Abstract
Importance
Resident fatigue has become a point of emphasis in medical education and its effects on otolaryngology residents and their patients require further study.
Objective
The purpose of our study was to evaluate the prevalence and nature of fatigue in otolaryngology residents, evaluate various quality of life measures, and investigate associations of increased fatigue with resident safety.
Study design
Anonymous survey.
Setting
Internet based.
Participants
United States allopathic otolaryngology residents.
Intervention
None.
Main outcome(s) and measures
The survey topics included demographics, residency structure, sleep habits and perceived stress. Responses were correlated with a concurrent Epworth Sleep Scale questionnaire to evaluate effects of fatigue on resident training and quality of life.
Results
190 residents responded to the survey with 178 completing the Epworth Sleep Scale questionnaire. Results revealed a mean Epworth Sleep Scale score of 9.9 ± 5.1 with a median of 10.0 indicating a significant number of otolaryngology residents are excessively sleepy. Statistically significant correlations between Epworth Sleep Scale and sex, region, hours of sleep, and work hours were found. Residents taking in-house call had significantly fewer hours of sleep compared to home call (p = 0.01). Residents on “head and neck” (typically consisting of a large proportion of head and neck oncologic surgery) rotations tended to have higher Epworth Sleep Scale and had significantly fewer hours of sleep (p = .003) and greater work hours (p < .001). Additionally, residents who reported no needle stick type incidents or near motor vehicle accidents had significantly lower mean Epworth Sleep Scale scores. Only 37.6% of respondents approve of the most recent Accreditation Council for Graduate Medical Education work hour restrictions and 14% reported averaging greater than 80 hours of work/week.
Conclusion and relevance
A substantial number of otolaryngology residents are excessively sleepy. Our data suggest that the effects of fatigue play a role in resident well-being and resident safety.
1
Introduction
Sleep deprivation among resident physicians is of growing concern. In 2003 and again in July of 2011, the Accreditation Council for Graduate Medical Education (ACGME) implemented work hour mandates that place limits on resident work hours with the goal of limiting sleep deprivation ( Table 1 ) . These guidelines have prompted an intense debate over their utility and consequences.
Rule | 2003 | 2011 |
---|---|---|
Maximum hours of work per week | 80 hours, averaged over 4 weeks including internal moonlighting | 80 hours per week, averaged over 4 week period including all moonlighting |
Maximum length of duty period | 30 hours with no new patients accepted after 24 hours, 6 additional hours reserved for transitional/educational activities | PGY-1: 16 hours; PGY-2 and up: 28 hours with no new patients accepted after 24 hours, 4 additional hours reserved for transitional/educational activities |
Maximum frequency of in-house night float | N/A | Not more than 6 consecutive nights of night float |
Maximum in-house on-call frequency | Every third night, on average | Every third night, no averaging |
At-home call | Time spent in hospital while on at-home call counts toward 80 hour work week | Every third night, no averaging |
Minimum time off between duty periods | 10 hours | 8 hours required, 10 hours recommended; 14 hours required after 24 hours of in-house duty |
Mandatory time free of duty | 1 day free in 7 days, averaged over 4 weeks including call | 1 day free every week, averaged over 4 weeks without at-home call assigned |
Sleep deprivation is characterized by weariness, tiredness and decreased work capacity, and appears to negatively affect numerous resident-related outcomes . Sleep loss has been associated with decreased ability to learn and think, poor task performance, and a decreased sense of personal well-being . Additionally, deficiencies in sleep have been specifically associated with an increased incidence of accidents such as percutaneous needle sticks or motor vehicle accidents .
The Institute of Medicine’s evaluations and recommendations published in 2009 provide a substantial compilation of evidence for sleep valuation in graduate medical education and played a substantial role in the ACGME duty hour revisions in 2011 . However, to date there has been no otolaryngology specific study utilizing a quantifiable measure of sleepiness such as the Epworth Sleepiness Scale in resident sleep assessment . The purpose of this study was to use a Web-based survey to evaluate the prevalence of sleep deprivation and fatigue specifically in otolaryngology residents and to assess its consequences. Otolaryngology residency is unique in many ways such as in variations of call schedules (in-house versus home call), the post-operative monitoring of microvascular free tissue transfers (aka free-flap checks), and the multitude of different otolaryngology subspecialty rotations. Our survey hoped to elucidate the distinctiveness of the specialty in relation to resident fatigue.
