Minimum Number Of Monitoring Days Necessary To Reliably Estimate On-duty Stepping Activities Among Police Officers

Abstract

It has been reported that police officers face the challenge of long shift (up to 13 hours) that is comprised of physical inactivity, which is associated with increased risk for cardiovascular diseases. Prior to implementing a workplace intervention, it is crucial to find a reliable method to track changes in the activity outcomes. PURPOSE: To apply the Generalizability (G) theory analysis to determine the minimum number of monitoring days necessary to reliably capture on-duty stepping activities in police offers. METHODS: Eighteen police officers from the local police department were recruited and completed a 7-day activity monitoring wearing an inclinometer on mid-thigh of the non-dominant leg. Based on the activity log, periods of on-duty shifts were adjusted by combining two days of inclinometry data, and on-duty step counts were computed for each on-duty shift. G-theory analysis was used to decompose total variance into the three components: person-to-person (true variability), day-to-day (systematic error), and person-by-day (random error). The estimated variance components were used to examine the change in reliable coefficient (i.e., generalizability (g) coefficient) by different measurement conditions (i.e., number of days). Prior to G-theory analysis, missing data were imputed using a likelihood-based expectation-maximization algorithm. RESULTS: A total of 15 subjects provided valid step counts data for a median of 3 on-duty days (range: 2-5 days), with an average wear time of 9.89 hours (≈96% of on-duty working hours). Overall, the average step counts accumulated per on-duty shift was 4087.48 ± 1084.58 steps. Estimated variance components based on G-theory analysis showed that a large amount of variability (87.8%) was attributed to person-by-day effect (i.e., random error), and true variability by the person was relatively small (6.5%). A follow-up decision study showed that a minimum of 53 on-duty days was required to obtain the stable (g-coefficient ≥ .80) measure of on-duty step counts. CONCLUSION: Results of the study revealed that a considerably longer monitoring period was necessary to reliably capture on-duty stepping activities of police officers, indicating the importance of determining the length of activity monitoring with an understanding of specific occupational environment.