Abstract Body: Enrichment is the prospective use of any patient characteristic to select a study population at higher risk in which detection of a drug effect is more likely than it would be in an unselected population. Patients with post-stroke cognitive impairment (PSCI) found to have higher risk of stroke recurrence in a recently completed meta-analysis. The goal of this study is to test whether PSCI based patient selection may represent enrichment strategy for secondary stroke prevention clinical trial. This is a subgroup analysis of Insulin Resistance Intervention after Stroke (IRIS) trial. In IRIS trial, patients were randomized to receive pioglitazone vs. placebo and had a baseline Modified Mini-Mental State Examination (3MS, where 3MS ≤ 88 was indicative of global PSCI. The primary endpoint of the study was recurrent stroke or MI. We estimated the hazard ratio (HR) for the effect of pioglitazone among those with global PSCI. To determine the sample size for a subsequent trial enriched by including only subjects with global PSCI, we make the following assumptions: (1) time to event follows an exponential distribution in both the pioglizone and placebo groups where the hazard rate for the placebo group is assumed to be the same as in the IRIS trial among those with global PSCI; (2) hazards for the pioglitizon and placebo groups are proportional over the course of the study; and (3) subjects are randomized to pioglitazone or placebo in equal proportions.Data on n = 3,338 patients of original cohort of n = 3,876 were analyzed, and n = 473 among them had PSCI at baseline. During 5-years of follow-up, n=246 patients experienced recurrent stroke, and n = 118 had MI. In patients with PSCI HR was 0.56 (95% CI 0.34 - 0.92) suggesting a 44% reduction in the hazard rate for secondary stroke or MI after 5 years of follow-up in the pioglitizone group compared to the placebo. If we conservatively assume that the true HR = 0.56 (closer to the null of HR = 1 than what was observed in the IRIS trial), then a total sample size of n= 967 willl proivde 90% power using a two-sided log-rank test at the 5% significance level. This conservative sample size corresponds to a 75% reduction in the sample size that was required for the IRIS Trial. PSCI screening may represent enrichment strategy for secondary stroke prevention clinical trial potentially reducing sample size by 75%. PSCI screening-based enrichment can be tested in phase 2 secondary stroke prevention trial.