2
Methods
We conducted an anonymous Web-based survey of residents at allopathic otolaryngology residency programs throughout the United States during the spring of 2014. The University of Mississippi University Medical Center Institutional Review Board (IRB) deemed the study exempt from IRB approval.
The survey was developed after consultation with persons familiar with health-related surveys. The survey included 48 questions concerning demographics, residency structure, sleep habits and symptoms, and quality of life. The final survey instrument is given in eAppendix 1. An included Epworth Sleepiness Scale (ESS) questionnaire was used as the primary measure of sleepiness (see eAppendix 1 questions 37–44) . The ESS is a summated scale of eight items. Questions are scored by the respondent on a scale of 0 (no chance of dozing) to 3 (high chance of dozing) for eight hypothetical situations. The total score ranges from 0 to 24. Higher scores indicate more sleepiness. A score ≤ 10 indicates a normal level of sleepiness. Scores of 11–16 indicate excessive sleepiness and scores of > 16 indicate severe sleepiness in previous studies and will serve as the categorization scheme of sleepiness in our study .
The survey was created through SurveyMonkey.com and kept secure with password protection. Residency program directors and coordinators at all 106 ACGME-accredited otolaryngology residency programs within the United States and Puerto Rico were contacted electronically and asked to circulate an invitation to participate to their residents. The list of program directors’ e-mail addresses was obtained using the FRIEDA online database. The invitation to participate included an introduction describing the overall objectives of the study and statements of anonymity of responses (see eAppendix 2). Specific instructions were to log onto our survey using a provided link. At various times reminder e-mails, also containing the link, were e-mailed to the residency directors and coordinators for re-distribution. After six weeks, the survey was closed and the data were extracted for analysis.
Demographics, residency structure, call schedule, and sleep related questions were analyzed and compared with the concurrent ESS questionnaire to identify factors impacting resident fatigue and outcome. Epworth Sleepiness Scale scores are reported as the mean ± standard deviation for categorical predictors. Comparisons were done with the independent samples t -test for dichotomous predictors and the one-way ANOVA with Tukey’s test for polytomous predictors. Tables include sample sizes for categorical variables along with percentage of n in each category. A spearman correlation was also drawn between ESS and conditions that may affect resident performance and fatigue. P-values less than 0.05 were considered statistically significant.
2
Methods
We conducted an anonymous Web-based survey of residents at allopathic otolaryngology residency programs throughout the United States during the spring of 2014. The University of Mississippi University Medical Center Institutional Review Board (IRB) deemed the study exempt from IRB approval.
The survey was developed after consultation with persons familiar with health-related surveys. The survey included 48 questions concerning demographics, residency structure, sleep habits and symptoms, and quality of life. The final survey instrument is given in eAppendix 1. An included Epworth Sleepiness Scale (ESS) questionnaire was used as the primary measure of sleepiness (see eAppendix 1 questions 37–44) . The ESS is a summated scale of eight items. Questions are scored by the respondent on a scale of 0 (no chance of dozing) to 3 (high chance of dozing) for eight hypothetical situations. The total score ranges from 0 to 24. Higher scores indicate more sleepiness. A score ≤ 10 indicates a normal level of sleepiness. Scores of 11–16 indicate excessive sleepiness and scores of > 16 indicate severe sleepiness in previous studies and will serve as the categorization scheme of sleepiness in our study .
The survey was created through SurveyMonkey.com and kept secure with password protection. Residency program directors and coordinators at all 106 ACGME-accredited otolaryngology residency programs within the United States and Puerto Rico were contacted electronically and asked to circulate an invitation to participate to their residents. The list of program directors’ e-mail addresses was obtained using the FRIEDA online database. The invitation to participate included an introduction describing the overall objectives of the study and statements of anonymity of responses (see eAppendix 2). Specific instructions were to log onto our survey using a provided link. At various times reminder e-mails, also containing the link, were e-mailed to the residency directors and coordinators for re-distribution. After six weeks, the survey was closed and the data were extracted for analysis.
