A Self-Reported Clinical Tool Predicts Falls in People with Parkinson's Disease

Abstract

Background: A 3-step clinical prediction tool including falling in the previous year, freezing of gait in the past month and self-selected gait speed <1.1 m/s has shown high accuracy in predicting falls in people with Parkinson's disease (PD). The accuracy of this tool when including only self-report measures is yet to be determined. Objectives: To validate the 3-step prediction tool using only self-report measures (3-step self-reported prediction tool), and to externally validate the 3-step clinical prediction tool. Methods: The clinical tool was used with 137 individuals with PD. Participants also answered a question about self-reported gait speed, enabling scoring of the self-reported tool, and were followed-up for 6 months. An intraclass correlation coefficient (ICC2,1) was calculated to evaluate test–retest reliability of the 3-step self-reported prediction tool. Multivariate logistic regression models were used to evaluate the performance of both tools and their discriminative ability was determined using the area under the curve (AUC). Results: Forty-two participants (31%) reported ≥1 fall during follow-up. The 3-step self-reported tool had an ICC2,1 of 0.991 (95% CI 0.971–0.997; P < 0.001) and AUC = 0.68; 95% CI 0.59–0.77, while the 3-step clinical tool had an AUC = 0.69; 95% CI 0.60–0.78. Conclusions: The 3-step self-reported prediction tool showed excellent test–retest reliability and was validated with acceptable accuracy in predicting falls in the next 6 months. The 3-step clinical prediction tool was externally validated with similar accuracy. The 3-step self-reported prediction tool may be useful to identify people with PD at risk of falls in e/tele-health settings.