Efficacy and effectiveness as aspects of cluster randomized trials with nursing home residents: Methodological insights from a pneumonia prevention trial

This report discusses how methodological aspects of study efficacy and effectiveness combine in cluster randomized trials in nursing homes. Discussion focuses on the relationships between these study aspects in the Pneumonia Reduction in Institutionalized Disabled Elders (PRIDE) trial, an ongoing cluster randomized clinical trial of pneumonia prevention among nursing home residents launched in October 2009 in Greater New Haven, Connecticut. This clinical trial has enrolled long-term care nursing home residents, over 65 years in age, who have either inadequate oral care or swallowing difficulty, previously identified risk factors for pneumonia. It has used a multicomponent intervention consisting of manual tooth/gum brushing, 0.12% chlorhexidine oral rinse administered twice daily by nurses, and upright feeding positioning at meals to reduce rates of radiographically documented pneumonia. Cluster randomization is attractive for nursing home intervention studies because physical proximity and administrative arrangements make it difficult to deliver different interventions to residents of the same nursing home. Implementing an intervention in an entire home requires integration into the daily life of residents and into the administrative procedures of the nursing home. This characteristic of nursing home cluster randomized trials makes them approximate "real-world" research contexts, but implementation can be challenging. The PRIDE trial of pneumonia prevention utilized specific methodological choices that include both efficacy and effectiveness elements. Cluster randomized trials in nursing homes having elements of both efficacy and effectiveness (i.e., hybrid designs) can address some of the methodological challenges of conducting clinical research in nursing homes: they have distinctive advantages and some limitations. (C) 2012 Elsevier Inc. All rights reserved.