What should be defined as good outcome in stroke trials; a modified Rankin score of 0-1 or 0-2?

Background and purpose Good outcome in stroke trials has been defined as a modified Rankin scale (mRs) score of 0-1 or 0-2. The aim of this study was to investigate the clinical meaning of these two dichotomies. Methods We studied 152 patients six months post stroke using the mRs and a new disability measure the AMC Linear Disability Scale (ALDS) item bank. Descriptive statistics were used to show the ALDS scores by the levels of the mRs. To investigate the clinical meaning of the different definitions of good outcome, the mean probability to perform activities of daily life (ADL) of all mRs grades and these two dichotomies was calculated. Results The ability to perform difficult ALDS items declined gradually with increasing mRs grade. When favourable outcome is defined as mRs 0-1, 15% of the cohort has a good outcome; of these patients 84 % were likely to perform outdoor activities. If good outcome is defined as mRs 0-2, the percentage of patients with good outcome increased to 37%, whereas 66% of these patients were likely to perform outdoor activities. Conclusion If good outcome is defined as the ability to perform outdoor activities mRs 0-1 should be chosen. If complex ADL are considered as good outcome mRs 0-2 is the outcome measure of choice. Independent of which outcome measure is chosen, the treatment effect in clinical trials must be large before good outcome is achieved. Therefore, it is likely that clinically important treatment effects can be missed in clinical trials with both these mRs endpoints.