International Stroke Trial Switzerland: some epidemiological data

Randomised trials provide the best evidence on the effects of treatment on a particular disease. They can also provide valuable data on outcome. In th e present article, data from 1631 Swiss patients randomised in the Internat ional Stroke Trial (IST) are presented. Baseline characteristics and outcom e in the Swiss patients were compared with the 17 804 patients randomised i n other countries. On average, compared with other countries, Swiss patient s were: 2.5 years older (CI: 1.9-3.1; p <0.001); more likely to have presen ted with total anterior circulation infarcts (28 vs 23%); less likely to ha ve lacunar strokes (16 vs 25%; p <0.00001); more likely to be randomised an d treated early (within 3 hours, 7 vs 4%, and within 6 hours, 25 vs 15%; p <0.00001). However, Swiss patients were more likely to be dead or dependent 6 months after the stroke (66.9 vs 62.2%; p <0.00001). The difference in d eath or dependency was almost entirely explained by the difference in age a nd baseline severity of strokes. These data highlight the difficulties inhe rent in between-country comparisons of outcome after stroke. They also emph asise the need for international multicentre trials to use methods of rando misation (either stratification or minimisation), as IST did, to ensure bal ance of prognostic factors within each participating country; to do otherwi se might confound the assessment of the effect of treatment.