Firefighter sleep: a pilot study of the agreement between actigraphy and self-reported sleep measures

Abstract

Study Objectives:

The aim of this study is to determine the extent of agreement between self-reported measurements of total sleep time (TST) and actigraphy in the fire and emergency services occupation.

Methods:

Twenty-four firefighters participated in an 18-day study. Four measurements were used to assess TST: Pittsburgh Sleep Quality Index, a newly developed habitual Extended Sleep Survey, a newly developed daily Emergency Services Sleep Diary (ESSD), and actigraphy. The Extended Sleep Survey and ESSD were constructed to address the specific job-related characteristics of fire and emergency services that other measurements cannot achieve (eg, multiple sleep episode in a single night).

Results:

The Pittsburgh Sleep Quality Index TST is least accurate compared to actigraphy. The Extended Sleep Survey TST shows improvement over Pittsburgh Sleep Quality Index TST, but was statistically different from actigraphy TST. No difference in mean TST was found between ESSD TST and actigraphy TST. Furthermore, ESSD TST and actigraphy TST correlated strongly together.

Conclusions:

Without modification, traditional self-reported measures may not be appropriate in the fire and emergency service occupation. This study suggests that the ESSD may serve as a useful alternative to actigraphy to measure TST.

Citation:

Billings JM. Firefighter sleep: a pilot study of the agreement between actigraphy and self-reported sleep measures. J Clin Sleep Med. 2022;18(1):109–117.

BRIEF SUMMARY

Current Knowledge/Study Rationale: To understand sleep in the fire and emergency services, we first must determine whether current sleep assessments can account for job characteristics that influence sleep such as multiple sleep episodes caused by emergency calls, irregular sleeping schedules, and different sleep environments at home and work.

Study Impact: Compared to actigraphy total sleep time, Pittsburgh Sleep Quality Index total sleep time is least accurate among the measures. A newly developed sleep diary designed to overcome previous limitations shows strong agreement with actigraphy. The impact of these results suggests improved accuracy in estimating total sleep time when using a sleep instrument designed specifically for the fire and emergency services occupation compared to traditional methods.

INTRODUCTION

The purpose of the research reported herein was to compare different measurements of total sleep time (TST) in the fire and emergency services occupation. Sleep duration is an important parameter because it is associated often with health and performance outcomes.¹ Sleep duration can be measured either by self-reported questionnaires or objectively through polysomnography and, less invasively, actigraphy. In addition, sleep can be measured habitually or daily. Daily assessments provide the opportunity to analyze sleep duration and other parameters throughout a multiday period, which allows a better understanding of circadian rhythm.² Habitual measurements, however, focus on overall sleep throughout a period. With growing consensus that sleep is an established pillar of health,³ it is important to determine whether current sleep instruments can accurately assess sleep among those serving in the fire and emergency services occupation.

Sleep in this occupation can be complex because firefighters 1) have irregular working schedules,⁴ 2) have irregular sleeping times,^5–8 3) may experience multiple sleep interruptions for emergency calls,⁹ 4) may experience multiple sleep bouts per night,¹⁰ 5) sleep in differing environments at home and work,¹¹ 6) may wake from alarms intended for different units (eg, multiple units in a station),¹² and 7) may nap and rest during downtimes throughout the workday.^11,13 While it is not surprising that firefighters suffer from poor sleep, these work characteristics can make assessing sleep using traditional measures much more challenging. For example, Figure 1 illustrates examples of actigraphy from firefighters (Figure 1A illustrates a single, consolidated sleep episode for comparison). Figure 1B illustrates 2 interruptions due to emergency calls. Figure 1C illustrates that, following the returning to the station after an emergency call, the firefighter attempted to sleep but could not. Figure 1D demonstrates an attempt to bank sleep, but the firefighter experienced interruptions along with an emergency call around midnight. After returning to the station, that firefighter experienced severe tossing and turning while sleeping. Traditional measures would not capture these events; therefore, sleep assessments relying on such measures would likely be inaccurate. While previous research has attempted to account for these considerations,¹⁰ to the author’s knowledge, no study has compared the agreement between different types of self-reported measurements to actigraphy in the fire and emergency services.

Figure 1. Actigraphy from firefighter participants.

Blue represents activity, yellow represents light, green and pink represent sleep. (A) a single, consolidated sleep episode for comparison; (B) 2 interruptions due to emergency calls; (C) following the returning to the station after an emergency call, the firefighter attempted to sleep but could not; (D) an attempt to bank sleep, but the firefighter experienced an emergency call around midnight followed by several interruptions.

