Objective: To examine the agreeement between two objective Parkinson’s disease (PD) monitoring systems (PKG and STAT-ON) and clinical information from a resident physician’s interview or medical records.

Background: Accelerometry devices like PKG and STAT-ON can be used to monitor motor symptoms in PD, but results must be interpreted.  The reliability of interpretations and the performance of different objective PD monitoring systems has not been well described.

Method: Fourteen patients with PD were recruited at a university hospital and provided with both PKG and STAT-ON sensors to use simultaneously. After use patients responded to a usability survey. Two resident physicians without knowledge of patients medical records evaluated patients’ PD symptoms and categorized the current global treatment effects at the completion of sensor use. Blinded sensor data was interpreted and categorized independently regarding global treatment effect by two movement disorder specialists (raters) experienced in interpreting PKG. Agreement of categorization was calculated with weighted kappa (k) statistics.

Results: Simultaneous PKG and STAT-ON registrations were available for 11/14 patients with PD and only PKG data from three. There was substantial agreement between medical records and assessments of global treatment effects by resident physicians (k 0.76, p = 0.004). One of the two raters had moderate agreement between PKG rating and medical records (k 0.43, p = 0.045) and the other rater had substantial agreement between STAT-ON and medical records (k 0.66, p = 0.017). There was substantial agreement between  STAT-ON and resident physician (k 0.79, p = 0.007) for one rater. Inter-rater agreement was moderate for PKG-based assessment of global treatment effect (k 0.45, p = 0.034) and of sleep disturbance (k 0.52, p = 0.018) as well as for assessment of falls with STAT-ON (k 0.55, p =, 0.025). PKG provided new information relative to medical records in 7/14 measurements and STAT-ON in 4/11. Both systems scored high on patient rated usability.

Conclusion: In this small study we confirm that objective sensors can provide new information in a clinical context. There was however considerable variations between clinicians’ interpretation of data from sensors. This points to a need for studies of the reliability of interpretations of objective sensor data for PD monitoring.

« Back to 2023 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/comparision-of-objective-parkinsons-disease-monitoring-systems-and-the-interpretations-of-results/