Pragmatic Clinical Trial Design and Outcomes in Acute Stroke

Abstract

Background: Pragmatic trials generate generalizable evidence applicable to routine practice, which informs clinical policy decisions rather than simply confirming a physiological or clinical hypothesis. These trials often employ patient-centered outcomes (such as endpoints reflecting pragmatic health care services utilized) rather than purely clinical or biological markers/symptoms. Research Aims: This study examined the uptake of pragmatic trial designs in Phase III acute stroke trials. The second aim examined the association between the type of thrombolysis and the pragmatic health care services utilized in acute ischemic stroke using data from AcT, a multicenter registry-linked acute stroke trial. Methods: A systematic review of published Phase III acute stroke trials was conducted to examine pragmatic uptake. The PRECIS-2 checklist was used to assess the pragmatic elements of the Phase III acute stroke RCT study designs along the pragmatic-explanatory continuum, with prespecified thresholds for the average PRECIS score (≥3) or the total PRECIS score (≥27). For the second aim, data from the AcT trial were used to examine the relationship between the type of thrombolysis received and pragmatic health care utilization outcomes. Administrative, clinical, and demographic data from the Canadian Institute of Health Information, as well as registry-linked secondary data, were used to address this objective. Outcomes included index length of stay, discharge disposition, 90-day and 1-year readmissions, and mortality. Multivariable regression models (logistic for binary outcomes; count/time-to-event models as appropriate) were adjusted for stroke severity (NIHSS), age, sex, morbidity status (none versus one comorbidity versus two or more comorbidities/multimorbidity), and site type (comprehensive versus primary care centres). All covariates were selected a priori based on clinical relevance and prior literature. Model estimates were used to evaluate the direction and magnitude of associations. Results: For Aim #1, of the 136 Phase III acute stroke trials included in the review, 52% of them were classified as pragmatic. The uptake of pragmatic design in acute stroke trials has increased over time, rising from 38% before 2016 to 62% after 2016. For Aim #2, we found that the thrombolytic agent (tenecteplase vs. alteplase) was not associated with greater pragmatic health services utilization. In multivariable-adjusted analyses, there were no significant differences between tenecteplase and alteplase in index length of stay, readmission status, or all-cause mortality at 90 days and 1 year. Age, stroke severity, and multimorbidity were the strongest predictors of hospital utilization across chosen outcomes. Conclusions: Phase III acute stroke trials are increasingly pragmatic, aligning trial design with real-world practice. In a large, registry-linked pragmatic RCT, tenecteplase and alteplase showed comparable downstream pragmatic health services utilization and mortality, providing robust evidence that tenecteplase is a safe and viable alternative to alteplase without increasing long-term healthcare burden, thereby informing recommendations and regulatory approvals in jurisdictions where alteplase remains the standard of care. Additionally, policy and implementation should consider patient risk profiles (e.g., age, severity, and multimorbidity) as influential in pragmatic health services utilization, and should also account for critical factors such as the logistical feasibility, ease of administration, and cost of the thrombolytic agent of choice.