Measurements of sleep

Pittsburgh Sleep Quality Index

Among the several sleep instruments available,¹⁴ the most frequently used in the fire and emergency services is the Pittsburgh Sleep Quality Index¹⁵ (PSQI) (Table 1). The PSQI contains many frequency-type questions that can be difficult to interpret for those who experience multiple sleep episodes a night without clarifying whether to judge the questions and responses per day/night or per sleep episode. In addition, firefighter responses depend on whether to judge sleeping habits at work, home, or both without additional clarification. For example, when asked retrospectively about average sleep duration over the past 30 days, accurate recollection is likely difficult for those who do not maintain a consistent sleep schedule and intact memory. Girschik and colleagues⁵⁵ suggest that respondents typically provide modal sleep time estimates rather than mean sleep time. It may be nearly impossible to recall an average sleep duration by a firefighter with an irregular work schedule, multiple sleep bouts each night, and different sleep latencies for each bout over a 30-day period. Moreover, when field testing a questionnaire for previous work,¹⁰ several firefighter participants asked if they should respond while thinking about sleep at work, home, or both. This feedback led to modification of the questionnaire to account for both home and work sleep so that an overall average can be calculated based on the shift schedule.

Table 1.

Frequently used measurements in the fire and emergency services sleep research.

Measurement	Studies (References)
PSQI15	7, 10, 16–43
ESS44	11, 17, 23, 25, 26, 28, 34, 45–47
ISI48	23, 25, 26, 28, 31, 46, 49
AIS50	23, 27, 51, 52
SSS53	28, 54

AIS = Athens Insomnia Scale, ESS = Epworth Sleepiness Scale, ISI = Insomnia Severity Index, PSG = Polysomnography, PSQI = Pittsburgh Sleep Quality Index, SSS = Stanford Sleepiness Scale.

While traditional measures may be feasible for those who experience 1 consolidated sleep episode per night in the same sleeping environment, these measures are not conducive to measure sleep from those who experience fragmented sleep, irregular sleep schedules, and differing sleep environments. Studies that compared the accuracy between TST as measured by PSQI and that measured by actigraphy find substantial disagreement;^1,56 however, no study has confirmed these results in the fire and emergency occupation. This leads to the first hypothesis, H1: Firefighter TST as measured by the PSQI is less accurate than TST measured by actigraphy.

Extended Sleep Survey

To address the characteristics of the fire and emergency services, a new retrospective habitual measure was created for this study. The Extended Sleep Survey modifies the PSQI by adding questions to account for the concerns mentioned above (eg, irregular work schedule, sleeping at home and work, and multiple sleep bouts) so that respondents may better estimate TST compared to PSQI. This leads to a second hypothesis, H2: TST measured by the Extended Sleep Survey is more accurate than TST measured by PSQI TST compared to actigraphy TST.

The Core Consensus Sleep Diary (CSD),² developed for those who suffer from insomnia, also was used to guide Extended Sleep Survey development. The CSD asks the following: “(1) the time of getting into bed, (2) the time at which the individual attempted to fall asleep, (3) sleep onset latency, (4) number of awakenings, (5) duration of awakenings, (6) time of final awakening, (7) final rise time, (8) perceived sleep quality (rated via Likert scale), and (9) open-ended comments from the respondents” (p. 290).² Contrary to the PSQI, the Extended Sleep Survey captures sleep habits at work and home separately as shown in the following:

Work

1. What is your typical in-bed time at work?

2. What is your typical out-of-bed time at work?

3. How many emergency calls do you typically experience a night at work?

4. What is the average duration of these calls you experience at work?

5. If you experience any other noncall interruptions during a typical night, what is the total duration on average?

6. How long does it take you to fall asleep the first time?

7. After a call, how long does it take you to fall back sleep?

   Home

8. What is your typical in-bed time at home?

9. What is your typical out-of-bed time at home?

10. How many interruptions do you typically experience a night at home?

11. What is the average duration of these interruptions you experience at home?

12. How long does it take you to fall asleep the first time?

13. After an interruption, how long does it take you to fall back sleep?

It is possible that even with enhanced questions, the Extended Sleep Survey may nevertheless not agree with actigraphy due to the occupationally induced characteristics that promote irregular sleeping patterns. This leads to a third hypothesis, H3: Firefighter TST as measured by the Extended Sleep Survey is less accurate than TST measured by actigraphy.

