Comparing actigraphy and diary to measure daily and average sleep in firefighters: a Bland-Altman analysis.
Level 4 - case-series / case-control
Observational method-comparison study without a diagnostic gold standard
PubMed 37950454 · doi:10.5664/jcsm.10916
What was done
Researchers evaluated the agreement between wrist actigraphy (Actiwatch-2) and the Consensus Sleep Diary among 60 career firefighters from an urban department working a 5/6 shift schedule. Data were collected over a 6-day between-shift recovery period, generating 329 paired daily measurements. Nine sleep parameters were evaluated using Bland-Altman analysis: wake after sleep onset (WASO), total sleep time (TST), sleep onset latency, sleep offset, in-bed time, lights-off time, out-of-bed time, wake time, and sleep efficiency.
What was found
Firefighters significantly overestimated sleep efficiency and underestimated WASO relative to actigraphy by amounts exceeding American Academy of Sleep Medicine clinical significance thresholds. All nine indices exhibited broad limits of agreement; specifically, the 95% confidence interval for diary and actigraphic total sleep time estimates spanned a 4.7-hour range. Exact mean bias numbers were not reported in the abstract.
Why it matters
Self-reported sleep diaries and wrist actigraphy diverge substantially during between-shift recovery in firefighters. Clinicians assessing shift work disorder or sleep disturbances in this population cannot treat these two standard measures as interchangeable.
Limits
The study lacked a gold standard (such as polysomnography) to confirm which tool was more accurate. Data were limited to a single urban department on a specific shift schedule, assessed only between-shift recovery sleep rather than on-shift sleep, and the abstract omits exact point estimates and standard deviations for individual sleep indices.
Cited by
- supports Wrist-movement-based sleep tracking underestimates wake time compared to EEG brain wave measurements, but overestimates wake time compared to self-report.