Telephone interview for stroke outcome assessment

Two groups of stroke patients were evaluated in order to test the reliability of telephone assessment of stroke disability with the modified Rankin scale. A first group of 53 cases was evaluated once by telephone and once by direct examination to test intermodality agreement, and a second group of 48 patients w'as interviewed by telephone by two independent observers to test interobserver agreement. As shown by kappa (k) statistics, the intermortality agreement was good (k = 0.72, 95% confidence interval 0.59-0.85) and the interobserver agreement moderate (k = 0.48, 95% confidence interval 0.32- 0.65). Thus the use of the Rankin scale by telephone instead of direct examination appears reliable. The low interobserver reproducibility of the scale indicates that the number of observers should be reduced which is possible in case of a telephone evaluation.