Emergency Services Sleep Diary

A sleep diary is considered the gold standard for self-reported measurement of sleep. A diary offers the advantage of reporting daily sleep parameters throughout a specified period. Though the CSD is limited by self-reports as in other self-report instruments, it is nevertheless seen as the most accurate among them.² The daily CSD offers greater accuracy over habitual measures since the respondent need only remember sleep from the previous night. While this is sufficient for those who experience 1 sleep episode each night, firefighter sleep is often fragmented due to occupational demands and related stressors (ie, interruptions due to a call, waking early to start/end work, anticipating emergency calls, experiencing stress/anxiety from a previous call, distractions from coworkers, etc). Firefighters may have multiple sleep episodes in a single night while at work, and each sleep episode should be treated individually. The reason for this is that the duration of emergency call interruptions at night should be eliminated from sleep data. Otherwise, this duration would be treated as wake after sleep onset (WASO), and thus inappropriately lower sleep efficiency. This would mischaracterize the sleep pattern and could lead to the possible misdiagnosis of a sleep disorder. Therefore, the Emergency Services Sleep Diary (ESSD) was developed for this study to address this limitation.

The CSD served as a guide to develop the ESSD, along with the author’s previous experience and firefighter feedback during field testing. The author predicts that the ESSD will be an accurate TST measurement and in fact, can compete with actigraphy in its accuracy. This leads to a fourth hypothesis, H4: No significant difference exists between TST as measured by the ESSD and TST as measured by actigraphy.

The ESSD contains questions relating to daily activities and sleep. Collecting data on individual sleep episodes and interruptions provides valuable insight into how characteristics of each sleep episode and interruption (eg, type of Fire/EMS calls, duration of interruption, and time of interruption) might impact subsequent sleep and performance. This would be useful in future studies investigating sleep patterns in the fire and emergency services. For the current study, the ESSD contains the following items to help respondents better estimate TST, as measured by last question, “How many hours do you feel you slept last night?”

1. How many calls did you respond to during the night (after 10 pm)? (with an option to indicate nonwork night)

2. A table to report, “Time Laid in Bed” (defined as attempting to fall asleep) and “Time Out of Bed” (defined as awake time) for each sleep episode.

3. How many sleep interruptions did you experience last night? (nonemergency calls)

4. How many hours do you feel you slept last night?

Actigraphy

Polysomnography (PSG) is the gold standard for assessing sleep; however, its use in the fire and emergency services presents challenges. PSG is less feasible due to the cost of equipment and technicians. In addition, PSG has limited practicality for firefighters wearing leads and equipment while performing their work responsibilities (eg, responding to emergencies).

Actigraphy is less costly and less intrusive compared to PSG. In studies comparing PSG to actigraphy, researchers found no statistical difference in TST measurement.^57,58 In addition, actigraphy records each sleep episode and interruption, thereby easily addressing the concerns above.

METHODS

Participants

The protocol for human subjects was approved by the Oklahoma State University Institutional Review Board. Informed consent was obtained from each participant prior to beginning data collection. Data were obtained from a longitudinal investigation of sleep in the fire and emergency services. Each participant completed 165 surveys over 54 days spread out in 3, 18-day rounds over a 10-month period. A fire department in the South Central United States classified by the United States Fire Administration as a “Career” department was recruited for this study based on its use of the 24 hours on and 48 hours off (24/48) shift schedule (Table 2), the most commonly used work schedule in the US fire service.⁴ The use of a well-defined shift schedule ensured that firefighter participants maintain a consistent work/home pattern that required them to sleep at the station when on shift (in comparison to volunteer fire departments staffed by firefighters who often do not adhere to fixed schedules). Finally, firefighters with at least 2 months of experience under the current shift schedule (so that their circadian rhythm was adjusted to the work schedule) qualified for participation in this study.

Table 2.

24/48 Shift schedule.

S	M	T	W	T	F	S
On	Off	Off	On	Off	Off	On
Off	Off	On	Off	Off	On	Off

Illustrates an example of 1 shift at a fire department operating on a 24 on/48 off shift schedule. This schedule has 3 shifts (or crews) that stager work days to provide 24/7 coverage. The tour repeats continuously according to a fire department’s designated work period.