Demographics, residency structure, call schedule, and sleep related questions were analyzed and compared with the concurrent ESS questionnaire to identify factors impacting resident fatigue and outcome. Epworth Sleepiness Scale scores are reported as the mean ± standard deviation for categorical predictors. Comparisons were done with the independent samples t -test for dichotomous predictors and the one-way ANOVA with Tukey’s test for polytomous predictors. Tables include sample sizes for categorical variables along with percentage of n in each category. A spearman correlation was also drawn between ESS and conditions that may affect resident performance and fatigue. P-values less than 0.05 were considered statistically significant.
3
Results
We received 190 responses; however, n = 12 (8%) subjects did not respond to any of the eight ESS questions and were eliminated from further analysis. Comparing these participants to those who did complete the ESS questions, there are no significant differences between the groups for age, race, gender or post-graduate year (PGY) so that a complete case analysis was performed with the assumption these twelve were missing at random. Of the remaining n = 178, most (n = 174) completed all eight questions for the ESS. There were n = 4 who answered all but one question. None of the four omitted the same question. To facilitate the analysis, a score of 1.5 was given to the omitted question response in order to estimate a total score. Therefore, the following analysis was based on the n = 178 who gave complete or nearly complete responses to the ESS.
Full demographic results with ESS correlation are listed in Table 2 . Of the groups only gender and a subset of regions were found to have a significant correlation with mean ESS. Females were found to have a significantly larger ESS of 10.80 when compared with males (ESS of 9.20). Additionally, mean ESS did vary by region, being the lowest in the midwest (8.08) and greatest in the northwest (13.07). However, only these two particular regions could be definitively shown to be significantly different. Age and race did not have a significant effect on mean ESS.
N = 178 (%) | ESS Mean | p-value | |
---|---|---|---|
Age | |||
25–28 | 55 (31%) | 10.27 ± 5.15 | 0.0697 |
29–30 | 64 (36%) | 8.76 ± 5.28 | |
> 30 | 59 (33%) | 10.81 ± 4.77 | |
Gender a | |||
Male | 70 (40%) | 9.20 ± 5.14 | 0.0409 |
Female | 106 (60%) | 10.80 ± 4.91 | |
Race a | |||
White | 135 (77%) | 9.87 ± 4.88 | 0.7660 |
Asian | 28 (16%) | 9.77 ± 5.31 | |
Other | 12 (7%) | 8.75 ± 6.73 | |
Year | |||
PGY1 | 30 (17%) | 11.15 ± 5.55 | 0.5078 |
PGY2 | 40 (22%) | 9.18 ± 5.21 | |
PGY3 | 42 (24%) | 9.74 ± 4.40 | |
PGY4 | 44 (25%) | 9.51 ± 5.51 | |
PGY5 | 22 (12%) | 10.64 ± 4.95 | |
Region a | |||
Northeast | 52 (29%) | 9.94 ± 5.03 | 0.0117 |
Southeast | 48 (27%) | 10.60 ± 5.08 | |
Midwest | 48 (27%) | 8.08 ± 4.70 | |
Southwest | 14 (8%) | 10.29 ± 4.98 | |
Northwest | 15 (9%) | 13.07 ± 5.68 |
3.1
Resident fatigue and sleep satisfaction
Overall mean ± s.d. ESS was 9.98 ± 5.16 ( Fig. 1 ). The ESS was normal (≤ 10) in 99/178 (55.6%), indicative of excessive sleepiness (11–16) in 57/178 (32%), and indicative of severe sleepiness (˃16) in 22/178 (12.4%) of residents. Resident reported sleep satisfaction and habits are reported in Table 3 . The majority (57%) of residents reported at least one sleep related symptom, the most common being frequent awakenings throughout the night (20.2%). About half (51.7%) of residents were at least somewhat satisfied with their sleep quantity and even more (65.2%) felt at least this amount of satisfaction toward sleep quality. Despite these seemingly positive self-evaluations of sleep satisfaction, 78.2% reported sleep impacted daily functioning. Most (71.8%) respondents reported never using sleep aids. Of those who reported using sleep aids, 62% reported the use of over the counter medications, 42% alcohol and 30% prescription medications.
