Cluster Randomized Trials in Comparative Effectiveness Research

BACKGROUND: The need for evidence about the effectiveness of therapeutics and other medical practices has triggered new interest in methods for comparative effectiveness research. OBJECTIVE: Describe an approach to comparative effectiveness research involving cluster randomized trials in networks of hospitals, health plans, or medical practices with centralized administrative and informatics capabilities. RESEARCH DESIGN: We discuss the example of an ongoing cluster randomized trial to prevent methicillin-resistant Staphylococcus aureus (MRSA) infection in intensive care units (ICUs). The trial randomizes 45 hospitals to: (a) screening cultures of ICU admissions, followed by Contact Precautions if MRSA-positive, (b) screening cultures of ICU admissions followed by decolonization if MRSA-positive, or (c) universal decolonization of ICU admissions without screening. SUBJECTS: All admissions to adult ICUs. MEASURES: The primary outcome is MRSA-positive clinical cultures occurring >or=2 days following ICU admission. Secondary outcomes include blood and urine infection caused by MRSA (and, separately, all pathogens), as well as the development of resistance to decolonizing agents. RESULTS: Recruitment of hospitals is complete. Data collection will end in Summer 2011. CONCLUSIONS: This trial takes advantage of existing personnel, procedures, infrastructure, and information systems in a large integrated hospital network to conduct a low-cost evaluation of prevention strategies under usual practice conditions. This approach is applicable to many comparative effectiveness topics in both inpatient and ambulatory settings.