Out of 56 qualified firefighters, 37 consented to participate (66% participation rate). At the conclusion of the larger study, 24 participants had submitted a complete set of relevant data (65% cooperation rate), which is representative of a national firefighter study conducted by Barger and colleagues.⁵¹ Thirteen participants were excluded from data analysis due to inconsistent data with actigraphy; failing to complete questionnaires; participation withdrawal, retirement, resignation, termination; or changed roles within the department during the study period.

Variables

PSQI TST is measured by question 4: “During the past month, how many hours of actual sleep did you get at night? (This may be different than the number of hours you spent in bed”, p. 209).¹⁵ PSQI TST is treated as an interval variable.

Extended Sleep Survey TST was calculated by the investigator from participant responses to each question. For example, if the response to question 1 was 10:00 pm and question 2 was 6:00 am, the initial TST for sleep would be 8 hours. If the firefighter averaged 2 calls (question 3) and each call lasted approximately 30 minutes on average (question 4), total interruption duration is 1 hour, reducing TST to 7 hours. If the firefighter reported an average initial sleep latency of 25 minutes (question 6) and used the bathroom at night for 5 minutes on average (question 5), 30 minutes are deducted from the TST to 6.5 hours. Lastly, if the firefighter reported 45 minutes sleep latency after returning from calls on average (question 7) and experienced 2 calls on average (question 3), TST is reduced further by 1.5 hours for an overall TST at work of 5 hours.

This process is repeated for TST at home. The Extended Sleep Survey TST is calculated based on the 1:2 (nights at work:nights at home) ratio (Table 2). Extended Sleep Survey TST is treated as an interval variable.

ESSD TST is measured by question 4: “How many hours do you feel you slept last night?” ESSD TST is treated as an interval variable.

ActiGraph wGT3X-BT (ActiGraph Corporation, Pensacola, FL) devices were purchased by the author for this study. ActiLife software⁵⁹ (Cole-Kripke algorithm) was used to score sleep parameters. TST, as scored by the Cole-Kripke algorithm, has been demonstrated to have an accuracy similar to PSG in healthy participants.⁶⁰

Data Collection Protocol

Participants offered informed consent and completed the PSQI, Extended Sleep Survey, and demographic questionnaire on the first day of the study. After completing the questionnaires, participants were fitted with a wGT3X-BT device and instructed to complete the ESSD morning section upon awaking each morning.

Participants wore the wGT3X-BT and completed the ESSD for 18 days. This allowed for 6 tours (24/48 consists of 3 days in the tour; see Table 2) of data collection. In the larger study, adherence to the shift schedule was essential since incomplete tours produced incomplete pictures of firefighter sleep. For example, if a firefighter working the 24/48 took a vacation over 1 work shift, they would have 5 consecutive days off, which would influence the results. A duration of 6 tours (18 days) was selected to ensure ample data for analysis should tour data need be eliminated due to vacation leave, sick leave, swapping shifts, or working overtime, which is common in the fire and emergency services.

The PSQI and ESSD examines habitual sleep over the past 30 days. Ideally, actigraphy and ESSD should also be 30 days and during the same period. Some error associated with nonsimultaneous measurements of TST is possible, as noted in a study that used a 3-day weighted average to represent 30 days.⁶¹ However, Girschik and colleagues⁵⁵ noted that participants are likely to report modal sleep with assessments that have a long recall period. Considering that firefighters likely have irregular schedules due to sleeping at work compared to home and having multiple sleep bouts of variable duration, it is unrealistic to assume sleep means can be accurately calculated over 30 days, which makes it more likely for participants to report modal sleep. Moreover, the tour of 24 hours on and 48 hours off repeats continuously; it is this tour pattern that structures their lives. Unless a participant took an extended leave, differences in reporting should be minimal. Nonetheless, all participants were screened on the first day for significant past work shift deviation.

Actigraphy Data Cleaning

Several steps were necessary to score actigraphy data. All actigraphy data were first validated using Choi (2011) wear time validation in ActiLife. Wear time was cross-referenced with the ESSD, in which participants were to indicate any removal of the device. Any inconsistencies were resolved by reviewing data with the participant. Next, each in-bed time and out-of-bed time per night were manually entered in ActiLife using Question 2 in the ESSD; a table (actigraphy log) was used for the sole purpose of manually indicating sleep times as recommended in the “SBSM Guide to Actigraphy Monitoring.”⁶³ Any inconsistencies (eg, if actigraphy illustrates extensive activity during in-bed and out-of-bed time) were resolved by reviewing data with the participant. Sleep data were exported to .csv files in which individual sleep episodes were combined to produce a single set of sleep parameters each night for each participant.

