The accuracy of self-report logbooks of adherence to prescribed home-based exercise in Parkinson's disease

Purpose: This study compared self-report logbooks of exercise performance to objective measures of performance to determine the accuracy of and patterns of misreporting in self-report logbooks in people with Parkinson's disease. Materials and methods: Fifteen participants from the intervention arm of a randomized control trial were prescribed a minimally supervised, 12-week, home-based upper limb exergame program (ACTRN 12614001048673). The exergame system provided an objective electronic measure of adherence for comparison with self-report logbooks. Results: Logbooks showed excellent to good accuracy of overall reported adherence to prescribed sessions (Intraclass correlation (ICC) = 0.83) and games (ICC 0.71). Logbooks were also a good to fair representation of weekly adherence across participants for both sessions (ICC 0.66) and games (ICC 0.56). Individual participant ICCs ranged from minimal to perfect agreement between logbooks and electronic records (ICC sessions range: -0.02 to 1; games range: -0.24 to 0.99). The pattern of logbook reporting suggested some participants were biasing entries to match prescribed exercise. Conclusions: Self-report logbooks may provide an accurate measure of overall adherence. However, the accuracy of individual logbooks was highly variable indicating caution is needed in using self-report measures to assess individual adherence in intervention studies and for clinical decision making. Clinical Trial Registration: Registered in Australia and New Zealand Clinical Trials Registry (https://www. anzctr.org.au/): Registration number: ACTRN12614001048673. IMPLICATIONS FOR REHABILITATION Adherence to prescribed unsupervised exercise is usually self-reported. Self-reported logbooks gave a good to excellent indication of exercise adherence overall when compared to electronically captured records for a group of participants with mild to moderate Parkinson's disease. There were high levels of variability in the accuracy of individual logbooks with a tendency to record prescribed rather than actual exercise. Logbooks may be accurate to monitor group adherence, but caution is required when using them to assess an individual's adherence to an exercise prescription.