The study evaluated several approaches for identifying IN and OUT of bedtimes, comparing their reliability, accuracy, and impact on calculating sleep and circadian parameters. In a 12-month observational study, 72 adults with advanced cancer wore wrist and thigh accelerometers for 72 hours and completed sleep diaries and questionnaires on chronotype, sleep quality, and daytime sleepiness. IN and OUT times were determined using patient diaries, wrist accelerometry event markers, and automated accelerometry algorithms. Automated algorithms identified IN and OUT times more consistently than patient-reported methods (93–100% vs. 48–83%), although both under- and over-estimation of timings occurred. Timings from all methods were significantly correlated (p < 0.001). Excellent agreement was observed between patient‑reports (patient diary and watch event marker, ICC 0.829–0.877), and patient-reports and thigh accelerometry (ICC 0.892). Event markers appeared more accurate than sleep diaries, showing closer agreement with associated DOWN and UP times. The choice of method for identifying IN and OUT times influenced the calculated sleep onset latency, percent sleep and dichotomy index values. These findings highlight the importance of accurate bedtime identification for reliable sleep and circadian analyses, particularly when diagnosed thresholds are used. Further research is required to determine the most accurate and clinically practical approach.
Gouldthorpe et al. (Tue,) studied this question.