Amin, Mohammad M. MD; Graber, Mark MD; Ahmad, Khalid MD; Manta, Dragos MD; Hossain, Sayeed MD; Belisova, Zuzana MD; Cheney, William PhD; Gold, Morris S. DSc; Gold, Avram R. MD
Despite the current duty hours regulations of the Accreditation Council for Graduate Medical Education (ACGME), which limit duty hours and endorse education regarding sleep loss for residents, these trainees continue to be sleep deprived and fatigued.1 Fatigue caused by insufficient sleep due to long work hours affects interns’ ability to learn as well as their general performance during patient care.2 Previous research has linked long work hours to increased medical errors,3 compromised patient safety,3,4 and increased self-reported injuries.5 Therefore, a strategy is needed to counter the effects of long work hours and fatigue on medical interns.
The 2003 ACGME duty hours standards have been used as the predominant mechanism to reduce fatigue.6 The ACGME limits duty hours and endorses education on sleep deprivation and its consequences. The Sleep, Alertness, and Fatigue Education in Residency (SAFER) program, developed by the American Academy of Sleep Medicine, constitutes an educational effort to decrease sleep deprivation and excessive daytime fatigue among residents.1 The SAFER curriculum covers the factors that put medical residents at risk for sleepiness and fatigue (e.g., circadian rhythm dysfunction and fragmented sleep), the effects of sleep loss on interns’ personal and professional lives (e.g., increase in needle stick injuries),5 the signs of sleepiness and fatigue (e.g., poor concentration and excessive yawning), and adaptive tools and strategies for managing alertness (e.g., napping before work). Although decreasing duty hours and increasing awareness of the dangers of and strategies for dealing with fatigue both seem to be reasonable steps in decreasing fatigue among residents, the ACGME’s duty hours standards and the SAFER program have not led to increased sleep time among medical residents.1
In industries associated with long work shifts, a mid-day nap has been used to resist fatigue. For example, among professional long-haul drivers, a three-hour mid-day nap improved psychomotor performance and alertness during simulated driving.7 Further, research has demonstrated that among medical interns, a nocturnal nap (of approximately three hours) during extended duty hours has increased sleep and decreased fatigue.8 However, to our knowledge, no one has examined the effect of a mid-day nap on the functioning of medical residents. Therefore, we tested the hypothesis that a short, mid-day nap during normal daytime duty hours would prove beneficial for internal medicine (IM) residents by decreasing their sleepiness and improving their cognitive functioning.
We performed a controlled study comparing the effect of a 20-minute (maximum) mid-day nap with that of a 20-minute period of rest (during which participants remained awake) on the alertness and cognitive functioning of first-year IM residents. We chose a duration of 20 minutes for the nap both to prevent the residents from developing sleep inertia that can result after prolonged periods of sleep9 and because research has shown that naps of this duration may improve cognitive functioning and alertness in the laboratory setting.10,11
We conducted the study on the inpatient medical service of the Department of Veterans Affairs Medical Center–Northport (New York). We isolated a room to serve as the study lab. We used blinds to darken the study lab, and we equipped the room with two identical MetroNaps Energy Pods (MetroNaps, New York, New York). The Energy Pod is a comfortable, easy-to-use reclining chair with a built-in timer that is available commercially for the purpose of mid-day napping at the cost of a few thousand dollars (Figure 1).
Participant recruitment and ethical considerations
We invited all of the 37 first-year IM residents to participate in the study. We assigned those who agreed to participate into two groups: a nap group of consecutive residents rotating through the medical service between September 2008 and April 2009, and a control group of consecutive residents rotating through the medical service between July 2009 and April 2010. Participants were working on the medical service during normal daytime duty hours (approximately 8:00 AM to 5:00 PM). We excluded residents working a rotating shift (days alternating with nights), serving as night floats (working only late at night to early morning), or doing a critical care rotation.
No systematic changes occurred in the residents’ work load or duty hours between the first and second years of the study (i.e., between the time when the nap group and the control group rotated through the medical service).
Participation was voluntary and independent of evaluation. We offered no compensation or incentive for participating. The institutional review board of the Department of Veterans Affairs Medical Center–Northport (New York) approved the research protocol used in this study.
Sleep medicine fellows (K.A., D.M., S.H., and Z.B.) executed the protocol; they approached first-year IM residents, explained what the residents’ role in the protocol would be, and elicited their consent.
