Measuring functional outcomes in stroke trials: Improving inter-rater reliability of the Modified Rankin Scale using a structured interview
Background & aims: The Modified Rankin Scale (MRS) (van Swieten et al, 1988) is widely used in clinical trials to rate disability and handicap after stroke. Although the MRS is a popular measure of functional outcome the categories of the scale are very broadly defined and open to interpretation by raters. Previous work with the Glasgow Outcome Scale indicates that the reliability of functional rating scales may be improved by use of a structured interview (Wilson et al, 1998). The purpose of the present study was to compare the inter-rater reliability of the conventional MRS with the inter-rater reliability of a newly developed structured interview for the MRS (MRS-SI) Methods: A structured interview was devised for the MRS covering five areas of everyday function. 63 patients with stable functional state after stroke (stroke 6 to 24 months previously) were recruited to the study and scored on the conventional MRS by two independent observers. These observers then underwent training in use of the MRS-SI. Eight weeks after the first assessment the same observers reassessed 58 of these patients using the MRS-SI. Results: To allow comparison between the assessments the analysis of results was restricted to the 58 patients who were rated on both the MRS and MRS-SI. Inter-rater reliability was measured using the kappa statistic (unweighted and weighted using quadratic weights). For the MRS, overall agreement between the two raters was 57% (unweighted kappa 0.44, weighted kappa 0.78); using the MRS-SI, overall agreement was 78% (unweighted kappa 0.70, weighted kappa 0.93). Conclusions: Variability between raters in assigning patients to Rankin grades appears to be reduced when using a structured interview for the Modified Rankin Scale. The use of the MRS-SI could potentially improve the quality of results from clinical studies in stroke. A multi-centre study to further establish the improvement in inter-rater reliability is ongoing.