The challenge of designing a treatment trial for warfarin-associated intracerebral hemorrhage.

BACKGROUND AND PURPOSE Warfarin-associated intracerebral hemorrhage (WICH) became more frequent in the past 2 decades. Interest in potential WICH treatment trials has grown, but the practicality of such trials has received less attention. We determined the number of patients that would be eligible for enrollment in hypothetical treatment trials for WICH using a population-based study. METHODS We identified all patients aged 18 years or older from the Greater Cincinnati/Northern Kentucky region with nontraumatic intracerebral hemorrhage in 2005. Three hypothetical WICH treatment trial criteria sets were used to determine eligibility for enrollment, varying from relatively strict to broadly inclusive. For the hypothetical trials, we assumed the comparison of a standard therapy to an alternative therapy. Sample size calculations assumed different rates of poor outcome depending on the criteria set, various effect sizes, a 2-sided alpha of 0.05, and 80% power. Given 5 years of trial enrollment, the population base needed to enroll the required subjects was then calculated. RESULTS Warfarin-associated intracerebral hemorrhage accounted for 54 of 286 (19%) cases of intracerebral hemorrhage within the Greater Cincinnati/Northern Kentucky region in 2005. Eligibility rates ranged from 2 of 54 WICH patients (4% of cases, strictest set) to 11 of 54 WICH patients (20% of cases, most inclusive set). Given these rates, a population base of at least 67 million persons would be required to conduct a 5-year trial for WICH with a 10% effect size using a moderately strict criteria set. CONCLUSIONS Any planned treatment trial for WICH should anticipate significant challenges in successfully enrolling adequate numbers of patients.