Browsing by Author "Grahn, Filip"

Filter results by typing the first few letters
Now showing 1 - 1 of 1
  • No Thumbnail Available
    Item
    Evaluation of Two Commercial Sensor Systems for Monitoring Parkinsonism and Their Possible Influence on Management of Parkinson’s Disease
    (2022-02-23) Grahn, Filip; University of Gothenburg / Institute of Medicine; Göteborgs universitet / Institutionen för medicin
    BACKGROUND Wearable sensors can be used to monitor motor symptoms in Parkinson’s Disease. The Parkinson’s KinetiGraph (PKG) and STAT-ON are two promising single-sensors. No previous studies have been made comparing these two, neither in terms of agreement nor in terms of usability. AIMS Compare agreement between PKG, STAT-ON, a resident physicians’ assessment, and patients’ medical records. Describe usability from the patients’ view. METHOD Ten patients recruited from Sahlgrenska University Hospital wore two sensor systems (PKG and STAT-ON) and reports were assessed for typical categories of motor symptoms. A resident physician categorized the patients in the same way after taking the patients’ history of motor symptoms and fluctuations. Agreement was evaluated with Cohen’s kappa, and usability surveys and self-assessment scales were filled out. RESULT Compared to information derived from the patients’ medical records, agreement was seen for the resident physician (kappa = 0.747, p = 0.015) and for one of two STAT-ON raters (kappa = 0.673, p = 0.023). Agreement between STAT-ON and the resident physician was significant for one of two raters (kappa = 0.783, p = 0.014). No significant agreement was seen between PKG and STAT-ON evaluations nor between PKG and the resident physician. Both sensors had a low mean rating score in the usability survey (1-5, lower better), PKG = 1.67 ± 0.56, and STAT-ON 2.01 ± 1.10. CONCLUSION The STAT-ON sensor provides information on motor symptoms that is more consistent with the resident physician than PKG. However, PKG might to a greater extent provide different information. Future studies are needed to understand the implications of this. The low mean rating score for usability indicates potentially high acceptability for sensors.