Inter-rater variability in motor function assessment in Parkinson's disease between experts in movement disorders and nurses specialising in PD management

Introduction and objective: In clinical practice, assessing patients with Parkinson's disease (PD) is a complex, time-consuming task. Our purpose is to provide a rigorous and objective evaluation of how motor function in PD patients is assessed by neurologists specialising in movement disorders, on the one hand, and by nurses specialising in PD management, on the other.

Methods: We conducted an observational, cross-sectional, single-centre study of 50 patients with PD (52% men; mean age: 64.7 ± 8.7 years) who were assessed between 5 January 2016 and 20 July 2016. A neurologist and a nurse evaluated motor function in the early morning hours using the Unified Parkinson's Disease Rating Scale (UPDRS) parts III and IV and Hoehn & Yahr (H&Y) scale. Tests were administered in the same PD periods (in 48 patients during the 'off' time and in 2 patients during the 'on' time). Inter-rater variability was estimated with the intraclass correlation coefficient (ICC).

Results: Forty-nine patients (98%) were classified in the same H&Y stage by both raters. Assessment times were similar for both raters. ICC for UPDRS-IV and UPDRS-III total scores were 0.955 (P<.0001) and 0.954 (P<.0001), respectively. The greatest variability was found for UPDRS-III item 29 (gait; ICC=0.746; P<.0001) and the lowest, for item 30 (postural stability; ICC=0.918; P<.0001).

Conclusions: Motor function assessment of PD patients by a trained nurse is equivalent to that made by an expert neurologist and takes the same time to complete.