Anticipating implementation: A mixed methods study of future prescribing determinants within an ongoing trial of clinical effectiveness

BACKGROUND: The Department of Veterans Affairs Cooperative Studies Program (CSP) added implementation science teams to select trials to accelerate the uptake of research evidence into practice. CSP study #2008, "Pentoxifylline in Diabetic Kidney Disease", evaluates whether adding pentoxifylline to usual care reduces the incidence of end-stage renal disease and mortality in patients with diabetic kidney disease. This implementation science sub-study aims to develop a VA-wide implementation plan by identifying key implementation outcomes, identifying and interviewing key stakeholders, and identifying critical barriers and facilitators to implementing findings. METHODS: Using an exploratory sequential mixed methods approach in this hybrid type 1 design, we conducted semi-structured interviews with nephrologists, primary care physicians, and pharmacists from ten VA facilities that were enrolling Veterans in the CSP 2008 clinical trial. The interviews were analyzed using the Tailored Implementation in Chronic Disease (TICD) framework. Findings informed a web-based survey distributed to nephrologists, primary care physicians, and pharmacists from 15 additional participating facilities. RESULTS: Interviews identified five facilitators: 1) providers are willing to prescribe an older medication; 2) providers are not concerned about the side effects of pentoxifylline; 3) providers believe that pentoxifylline is inexpensive; 4) providers report widespread comfort with off-label prescribing; and 5) providers state they would be comfortable prescribing pentoxifylline if colleagues they trust recommend it. Survey data confirmed that providers are not concerned with prescribing an older medication, but there were mixed responses about off-label prescribing. Barriers included: 1) concerns about possible drug interactions; 2) providers self-identified as late adopters; 3) concerns about polypharmacy; 4) prescribers prefer newer drugs; and 5) providers want ample evidence that pentoxifylline works for this indication. Survey data confirmed all the identified barriers except for providers identifying as late adopters. CONCLUSIONS: Interview data provided early evidence of facilitators that may support pentoxifylline adoption if the trial shows positive results. Survey data reinforced and broadened these insights across a larger clinician sample. Together, these findings will guide development of targeted implementation strategies to support effective and timely dissemination of trial findings across the VA health system. TRIAL REGISTRATION: The CSP 2008 study ClinicalTrials.gov ID is NCT03625648.