Manually scoring actigraphy in the absence of a sleep diary: Reliability analysis in Gulf War veterans

Abstract

IntroductionThe process of manually scoring actigraphy data can be subjective with few studies utilizing and reporting clear actigraphy scoring guidelines. Clinical populations often have challenges with accurately and consistently completing sleep diaries which support actigraphy scoring. ObjectiveTo promote internal consistency and replicability in actigraphy scoring procedures, this study developed a detailed actigraphy scoring protocol for cases without sleep diary data and performed an inter-rater reliability analysis using a sample of veterans with Gulf War Illness (GWI). MethodsOne hundred fifty-nine nights of actigraphy data from a random subsample of 25 veterans with GWI were independently, manually scored using the protocol. Mean values for the start and end of rest intervals and derived sleep parameters—time in bed (TIB), total sleep time (TST), and sleep efficiency (SE)—and mean differences were calculated. Inter-rater reliability was evaluated using intra-class correlation (ICC). ResultsICC demonstrated excellent agreement between manual scorers for rest interval start (ICC = 0.98) and end times (0.99), TIB (0.94), TST (0.98), and SE (0.97). There were no clinically important differences (greater than 15 min) between manual scorers for the start of rest (average difference: 6 mins ± 28) or the end of rest (2 mins ± 23). ConclusionThis protocol may serve as a reproducible set of guidelines for researchers manually scoring actigraphy and enhance internal consistency for studies, especially for those working with clinical populations with significant sleep disturbance and related difficulties yielding quality sleep diary data.