Measuring disability in stroke: relationship between the modified Rankin scale and the Barthel index

Abstract

The effectiveness of therapeutic interventions in acute stroke trials is traditionally measured with the modified Rankin scale (mRs) and the Barthel index (BI). The mRs is a global disability scale divided into six steps from total independence to total dependence. The BI assesses ten basal activities of daily living, of which eight assess level of dependence (bathing, grooming, using stairs, dressing, feeding, toilet use, transfers and walking). The aim of this study was to investigate the relationship between the mRs and the total scores and item-scores of the BI. During a period of 3 months mRs and BI scores were collected from two multicentre randomised, placebo-controlled trials with lubeluzole (515 and 519 patients). In each patient we compared the mRs grades with the total BI score and the scores on the ten subitems. For both trials there was extensive overlap of BI scores between mRs grades and a wide range in BI scores among patients with mRs grades 3 and 4. We also found discrepancies between the BI item-scores and mRs grades. About 40% of patients with mRs grades 1 (able to carry out all usual activities) and 2 (able to look after own affairs without assistance) were not independent on at least one activity of the BI. In both studies, about 30% of the patients needed help or supervision for walking, although they were classified as mRs 3 (requiring some help but able to walk without assistance). Investigators in stroke trials use the mRs as a subjective global disability scale, and they do not strictly take into account limitations in performing specific basal activities of daily living, as assessed by the BI, to assign mRs grades.