To characterize the residents’ quality of sleep during the four weeks before the intervention, we administered the Pittsburgh Sleep Quality Index (PSQI) on the day they took their mid-day nap or rest. The PSQI is a validated, 19-item self-report questionnaire that assesses the severity of sleep disturbance on an increasing scale of 0 to 21.12 The PSQI comprises seven subscales (sleep duration, sleep efficiency, sleep latency, daytime sleepiness, sleep efficiency, need for medication, and overall sleep quality), each scored 0 to 3. The PSQI threshold for abnormal sleep quality is a score of 5.12 Sleep quality score is an independent measure that cannot predict the result of a sleep intervention, such as our short, mid-day nap.
Before reporting for their day-shift duty hours, participants arrived at approximately 8:00 AM at the study lab where one of the sleep medicine fellows (K.A., D.M., S.H., or Z.B.) connected them to a portable sleep monitoring device (VIASYS-JAEGER sleep screen monitor, Hochberg, Germany) by two electroencephalographic (EEG) electrodes and two electroculographic (EOG) electrodes. The residents then proceeded with their normal duty hours routines (e.g., rounding on and caring for patients). From approximately 8:30 AM to 1:00 PM, the monitor recorded every occurrence of slow eye movements (SEMs), an indicator of attention failure, during the residents’ wakefulness (see also “Attention failures,” below).13
Participants returned to the study lab at approximately 1:00 PM. There, they took Conner’s Continuous Performance Test Version 5 (CPT II) to evaluate for cognitive functioning.14–16 The CPT II test asks each participant to respond to a letter flashed on a computer screen by either hitting a key or refraining from hitting a key, depending on the letter flashed. The test takes approximately 14 minutes to complete. The CPT II test protocol provides a task-oriented assessment of attention problems and is widely used in research and clinical testing.14–16 CPT II has good age-corrected normative data and good validity.15 It is sensitive and can pick up specific and subtle impairments.16
After the CPT II test, the residents in the nap group were asked to nap in the Energy Pod, and the Energy Pod’s timer awakened each of them after 20 minutes. Participants in the control group were also asked to lie in the Energy Pod, but one of the investigators (S.H. or Z.B.) prevented these residents from falling asleep by conversing with them. All participants repeated the CPT II test after the 20-minute nap or rest period and then returned to their duties at approximately 2:00 PM. During the protocol (from about 1 to about 2 PM), any investigator who was available carried the first-year residents’ pagers, and the supervising second-year residents covered any emergencies.
From about 2:00 PMto approximately 4:30 PM, the sleep monitors again recorded any occurrence of SEMs during the residents’ wakefulness. At 4:30 PM, the study ended and the sleep medicine fellows (K.A., D.M., S.H., and Z.B.) removed the recording equipment.
Data analysis and outcome parameters
Nap data. We determined the nap residents’ wakefulness and sleep during the mid-day nap attempt, using the EEG recording. We divided the entire nap period into 30-second epochs, and we defined a sleeping epoch as at least 15 seconds of non–rapid eye movement (NREM) sleep within the epoch. We also scored every epoch for the depth of sleep using Rechtschaffen and Kales’17 criteria, where NREM stage 1 is shallow sleep, followed by NREM stage 2, then slow wave sleep as the deepest sleep. For each participant, we used the total number of sleeping epochs to define the nap duration, and we used the sleep staging data to determine the depth of sleep.
Attention failures. We defined an attention failure as a 30-second wake period (observed in the EEG) with one or more SEMs. SEMs consist of conjugate, reasonably regular, slow sinusoidal eye movements observed in the EOG recording.18 A board-certified sleep medicine physician masked to the identity of the participant determined the number of attention failures. To determine the prevalence of attention failures, we divided the number of attention failures by the total number of 30-second wake periods and multiplied the quotient by 100 to express the result as a percentage. For each participant, we compared the prevalence of attention failures before and after the 20-minute mid-day intervention. We compared the relative frequency of attention failures between the nap and control groups.
CPT II test parameters. We compared the following parameters of the CPT II test before and after the 20-minute intervention between the two groups:
1. Hit reaction time (HRT): The mean response time (in milliseconds) for all target responses (hitting the key after observing the letter). A long HRT indicates a loss of vigilance.
2. Omission: The number of letters to which the individual should have responded, but did not. A high omission rate, in combination with a slow HRT, indicates inattention.
3. Commission: The number of letters to which the individual should not have responded, but did. A high commission rate, in combination with a slow HRT, indicates inattention.
