To investigate the reliability of a scoring system to differentiate movement severity by observing the signals and transforms of instrumentation to identify the position in space of the extremities of people with Parkinson’s disease (PD)

The diagnosis and treatment of PD may be enhanced by the development of a procedure for trained raters to score the output of instrumentation for motor assessments without information about the patients.

Thirty-five trained raters certified in the MDS-UPDRS applied a similar scoring system  for the evaluation of signals and fast Fourier transforms (FFTs) and continuous wavelet transforms (CWTs) of accelerometers attached to the upper and lower extremities of 10 patients with PD  (McKay, et al., MethodsX 6 (2019) 169-189). The raters were presented the signals and FFTs and CWTs as panels with the x, y, and z output of five repetitive tasks {upper extremity [pointer finger and wrist]: (3.4FT: Finger tapping, 3.5HM: Hand movements, 3.6PS: Pronation-supination of the hands) and lower extremity [big toe and ankle]: (3.7TT: Toe tapping, 3.8LA: Leg agility)}. Raters were allowed to take as much time as needed to perform the ratings independently without consultation with others. The results were collected and classified into two categories for each of FFT and CWT as following (< mild, > moderate).

 Optimal reliability was attained for ratings < mild. The mean percentage agreements for the ratings of < mild were (A) FFT: 3.4FT = 74%, 3.5HM = 82%, 3.6PS = 79, 3.7TT = 81%, 3.8LA = 76%) and (B) CWT: 3.4FT = 68%, 3.5HM = 68%, 3.6PS = 80%, 3.8LA = 72%). Ratings > moderate were not significant.

Trained raters can differentiate level of severity of repetitive movements of people with PD by visual observation of accelerometry output. This protocol provides the foundation for optimal clinical trials of PD and related conditions.