How Pragmatic are Trials in International Nursing Home Settings?

KATE MAGID
Health Science Specialist, Rocky Mountain Regional Veterans Affairs Medical Center

At the 2019 AMDA-The Society for Post-acute and Long-term Care Medicine symposium, researchers discussed the implementation of pragmatic trials in nursing homes. Relatively few pragmatic trials have been conducted in nursing homes. In this abstract, we review the extent to which the design and implementation of trials presented at the AMDA symposium were pragmatic with a goal of describing approaches to improve pragmatic nursing home research study design.

All trials were conducted in nursing homes, with sample sizes ranging from 12- 175 homes.1-6 Trials were conducted Europe and the United States. Participants varied by study and included nursing home residents, nursing home staff, and caregivers.

We used the Pragmatic Explanatory Continuum Indicator Summary (PRECIS-2) criteria to rate the pragmatic nature of the trials.7-8 Given that these trials were conducted at the level of the nursing home, reviewers rated recruitment and eligibility for nursing homes and residents separately. Subsequently, the reviewers discussed ratings and came to consensus. A PRECIS wheel was constructed for each trial to visually represent where the trial aligned on the explanatory-pragmatic continuum.

Figure 1 shows the PRECIS-2 wheels summarizing the pragmatic nature of the trials. Using PRECIS-2 criteria, the domains identified as most pragmatic across the trials were setting, primary outcome, and primary analysis. All studies were conducted in nursing homes that resembled usual care, collected primary outcomes relevant to participants, and almost all conducted an intention-to-treat analysis. In contrast, organization, nursing home recruitment, and nursing home eligibility were the least pragmatic. Several trials hired staff for intervention delivery or evaluation, implemented trainings, or provided resources beyond what is available in usual care, thus making the organization domain more explanatory. The eligibility criteria and recruitment for nursing homes were less pragmatic due to excluding homes based on; resident census, location, existing interventions and recruiting homes through advertisements or mailings. Four trials required residents or their legal guardian to provide informed consent to participate, making resident recruitment less pragmatic.

The application of PRECIS-2 criteria to the trials presented at the AMDA symposium suggests that the implementation of pragmatic trials in nursing homes have pragmatic and explanatory components. Future studies should explore how requiring residents to provide consent might lead to participants not being fully representative of the usual care population, especially given the prevalence of dementia in nursing homes. To better suit cluster pragmatic trials, researchers should rate the eligibility and recruitment of facilities and participants separately.