We compared continuous demographic variables between the nap and the control groups using unpaired t tests. We tested the difference between groups in the male-to-female ratio through Fisher exact test. We summarized measures of alertness using simple summary statistics. We compared the means for each cognition parameter (HRT, omission, commission) between the two groups with analysis of variance, using main-effect-of-treatment and baseline as covariate using SAS software (version 9.2, Carey, North Carolina).
Of the 37 first-year IM residents who rotated through the medical service in 2008–2009 and in 2009–2010, 33 accepted our invitation. We administered the PSQI to these 33 residents and excluded 1 because of poor sleep quality. After we began the protocol, we excluded 3 more residents because they were not willing to continue with EEG/EOG monitoring while attending to patient care. Thus, 29 residents (78% of 37) participated: 18 consecutive IM residents in the nap group who rotated through the medical service between September 2008 and April 2009, and a control group of 11 consecutive residents who rotated through the medical service between July 2009 and April 2010.
Demographics and sleep
Table 1, demonstrating the anthropometric data of both sets of residents, shows that the two groups did not differ significantly by gender, age, or body mass index (these variables affect prevalence of sleep disorders). Table 2 demonstrates that both groups had mildly abnormal sleep quality, including some daytime sleepiness, as determined by the PSQI.
Participants assigned to the nap group slept for 8.4 ± 3.0 minutes (in NREM) of the 20 minutes allotted (range: 3–14 minutes). On average, these participants had 7.4 ± 2.4 minutes of NREM stage 1 sleep and 1.0 ± 1.0 minutes of NREM stage 2 sleep.
We detected no significant correlations between either the baseline PSQI or sleep time during naps and any of the outcome parameters (below).
The nap group demonstrated a decrease in the prevalence of attention failures from 15.4% ± 2.3% before napping to 10.8% ± 3.7% after napping compared with the control group residents, who demonstrated an increase in the prevalence of attention failures from 14.9% ± 2.6% before resting to 16.6% ± 3.9% after resting (P = .002; Figure 2).
The nap residents also demonstrated improvement after napping in the outcome parameters of the CPT II test. The nap group demonstrated a decrease in the HRT from 341.3 ± 37.2 milliseconds before napping to 313.9 ± 30.1 milliseconds after napping compared with the controls, whose HRT remained relatively unchanged, at 361.5 ± 50.5 milliseconds before resting and 360.9 ± 52.5 milliseconds after resting (P = .004; Figure 2). The nap group demonstrated a decrease in omissions from 2.4 ± 2.1 before napping to 1.2 ± 1.7 after napping compared with the control group residents, whose omissions remained relatively unchanged at 1.3 ± 1.6 before resting and 1.3 ± 1.4 after resting (P = .01; Figure 2). The nap group demonstrated a decrease in commissions from 17.0 ± 6.7 before napping to 13.2 ± 7.0 after napping compared with the controls, whose commissions increased from 16.6 ± 6.8 before resting to 17.4 ± 5.5 after resting (P = .007; Figure 2).
Discussion and Conclusions
In this study, we measured the effect of a brief, mid-day nap during normal duty hours on cognitive functioning and alertness among first-year IM residents. We found that, compared with the resting-but-awake residents, the residents who actually napped experienced fewer attention failures during their work later in the day as determined by a monitor of SEMs. Further, we found that, compared with controls who rested but stayed awake for 20 minutes, residents who had the opportunity to nap for a maximum of 20 minutes demonstrated a faster reaction time and made fewer errors of omission and commission as determined by a validated test of cognitive functioning. These findings suggest that a short, mid-day nap may improve first-year residents’ performance during their clinical duties.
Preventing fatigue among medical residents has long been viewed as a method of preventing medical errors, enhancing patient safety, and improving medical education.19 The ACGME implemented duty hours standards and encouraged the SAFER program, respectively, to provide medical residents with more time for sleep before and after long shifts associated with sleep loss and to educate residents on both preventive and recuperative steps they might take to mitigate the effects of chronic sleep loss.1 In 2007, Arora and associates1 published their study evaluating whether medical residents at the University of Chicago obtained enough preventive sleep before long shifts and enough recuperative sleep after those shifts. In addition, they evaluated the effect of a one-time SAFER program presented to the residents. Using actigraphy (i.e., a noninvasive method to measure periods of activity and rest in humans by recording movements and using periods of rest as a surrogate for sleep) to objectively record the residents’ sleep and wake time, they found that the residents did not obtain adequate preventive sleep (a minimum of 7 hours) the night before their long shift or enough recuperative sleep during their two nights after the shift (a minimum of 16 hours).1 Further, by comparing residents’ actigraphy before and after the SAFER program, they found no change in the residents’ sleep behavior.1 The findings of Arora and colleagues’ study suggest that the ACGME duty hours standards and the SAFER program are not adequate to prevent fatigue among medical residents and that alternative approaches, such as a short, mid-day nap, may be needed.
