Visual Abstract
Design of a pragmatic trial requires planning of key components (cohort identification, participant recruitment, consent and enrollment, intervention delivery, and outcomes measurements). The delivery of healthcare was significantly disrupted by COVID-19; this also affected research. We applied health services and outcomes research (HSOR) methods to inform design adaptations of trial components. The trial is an intervention to improve self-care of patients with diabetes mellitus (DM), discharged from the hospital on insulin. Key informant interviews, clinical observations (N=5), patient tracers (N=5), and operational data were gathered from health system stakeholders (N=9), clinicians (N=20), and measurement experts (N=2). Table 1 describes the trial components, HSOR approaches, data sources, and results. A cohort identification algorithm was created using enterprise data warehouse (EDW) data and validated by chart review, and 3 clinical care process maps were produced. Multiple options for each trial component, allowing for flexibility should care disruptions occur, were identified and pretested (N=5). HSOR approaches are effective to rapidly adapt trial design components, even during healthcare delivery disruptions that threaten research. Application of these approaches should be considered as a strategy to design resilient, flexible, and effective pragmatic trials.
P. Wax: None. C. Barnard: None. R. T. Ackermann: None. J. L. Holl: None. A. Wallia: Research Support; Self; Eli Lilly and Company, Novo Nordisk, UnitedHealth Group. R. Khorzad: None. C. M. Yu: None. E. K. Touma: None. A. Rosales: None. G. Prince: None. K. Coyne: None. S. Bailey: Consultant; Self; Luto UK, Pfizer Inc., Sanofi US, University of Westminster (UK), Research Support; Self; Eli Lilly and Company, Gordon and Betty Moore Foundation, Lundbeck, Merck & Co., Inc., National Institutes of Health, Pfizer Inc. T. Pollack: None.
Agency for Healthcare Research and Quality (3R18HS026143-02S1)
