Psychoeducation for adults with type 1 diabetes and problematic hypoglycemia: implementation analysis of a clinical trial (HARPdoc).

Background: The Hypoglycemia Awareness Restoration Program for people with type 1 diabetes and problematic hypoglycemia with severe episodes persisting despite optimal care (HARPdoc) uniquely focusses on addressing cognitive and motivational barriers to hypoglycemia avoidance associated with impaired awareness to hypoglycemia. We aimed to compare perceptions of acceptability, feasibility, and appropriateness of HARPdoc intervention to an existing program, Blood Glucose Awareness Training (BGAT) and understand how these implementation outcomes relate to cognitive and mental health clinical outcomes. Methods: The HARPdoc trial was a hybrid randomized clinical trial delivered in the United Kingdom and United States between July 2018 and December 2019. Implementation outcomes, including perceived acceptability, appropriateness, and feasibility, were measured using published validated surveys. These surveys were completed by the people with diabetes, healthcare professionals, and relatives of participants. Clinical outcomes, including attitudes to awareness, diabetes distress, anxiety, and depression, were measured using validated self-reported questionnaires. We explored differences of perceived implementation outcomes between HARPdoc and BGAT and associations between implementation and clinical outcomes using quantile and linear regression. We also assessed whether the effect of HARPdoc on cognitive and mental health outcomes were mediated by implementation outcomes. Results: HARPdoc was perceived as more appropriate than BGAT at 12 months, with a median difference of 0.75 (95% CI 0,26,1,24) by both those involved in delivering the programs and the HARPdoc participants. All stakeholder groups also perceived HARPdoc intervention as more acceptable (MD 0.50 95% CI 0.13, 0.87), but as feasible (MD 0.00; 95% CI -0.31, 0.31) as BGAT. Each of perceived acceptability, appropriateness, and feasibility were significantly linked to improvements in clinical outcomes (feasibility- anxiety: Mean Difference: -1.07; 95% CI: -2.03, -0.10); feasibility-depression (MD: -5.25, 95% CI -9.09, -1.41)). No evidence of mediation was observed. Conclusions: HARPdoc compared to BGAT was perceived as more appropriate and acceptable and subsequently higher perceived appropriateness, acceptability and feasibility was linked to better cognitive and mental health outcomes. Our findings provide important insights for the development of an implementation blueprint and the expansion of HARPdoc and BGAT programs into routine healthcare services and highlight the need for larger, better-powered hybrid trials.