Our hypothesis—that a short, mid-day nap improves cognitive functioning and alertness among medical residents—does not seem to emanate naturally from the previous studies of long-haul drivers7 and medical residents8 working extended duty hours. In those studies, drivers and residents took long naps of approximately three hours. Imagining that a three-hour nap can be restorative and can improve cognition, psychomotor performance, and alertness is not difficult. Nevertheless, our study’s findings are consistent with previous nap studies done in a sleep laboratory setting. In two separate studies, a 10-minute nap (in a sleep laboratory bed in a darkened room) after a night of mild sleep restriction11 and a 15-minute nap (under similar conditions) after a night of unrestricted sleep10 both resulted in improved cognitive functioning and improved alertness that persisted for approximately three hours. Our hypothesis follows naturally from these two previous sleep laboratory studies examining short naps, and our results extend their findings to a work setting.
Although our findings show that a short period of light sleep may improve cognitive functioning and alertness in IM residents, they do not explain how or why that result may occur. The existing two-process model of sleep regulation20 postulates two contributing processes: (1) a circadian process, through which sleepiness increases and decreases depending on the time of day (Process C), and (2) a homeostatic process, through which sleepiness increases with lengthening wakefulness (Process S). If we postulate that napping decreases sleepiness through its effect on Process S, then it is difficult to imagine a very brief period of light sleep decreasing sleepiness and increasing alertness very much. Likewise, if we postulate that napping interrupts Process C, then a very brief nap may not decrease sleepiness.
To explain the restorative effects of a short, light nap, Lovato and Lack9 have postulated a third process related to the onset of sleep (Process O). According to their hypothesis, decreased alertness and cognitive functioning may result from an adaptation occurring in the wake on / sleep on switch in the hypothalamus.21 During a few hours of wakefulness, the neurons of the wake on switch gradually become less excitable, resulting in decreased alertness. With sleep onset, the excitability rapidly returns to these neurons so that after a nap lasting even a few minutes, much of the excitability has been restored to the wake on neurons, which results in improved cognition and alertness that can last a few hours as the neurons readapt. However, Lovato and Lack’s9 explanation is, as yet, unproved; thus, the existing paradigm of sleep regulation is not adequate to explain the restorative effects of a short, mid-day nap. Further modeling is needed.
Considerations and limitations
To apply our findings to medical residency programs, several aspects of our study must be considered. First, our participants’ mildly elevated PSQI scores, which reflected increased sleepiness (similar to that of other first-year medical residents1), exemplify the sleep deprivation that continues despite the ACGME’s regulations.
In designing our study, however, we did not control for at least two confounding factors: caffeine intake and workload. Also, we performed our study across two different, but sequential years—the first year with our nap group, and the following year with our control group—rather than simultaneously. Although the Department of Medicine at the Department of Veterans Affairs Medical Center did not change its policies regarding the first-year residents’ work hours or workload during the time of the study, we cannot be certain that the patient load was the same during each academic year. Furthermore, we studied the effect of sleep on cognitive function and alertness (the nap group), comparing it against rest without sleep (the control group that rested in the Energy Pod). While a rest group is one control, we suspect that most residents do not rest during the day, but continue working. Even resting could be better than continuous work. To know the frequency of attentional failures and cognitive performance in a group of residents who continued to work without either rest or sleep would be interesting. Finally, although we demonstrated some benefit to cognitive functioning and alertness resulting from a short, mid-day nap, we do not know what effect, if any, such a benefit would have on patient safety or the frequency of medical errors. We limited our protocol to one day, and we did not collect patient outcomes or medical error data.
Despite its limitations, our study demonstrates the potential benefit of a short, mid-day nap for residents and highlights the need for further research to address the effect of a mid-day nap on the quality of patient care, resident education, and longer work shifts.
Acknowledgments: The authors acknowledge the technical assistance of Ms. Cori Abbondanza, a registered polysomnographic technologist.