Statistical Analysis

Analysis of sleep data was conducted using STATA.⁶⁴ For each hypothesis, a paired t-test was used to analyze the difference of means at the .05 level. To test H1 and H3, a paired t-test was performed using actigraphy TST aggregated by participants. H4 was also analyzed by paired t-test in addition to Pearson product-moment correlation.

RESULTS

Participant characteristics are summarized in Table 3. Firefighter respondents overestimated sleep in all 3 self-reported measurements compared to actigraphy. Using actigraphy, firefighters slept an average 6.38 hours each night. The mean PSQI TST is 7.52 hours. The mean Extended Sleep Survey TST is 6.97 hours. The mean ESSD TST is 6.43 hours (Table 4).

Table 3.

Characteristics of firefighter respondents.

Sex male, n	24
Mean age, years (range)	34 (20–52)
Marital status, n (%)	19 (79%) married
	5 (21%) not married
Children, n (%)	16 (67%) with
	8 (33%) without
Second job, n (%)	10 (42%) with
	14 (58%) without
Mean years of service (range)	10.5 (3 months to 29 years)
Alcohol use, n (%)	18 (75%) consume
	6 (25%) do not consume
Tobacco, n (%)	7 (29%) use
	17 (71%) do not use
Caffeine, n (%)	23 (96%) consume
	1 (4%) do not consume
Mean body mass index	28.4 kg/m2
Mean wake to work latency	96.5 minutes
Mean station travel duration	28 minutes
Station rotation, n (%)	4 (17%) rotate
	20 (83%) do not rotate
Individual night call volume	1.0 call
Individual day call volume	2.6 calls

Table 4.

Measurements and total sleep time.

Measurement	Mean Hours	Standard Deviation	Min	Max
PSQI TST	7.52	1.16	4.5	9.3
Extended Sleep Survey TST	6.97	0.86	5	8.2
ESSD TST	6.43	0.70	5.3	7.8
Actigraphy TST	6.38	0.78	4.9	7.5

Actigraphy and ESSD data aggregated by participant. ESSD = Emergency Services Sleep Diary, PSQI = Pittsburgh Sleep Quality Index, TST = total sleep time.

Each of the three self-reported measurements were compared to actigraphy. The differences are illustrated in the Figure 2 box plot. In addition, paired t-tests were performed to determine whether a statistically significant difference in means exists between measurements. The PSQI TST had a statistically significant difference from actigraphy (H1).

Figure 2. Agreement with actigraphy.

ESSD TST = Emergency Services Sleep Diary, PSQI TST = Pittsburgh Sleep Quality Index Sleep Duration component, TST = total sleep time.

The results of paired t-test for H2 indicates a statistically significant difference of means between the Extended Sleep Survey TST and PSQI TST of 0.55 hours (33 minutes) (standard deviation: 0.60) [t(24) = .55, P < .001]. This suggests that the Extended Sleep Survey TST is more accurate than the PSQI TST compared to actigraphy TST. Nevertheless, the results of paired t-tests suggest that the Extended Sleep Survey TST was statistically different from actigraphy TST (H3).

One participant overestimated their sleep by 2.71 hours (illustrated as the outlier on the boxplot). For this person, they recorded sleeping approximately 7.58 hours in the Extended Sleep Survey, but actigraphy shows average sleep of 4.87 hours. Their response to the ESSD was 5.5 hours and the PSQI TST was 8.5 hours, which was also the largest difference (3.63 hours) for that measurement among the participants. It is unknown why this participant had such inaccurate TST estimates.

The results of the paired t-test for H4 failed to reject the null hypothesis, suggesting no difference in means exists between TST as measured by ESSD and that measured by actigraphy. The mean difference is 0.05 hours (3 minutes) (standard deviation: 0.72). A Pearson product-moment correlation coefficient was performed to assess the relationship between ESSD TST and actigraphy TST with unaggregated data. There was a strong correlation between the 2 [r = .753, n = 304, P < .001] (Figure 3).

Figure 3. Total sleep time between Emergency Services Sleep Diary and actigraphy.

Unaggregated participant sleep data. TST = total sleep time.

DISCUSSION

On the basis of this pilot study of 24 firefighters from 1 fire department in the South Central United States, the ESSD proved to be a favorable, new method of estimating firefighter TST when compared to actigraphy. The results demonstrate a strong, significant correlation between the ESSD-estimated TST and actigraphy TST. The results also demonstrate no significant difference in means exists between the self-reported ESSD TST measurement and actigraphy TST. Furthermore, no notable difference was found between estimates of ESSD TST at home compared to work despite possible influence from adverse effects of sleep loss (eg, sleep inertia, decreased performance, poor memory recall^65–67).

