Optimising fall risk classification models in Parkinson's disease using clinical and mobility outcomes.

Research overview

The text below is the source article’s original abstract under a verified reuse license. cc by

BackgroundFalls are a serious concern in Parkinson's disease (PD), often leading to hospitalisation, dependence and reduced quality of life. Effective fall management requires identification of those at risk. This cross-sectional, discriminative study aimed to evaluate which selection of outcomes best discriminate retrospective fallers from non-fallers, to inform a future prospective clinical prediction tool.MethodsPeople with PD were recruited (ICICLE-GAIT; 54- and 72-month follow up assessments). Fallers and non-fallers were stratified based on prospective fall reports. A total of 299 outcomes across 4 domains were collected: clinical (n = 9), lab-based mobility (gait n = 60, turning n = 99), real-world mobility (n = 131). Receiver operating characteristic analysis evaluated classification models distinguishing fallers from non-fallers. Area under the curve (AUC) determined which models were optimal. Models were re-applied at 72-months.ResultsOf the 48 participants, 32 (67%) were classified as fallers and 16 (33%) as non-fallers. Significant group differences (faller vs. non-faller) were found in all domains at 54-months; clinical (n = 2/9), lab-based gait (n = 8/60), lab-based turning (n = 19/99) and real-world mobility (n = 5/131). At 54-months, turning was the strongest single-domain model (apparent AUC = 0.86, sensitivity = 0.74, specificity = 0.93; optimism-corrected AUC = 0.61), followed by real-world mobility (AUC = 0.82, sensitivity = 0.68, specificity = 1.00; optimism-corrected AUC = 0.72). All multi-domain combinations including turning showed acceptable discrimination, with AUC values > 0.70. At 72-months, turning alone retained apparent perfect discrimination (AUC = 1.00), whereas clinical (AUC = 0.52), gait (AUC = 0.51) and real-world (AUC = 0.69) models declined substantially.ConclusionTurning showed strong discriminative power in classifying PD fallers remaining robust over time and outperforming other assessments. Real-world mobility also had strong discriminative value, highlighting the importance of ecologically valid continuous monitoring. As this study was discriminative and exploratory in nature, these findings should be interpreted as hypothesis-generating pending external validation. Future models should explore whether real-world turning provides superior discriminative value to lab-based turning.

Related topics