Validity and reliability of the Simpson‐Angus Scale (SAS) in drug induced parkinsonism in the elderly

Abstract

Quantification of drug induced parkinsonism (DIP) for study purposes is difficult. The most often used Simpson Angus Scale (SAS) lacks proper clinimetric evaluation. The newer Schedule for Assessment of Drug-Induced Movement Disorders (SADIMoD) shows good clinimetric characteristics, but has not been used in published clinical studies, probably due to the complexity of the scale. To evaluate internal consistency and inter-rater reliability of the SAS and the correlation ot f the SAS with the parkinsonism subscale of the SADIMoD in elderly. Fifteen elderly diagnosed with DIP were recruited. The patients were assessed three times with the SAS by three independent investigators. The resident also performed the SADIMoD. Internal consistency was measured by Cronbach's alpha-coefficient, inter-rater variability was examined with weighted kappa values and percentage of agreement and correlation to SADIMoD by Spearman's correlation coefficient. SAS demonstrated good internal consistency reliability (Cronbach's alpha coefficients 0.83). Inter-rater reliability for sum score was good. For individual items slight agreement on the item salivation and moderate to very good agreement on remaining items calculated by weighted kappa values was reached. We found 87-100% agreement on the individual items with acceptance of 1 point difference between raters. The SAS demonstrated acceptable correlation with the SADIMoD parkinsonism subscale scores (Spearman's rho = 0.66; p < 0.01). The SAS appears to be a valid and by different instructed health care professionals easy to perform research tool to evaluate DIP.