Funding/Support: The authors declare that they have no competing interests. Dr. Morris Gold’s participation in this study was entirely voluntary. Novartis Consumer Health provided no funding or other resources for this study. MetroNaps (New York, New York) loaned two MetroNaps Energy Pods to the Department of Veterans Affairs Medical Center for the two-year research period.
Ethical approval: The institutional review board of the Department of Veterans Affairs Medical Center–Northport (New York) approved the research protocol used in this study.
Disclaimer: The views expressed in this report are those of the authors alone and do not reflect those of the U.S. Department of Veterans Affairs.
Previous presentations: The results of this study were presented at the Sleep Health and Safety meeting of the National Sleep Foundation, March 2011,Washington, DC.
1. Arora VM, Georgitis E, Woodruff JN, Humphrey HJ, Meltzer D. Improving sleep hygiene of medical interns: Can the sleep, alertness, and fatigue education in residency program help? Arch Intern Med. 2007;167:1738–1744
2. Feddock CA, Hoellein AR, Wilson JF, Caudill TS, Griffith CH. Do pressure and fatigue influence resident job performance? Med Teach. 2007;29:495–497
3. Barger LK, Ayas NT, Cade BE, et al. Impact of extended-duration shifts on medical errors, adverse events, and attentional failures. PLoS Med. 2006;3:e487
4. Parshuram CS. The impact of fatigue on patient safety. Pediatr Clin North Am. 2006;53:1135–1153
5. Ayas NT, Barger LK, Cade BE, et al. Extended work duration and the risk of self-reported percutaneous injuries in interns. JAMA. 2006;296:1055–1062
6. Philibert I, Friedmann P, Williams WTACGME Work Group on Resident Duty Hours. Accreditation Council for Graduate Medical Education. . New requirements for resident duty hours. JAMA. 2002;288:1112–1114
7. Macchi MM, Boulos Z, Ranney T, Simmons L, Campbell SS. Effects of an afternoon nap on nighttime alertness and performance in long-haul drivers. Accid Anal Prev. 2002;34:825–834
8. Arora V, Dunphy C, Chang VY, Ahmad F, Humphrey HJ, Meltzer D. The effects of on-duty napping on intern sleep time and fatigue. Ann Intern Med. 2006;144:792–798
9. Lovato N, Lack L. The effects of napping on cognitive functioning. Prog Brain Res. 2010;185:155–166
10. Takahashi M, Fukuda H, Arito H. Brief naps during post-lunch rest: Effects on alertness, performance, and autonomic balance. Eur J Appl Physiol Occup Physiol. 1998;78:93–98
11. Tietzel AJ, Lack LC. The recuperative value of brief and ultra-brief naps on alertness and cognitive performance. J Sleep Res. 2002;11:213–218
12. Buysse DJ, Reynolds CF 3rd, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: A new instrument for psychiatric practice and research. Psychiatry Res. 1989;28:193–213
13. Lockley SW, Cronin JW, Evans EE, et al.Harvard Work Hours, Health and Safety Group. Effect of reducing interns’ weekly work hours on sleep and attentional failures. N Engl J Med. 2004;351:1829–1837
14. Conners CK. The computerized continuous performance test. Psychopharmacol Bull. 1985;21:891–892
15. Homack S, Riccio CA. Conners’ Continuous Performance Test (2nd ed.; CCPT-II). J Atten Disord. 2006;9:556–558
16. Egeland J, Kovalik-Gran I. Validity of the factor structure of Conners’ CPT. J Atten Disord. 2010;13:347–357
17. Rechtschaffen A, Kales A A Manual of Standardized Terminology, Techniques and Scoring System for Sleep Stages of Human Subjects. 1968 Bethesda, Md US Department of Health, Education and Welfare
18. Silber MH, Ancoli-Israel S, Bonnet MH, et al. The visual scoring of sleep in adults. J Clin Sleep Med. 2007;3:121–131
19. Iglehart JK. Revisiting duty-hour limits—IOM recommendations for patient safety and resident education. N Engl J Med. 2008;359:2633–2635
20. Achermann P. The two-process model of sleep regulation revisited. Aviat Space Environ Med. 2004;75(3 suppl):A37–A43
21. Saper CB, Chou TC, Scammell TE. The sleep switch: Hypothalamic control of sleep and wakefulness. Trends Neurosci. 2001;24:726–731