CAN SIMPLE QUESTIONS ASSESS OUTCOME AFTER STROKE

We assessed the validity and reliability of simple questions to assess outcome after stroke in a group of 60 patients. 100 consecutive patie nts admitted to hospital with acute stroke were prospectively studied, and survivors were randomised to follow-up either by postal questionn aire or telephone interview. The responses to the postal or telephone assessment from the 60 patients who could participate in the study wer e compared to the Barthel Index (BI) and Oxford Handicap Scale (OHS) o btained from a face-to-face interview by a trained nurse (who was 'bli nd' to the postal or telephone responses), The response to the postal questionnaire or telephone interview item: 'In the last 2 weeks, did y ou require help from another person for everyday activities?', had an accuracy of 75% in identifying patients with bad outcome (defined by a BI <20/20) and an accuracy of 83% when bad outcome was defined by an OHS category of 3, 4 or 5. The answer to the item 'Do you feel that yo u have made a complete recovery from your stroke?' had an accuracy of 90% when compared with an OHS category of 0 obtained at the nurse inte rview. The responses to the simple 'dependency' and 'recovery' questio ns obtained by postal questionnaire and telephone interview agreed wel l with the responses to the same questions administered by the nurse ( kappa = 0.62 and 0.73, respectively). There was no difference between outcome assessed by postal or telephone methods, and no clinically imp ortant difference between self- or carer-completed questionnaires. The se results suggest that very simple measures of outcome after stroke c an be developed. The two simple questions were practical, valid, relia ble and could be administered effectively by post or telephone, withou t the need for a face-to-face interview; this method would therefore b e suitable for all types of very large-scale studies of outcome after stroke (e.g. randomised trials, observational studies and multicentre audits of stroke care).