Sleep variable | % (N = 178) |
---|---|
Sleep related symptoms | |
Frequent awakenings | 20.2% |
Snoring ≥ 3 nights per week | 12.4% |
Morning headache | 11.8% |
Difficulty falling asleep | 12.4% |
Sleep quantity satisfaction | |
Not satisfied at all | 21.4% |
Slightly satisfied | 27.4% |
Somewhat satisfied | 38.1% |
Very satisfied | 12.9% |
Extremely satisfied | 0.6% |
Sleep quality satisfaction | |
Not satisfied at all | 12.4% |
Slightly satisfied | 22.5% |
Somewhat satisfied | 31.4% |
Very satisfied | 28.1% |
Extremely satisfied | 5.6% |
Sleep effect on daily functioning | |
Very frequently | 6.2% |
Frequently | 24.2% |
Occasionally | 47.8% |
Rarely | 15.2% |
Very rarely | 5.0% |
Never | 1.6% |
Frequency of naps | |
0 per week | 54.5% |
1 per week | 27.5% |
2 per week | 12.9% |
3 or more per week | 5.1% |
Frequency of sleep aid use | |
Very frequently | 3.4% |
Frequently | 2.8% |
Occasionally | 9.0% |
Rarely | 5.1% |
Very rarely | 7.9% |
Never | 71.8% |
3.2
Resident and residency variables
Correlation of ESS with various sleep and residency characteristics are recorded in Table 4 . The majority (83%) of respondents reported no sleep-related comorbidities and comorbidities did not correlate with ESS. Bruxism was the most commonly reported comorbidity at 12.9% while 1.8% reported insomnia. A statistically significant difference in mean ESS was found when analyzing average sleep per night, sleep on call, and work hours per week. Residents who received less than six hours per night had a significantly greater mean ESS (11.93) indicating they are exceptionally more fatigued than residents who slept longer than six hours per night (mean ESS 8.68).
Total = 178, N (%) | ESS, mean ± s.d. | p-value | |
---|---|---|---|
Comorbidities a | |||
None | 142 (83%) | 9.97 ± 5.17 | 0.9953 |
At least one | 29 (17%) | 9.97 ± 5.34 | |
Average sleep on call | |||
0–2 hours | 15 (8%) | 10.93 ± 6.03 | 0.0180 |
2–4 hours | 62 (35%) | 10.24 ± 4.92 | |
4–6 hours | 54 (30%) | 10.18 ± 4.84 | |
6–8 hours | 21 (12%) | 6.38 ± 3.83 | |
N/A | 26 (15%) | 10.79 ± 5.74 | |
Number of incidents | |||
0 | 100 (56%) | 8.18 ± 4.36 | < 0.001 |
1 | 50 (28%) | 11.52 ± 4.96 | |
2 + | 28 (16%) | 13.18 ± 5.58 | |
Frequency of free flap checks a | |||
Every 1–3 hours | 21 (13%) | 8.62 ± 5.33 | 0.3529 |
Every 4 hours | 38 (23%) | 9.17 ± 4.92 | |
Every 5–6 hours | 20 (12%) | 10.63 ± 5.41 | |
Only once | 40 (24%) | 11.04 ± 5.43 | |
None | 46 (28%) | 10.01 ± 4.78 | |
Current avg. sleep a | |||
2–6 hours | 69 (39%) | 11.93 ± 5.56 | < 0.0001 |
6–10 hours | 108 (61%) | 8.68 ± 4.38 | |
Current avg. work hours/week a | |||
0–50 | 15 (8%) | 9.33 ± 4.62 | 0.0002 |
50–60 | 35 (20%) | 6.90 ± 3.67 | |
60–70 | 57 (32%) | 9.94 ± 4.65 | |
70–80 | 46 (26%) | 11.22 ± 5.31 | |
80 + | 24 (14%) | 12.33 ± 6.00 | |
After hours work | |||
0–1 | 29 (16%) | 9.41 ± 5.22 | 0.3069 |
1–2 | 87 (49%) | 9.37 ± 5.01 | |
2–3 | 36 (20%) | 11.08 ± 5.71 | |
3 + | 26 (15%) | 10.62 ± 4.45 | |
Call type a | |||
All home | 121 (68%) | 10.01 ± 5.17 | 0.2890 |
In-house | 39 (22%) | 10.13 ± 4.82 | |
Night float | 11 (6%) | 10.54 ± 5.97 | |
Other | 6 (4%) | 6.00 ± 4.00 | |
Current rotation | |||
Head and neck | 42 (24.1%) | 11.14 ± 5.21 | 0.0925 |
All other rotations | 132 (75.9%) | 9.60 ± 5.10 |