner A, Evans L, Maislin G, Chugh D, Lyon K,
Smith P, Schwartz A, Redline S, Pack A, Dinges DF. An instrument to
measure the functional status outcomes for daytime sleepiness. Sleep
1997;835-843.
24. Findley L, Unverzagt M, Fabrizio M, Guchu R. Patients with
obstructive sleep apnea have impaired performance on a computer driving simulator. Am Rev Resp Dis, 1990;141:A375.
25. Haraldsson PO, Carenfelt C, Laurell H, Tornros J. Driving vigilance
simulator test. Acta Otolaryngology (Stockholm), 1990;110:136-140.
26. Findley LJ, Unverzagt ME, Suratt PM. Automobile accidents
involving patients with obstructive sleep apnea. Am Rev of Resp Dis,
1988;138:337-340.
27. George CF, Nickerson PW, Hanly PJ, Millar TW, Kryger MH. Sleep
apnea patients have more automobile accidents. Lancet, 1987;447.
28. Haraldsson PO, Carenfelt C, Diderichsen F, Nygren A, Tingvall C.
Clinical symptoms of sleep apnea syndrome and automobile accidents.
ORL J Otorhinolaryngol Relat Spec. 1990;52:57-62.
29. Aldrich MS. Automobile accidents in patients with sleep disorders.
Sleep, 1989;12:487-494.
30. Bartels E, Kusakcioglu, O. Narcolepsy: a possible cause of autombile accidents. Lahey Clinic Foundation Bulletin, 1965;1:21-26.
31. Broughton R, Ghanem Q, Hishikawa Y, Sugita Y, Nevsiomalova S,
Roth B. Life effects of narcolepsy in 180 patients from North America,
Asia, and Europe compared to matched controls. Le Journal Canada des
Sciences Neurologiques, 1981;8:299-304.
32. Broughton R, Guberman A, Roberts J. Comparison of psychosocial
effects of epilepsy and narcolepsy/cataplexy. Epilepsia, 1984;25:423433.
SLEEP, Vol. 22, No. 6, 1999
809
Time-on-Task Decrements—Findley et al.