Interestingly, but not surprisingly, it appears that participants typically rounded ESSD TST estimates in increments of 30 minutes (Figure 3, eg, hours 4, 5, 5.5, 6, 6.5, etc). Further research is needed to explore possible influences or covariates in estimating TST in the fire and emergency services. Previous research finds several covariates that can influence estimation of TST,¹ whether this can be generalized to firefighters has not been confirmed. It is clear that more work is needed to improve reporting accuracy and to understand variations in TST.

The convenience of simpler instruments lies in the limited number of questions included in them, eg, how many hours do you feel you slept last night? In comparison, the ESSD includes more questions that, although it lengthens the time required to complete the survey, appear to guide the respondent in recording data that lead to more accurate TST measurements, as reflected in the correlation results (r = .753, n = 304, P < .001). While question 2 (actigraphy log) in the ESSD was used exclusively for ActiLife, it is possible that the table provided participants a time range that they used to estimate their sleep time, but this is not a weakness but rather a strength in that it helps firefighters to recall sleep data more accurately. In addition, many firefighters indicated that using the fire department computer-aided dispatch to indicate what time they were alerted for an emergency call and when they returned to the station were important in completing the ESSD and thus improving TST measurements. It is possible, then, that the ESSD structure allowed participants to calculate TST rather than guessing TST.

To the author’s knowledge, no self-reported TST measurement exists that addresses the job-related aspects of firefighter sleep such as multiple sleep bouts, sleeping at home and work, and suffering unintentionally forced awakenings when other units respond to emergency calls from the same station. The PSQI estimates TST by asking 1 average in-bed and 1 out-of-bed time, which may be an appropriate method for most individuals with traditional schedules but much less so for firefighters who have irregular work schedules and irregular sleep schedules. This at least partially explains the significant difference in mean scores. To date, however, the PSQI is the most used sleep instrument for firefighter sleep research; it is unknown how other researchers have addressed these concerns.

While exploratory, this study provides new knowledge in survey research design. In a community sample study, Buysse and colleagues⁶⁸ concluded that “the PSQI and ESS [Epworth Sleepiness Scale], two of the most widely used self-report measures of sleep and sleepiness, capture orthogonal dimensions of sleep-wake experience, and that both of these dimensions are distinct from what is measured with actigraphy and PSG” (p. 568).⁶⁸ The results here support their conclusion in a firefighter sample. PSQI significantly overestimates TST compared to actigraphy and also deviated from TST estimates derived from Extended Sleep Survey and ESSD. The implication of these results questions the accuracy of previously reported TST measurements in fire and emergency services research that use this, or similar, instruments. Future research is needed to explore the validity of self-reported measures of TST.

In addition to research implications, this research is also useful for practitioners. It is possible that new instruments such as the ESSD are more beneficial to fire service leadership who are increasingly interested in the relationships among sleep, health, and performance.

This pilot study admits limitations of its findings. Only 1 department was selected to participate, and the sample size was limited. With the extensive data recording requirements of this study, it was expected that some participants would not complete all requirements. Of those excluded due to data deficiencies, it is believed that participants were more fatigued with the study’s duration than they were with reporting sleep diary data. Ongoing and future research by this author will reevaluate the research question and hypotheses with a larger sample size and with additional fire departments.

CONCLUSIONS

This study presents evidence that suggests the popular PSQI may not be an appropriate sleep instrument in the fire and emergency services occupation. The results illustrate that firefighters overestimated PSQI TST by 1.14 hours compared to actigraphy (7.52 hours vs 6.38 hours, respectively). This study also presents 2 newer measurements to assess TST that were designed specifically to address the job-related characteristics that influence firefighter sleep. The ESSD appears to offer the benefits of both self-reported and objective measures and that work is needed to improve the habitual Extended Sleep Survey.

DISCLOSURE STATEMENT

The author has seen and approved the final manuscript. The author reports no conflicts of interest.

ABBREVIATIONS

24/48 Shift Schedule

24-hours on shift, 48-hours off shift

CSD

Core Consensus Sleep Diary

ESSD

Emergency Services Sleep Diary

H

Hypothesis

PSG

Polysomnography

PSQI

Pittsburgh Sleep Quality Index

TST

total sleep time