Background: Different screening tools to identify advanced Parkinson's disease (APD) have emerged in recent years. Among them, wearable medical devices, such as STAT-ON™, have been proposed to help to objectively detect APD.
Objectives: To analyze the correlation between STAT-ON™ reports and other assessment tools to identify APD and to assess the accuracy of screening tools in APD patients, using the STAT-ON™ as the gold standard.
Methods: In this retrospective, observational study, data from the University Hospital Complex of Pontevedra database on 44 patients with potential APD who wore STAT-ON™ were extracted. Data were collected according to different sources of tools for identifying APD: (1) STAT-ON™, (2) information provided by the patient, (3) questionnaire for advanced Parkinson's disease (CDEPA), (4) 5-2-1 Criteria, and (5) Making Informed Decisions to Aid Timely Management of Parkinson's Disease (MANAGE-PD). Considering STAT-ON™ recordings as a reference, the sensitivity, specificity, and positive and negative predictive values for each tool were calculated. The kappa index assessed the degree of agreement between the gold standard and the other instruments.
Results: Although no statistically significant association was found between STAT-ON™ recordings and any screening methods evaluated, the CDEPA questionnaire demonstrated the highest sensitivity and VPN values to detect patients with APD candidates for second-line therapy (SLT). According to the correlation analyses, MANAGE-PD demonstrated the highest degree of concordance with STAT-ON™ recordings to identify the SLT indication and to predict the SLT decision.
Conclusion: STAT-ON™ device may be a helpful tool to detect APD and to guide treatment decisions.
Keywords: Parkinson disease; levodopa; machine learning; motor complications; wearable electronic devices.
Copyright © 2023 Cabo-Lopez, Puy-Nuñez, Redondo-Rafales, Teixeira Baltazar and Calderón-Cruz.