Designing research studies in writer’s cramp dystonia: an analysis of automated writing measures

Journal Article

ABSTRACTBackgroundWriter’s cramp (WC) dystonia presents with abnormal postures during the task of writing and is an ideal dystonia subtype to study disease mechanisms for all forms of focal dystonia. Development of novel therapies is contingent on identifying sensitive and specific measures that can relate to the clinical syndrome and achieve a realistic sample size to power clinical research study for a rare disease. Although there have been prior studies employing automated measures of writing kinematics, it remains unclear which measures can distinguish WC subjects with high sensitivity and specificity and how these measures relate to clinician rating scales and patient-reported disability. The goal of this study was to: 1-identify automated writing measures that distinguish WC from healthy subjects, 2-measure sensitivity and specificity of these automated measures as well as their association with established dystonia rating scales, and 3-determine the sample size needed for each automated measure to power a clinical research study.Methods21 WC dystonia and 22 healthy subjects performed a sentence-copying assessment on a digital tablet in a kinematic software and hand recognition software. The sensitivity and specificity of automated measures was calculated using a logistic regression model. Measures were then correlated with examiner and patient rating scales. Power analysis was performed for 2 clinical research designs using these automated measures.ResultsOf the 23 automated writing measures analyzed, only 3 measures showed promise for use in a clinical research study. The automated measures of writing legibility, duration, and peak acceleration were able to distinguish WC from healthy controls with high sensitivity and specificity, correlated with examiner-rated dystonia sub-score measures and demonstrated relatively smaller sample sizes suitable for research studies in a rare disease population.DiscussionWe identified novel automated writing outcome measures for use in clinical research studies of WC subjects which capture key aspects of the clinical disease and can serve as important readout of dystonia disease mechanism as well as future disease interventions.

Full Text

Duke Authors

Cited Authors

  • Bukhari-Parlakturk, N; Lutz, M; McConnell, A; Al-Khalidi, H; Wang, JE-H; Scott, B; Termsarasab, P; Appelbaum, L; Calakos, N

Published By

Digital Object Identifier (DOI)

  • 10.1101/2021.03.02.21252036