Pragmatic enrollment strategies enroll older adults in hypertension trial

Clinical trials of blood pressure treatments often enroll patients who look little like the people doctors actually treat, and the mismatch can limit how confidently trial results apply to everyday care. A team testing remote monitoring and virtual collaborative care for hypertension in adults 65 and older tried to close that gap by building pragmatic enrollment and intervention strategies into the design of a hybrid type 1 effectiveness-implementation study, and then measuring how well the enrolled cohort matched the eligible primary care population it was drawn from.

The study, described in BMC Medical Research Methodology, enrolled adults 65 years or older with inadequately controlled hypertension from primary care clinics within two large health systems. Rather than relying on the usual funnel of clinician referrals and in-person screening visits, the investigators prespecified a set of strategies intended to widen reach: broad inclusion criteria, automated identification of potentially eligible patients, an opt-out approach to initial contact, low study burden, phone-based study visits, and stakeholder engagement throughout the process.

To judge whether those choices worked, the researchers performed a cross-sectional analysis of enrollment data and compared the characteristics of enrolled patients with those of all eligible primary care patients who met criteria for guideline-directed blood pressure lowering. The comparison covered sex, Black race, Charlson comorbidity index, diabetes, coronary artery disease, stroke history, age, and estimated glomerular filtration rate, giving a structured picture of where the trial cohort and the source population diverged and where they aligned.

The reach numbers show how much of the eligible population the trial actually touched. Among 64,671 patients 65 years or older who had primary care visits during the enrollment period, 6,006 were potentially eligible, and 1,000 were enrolled and randomized. That works out to a source-population reach of 16.7%, meaning roughly one in six potentially eligible older adults ended up in the trial, a figure the authors present as the product of the automated identification and opt-out contact approach rather than chance.

On representativeness, the results were largely reassuring. Compared with eligible non-enrolled patients, enrolled patients were slightly younger, with a mean age of 73 plus or minus 6 years versus 75 plus or minus 7 years, and had a lower prevalence of stroke, 5.9% versus 8.9%. Those are modest differences, but they are differences nonetheless, and the authors report them plainly rather than claiming perfect equivalence.

Everywhere else the comparison held. The enrolled and eligible non-enrolled groups were similar with respect to sex, Black race, Charlson comorbidity index, diabetes, coronary artery disease, and estimated glomerular filtration rate. In a trial population where comorbidity burden and kidney function often differ sharply between participants and the broader patient pool, that alignment across six measured characteristics is the central technical finding of the analysis.

The design itself is worth unpacking. A hybrid type 1 effectiveness-implementation study puts primary weight on testing whether an intervention works while also collecting information on how it could be implemented in routine care. Pairing that framework with pragmatic enrollment choices, broad criteria that avoid excluding common conditions, automated screening that does not depend on busy clinicians remembering to refer, opt-out contact that lowers the barrier to hearing about the study, and phone visits that remove travel burden, is what the authors argue produced the reach and representativeness they observed.

The trial behind the analysis, registered as NCT05138601 under the name Remote Monitoring and Virtual Collaborative Care for Hypertension Control to Prevent Cognitive Decline Phase II, is testing whether remote monitoring and virtual collaborative care can control blood pressure in older adults, with cognitive decline as an outcome of interest. The enrollment analysis reported here concerns the composition of the cohort, not the intervention’s effects, which the trial has not yet reported.

The authors conclude that logically integrated, prespecified, pragmatic enrollment strategies supported enrollment of a cohort of older adults with hypertension that was similar to the eligible primary care population across key measured characteristics, and they suggest these strategies may serve as a framework for trials seeking to improve external validity. The evidence base for that suggestion is a single multicenter randomized trial in two health systems, so the framework’s portability to other settings, other conditions, and health systems with different electronic record infrastructures remains to be demonstrated.

What the analysis does establish is a measurable benchmark. A trial of an older, comorbid, real-world-adjacent population achieved 16.7% source-population reach and differed from its eligible population on only two of eight compared characteristics, both modestly. For trialists wrestling with the generalizability problem, the study offers a concrete, quantified example of what prespecified pragmatic design choices can and cannot deliver, and it does so with the kind of transparent accounting, eligible counts, enrolled counts, and characteristic-by-characteristic comparisons, that makes the claim checkable rather than rhetorical.

Subject of Research: Pragmatic enrollment strategies to improve reach and representativeness in a hybrid type 1 hypertension trial in older adults

Article Title: Pragmatic strategies to improve reach and representativeness in a hybrid type 1 hypertension trial in older adults

Article References: Gupta, A., Song, X., Young, K., Kumar, D. P., Adomako, E. A., Ramakrishnan, M., Chandaka, S., Barlocker, J., King, J. B., Thompson, J., Liu, T., Conroy, M. B., Supiano, M. A., Mahanken, J., & Burns, J. M. (2026). Pragmatic strategies to improve reach and representativeness in a hybrid type 1 hypertension trial in older adults. BMC Medical Research Methodology. https://doi.org/10.1186/s12874-026-03063-0

Image Credits: AI Generated

DOI: 10.1186/s12874-026-03063-0

Keywords: hypertension, older adults, clinical trial, generalizability, primary care, remote monitoring, implementation science, enrollment bias, source-population reach, hybrid type 1 design, Pragmatic, strategies