Can the Dyskinesia Impairment Scale be used by inexperienced raters? A reliability study

The Dyskinesia Impairment Scale (DIS) is a new scale for measuring dystonia and choreoathetosis in dyskinetic Cerebral Palsy (CP). Previously, reliability of this scale has only been assessed for raters highly experienced in discriminating between dystonia and choreoathetosis.

The aims of this study are to examine the reliability of the DIS used by inexperienced raters, new to discriminating between dystonia and choreoathetosis and to determine the effect of clinical expertise on reliability.

Twenty-five patients (17 males; 8 females; age range 5-22 years; mean age = 13 years 6 months; SD = 5 years 4 months) with dyskinetic CP were filmed with the DIS standard video protocol. Two junior physiotherapists (PTs) and three senior PTs, all of whom were new to discriminating between dystonia and choreoathetosis, were trained in scoring the DIS. Afterward, they independently scored all patients from the video recordings using the DIS. Reliability was assessed by (1) Intraclass Correlation Coefficient (ICC), (2) Standard Error of Measurement (SEM) and Minimal Detectable Difference (MDD) and (3) Cronbach's alpha for internal consistency.

Interrater reliability for the total DIS, and for the dystonia and choreoathetosis subscales was good for the junior PTs and moderately high to excellent for the senior PTs. SEM and MDD values for the total DIS were 6% and 15% respectively for the junior PTs and 4% and 12% respectively for the senior PTs. Cronbach's alpha ranged between 0.87 and 0.95 for the junior PTs and between 0.76 and 0.93 for the senior PTs.

Reliability of the DIS scores for the inexperienced junior and senior PTs was sufficient in comparison with scores from the experienced raters in the previous study, indicating that the DIS can be used by inexperienced PTs new to discriminating between dystonia and choreoathetosis, and also that its reliability is not dependent on clinical expertise. However, based on the measurement errors and questionnaire data, familiarity with operational definitions of dystonia and choreoathetosis is crucial to improve scoring reliability.