Objective: To compare adherence to exercise between participant logbook records and device recorded measures to determine: if participant logbooks were accurate, and if there were any patterns to logbook accuracy, in people with Parkinson’s disease (PD).

Background: Exercise is commonly prescribed as part of the overall management for people with PD. Optimal exercise prescription relies on clinician knowledge of the extent of patient adherence, typically monitored via logbooks kept by patients. The accuracy of these logbooks is unknown.

Methods: Data from 15 participants with PD from the intervention arm of a randomized controlled trial investigating a minimally-supervised upper limb exergame intervention [1] were analyzed. Participants were prescribed at least 3 sessions of 12 games per week for 12 weeks. Each participant’s completed games and sessions were recorded in two ways: by the person in a paper-based logbook; and by the gaming device in an electronic record unseen by the participant. Logbook and device recorded adherence were calculated. Intra-class correlation coefficients (ICCs) were calculated to determine logbook accuracy. Bland Altman plots were used to examine for systematic bias between logbook and electronic records.

Results: Overall adherence was similar between logbook and electronic records, however there was substantial individual variability (mean (95%CI) for sessions: logbook 107% (86-144%), electronic 105% (72-161%); for games: logbook 116% (86-178%), electronic 132% (97-197%)). Similarly, ICCs for the overall session and game data were fair to good, but individual participant ICCs (95%CI) were highly variable, ranging from minimal to perfect agreement between logbook and electronic records (sessions: ICC 0.656 (-0.02–1.0); games: ICC 0.563 (-0.24–0.99)). The pattern of errors in the logbooks suggested that participants were biasing entries to match the prescribed exercise.

Conclusions: Self-report logbooks provided an accurate measure of overall adherence. However, the accuracy of individual logbooks was highly variable, and demonstrated a tendency for participants to record the prescribed exercise regardless of whether they had completed more or less. Logbooks should not be used in isolation to assess a person’s adherence to exercise.

References: 1. Allen NE et al, Parkinsonism Relat Disord 2017;44:66-72.

« Back to 2018 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/how-accurate-are-exercise-logbooks-a-comparison-of-participant-and-device-recorded-exercise-records-in-people-with-parkinsons-disease/