Using e-modules for acquisition of complex diabetes skills in diabetes care providers in Rwanda

Type 1 Diabetes (T1D) is life-threatening without appropriate treatment. Though pediatric endocrinology care is limited in Rwanda, a decentralized health system allows access to local non-communicable disease (NCD) nurses through a network of 42 district hospitals. Recent rapid expansion of internet access in the country makes virtual diabetes education initiatives possible. We investigated whether Rwandan NCD nurses receiving diabetes education via online e-modules could make similar educational gains in insulin adjustment skills (IAS) compared to NCD nurses educated in a conference-style setting, and whether they would maintain equivalent competency at 1 year after education. We randomized 21 district hospitals and their NCD nurses to participate in a 1.5-day educational conference centered around care of type 1 diabetes (Group 1), while nurses from the remaining 21 hospitals (Group 2) received accommodation and access to equivalent educational materials in e-module form. Both groups were requested to review initial course materials at 4, 8, and 12 months. Ten-point IAS assessments were administered before and after education or review at each time point. Groups 1 and 2 had equal improvement after education (+2.0 vs. +2.0, p = 0.47) and equal final score at baseline (6.0 vs. 6.0, p = 0.74). However, both groups showed a diminishing improvement over time, so that any gains were lost by 4 months in Group 1 and 8 months in Group 2. Group 1 showed greater attrition in participation over time (19% vs 58% continued participation at one year, p = 0.002). Groups did not differ in subjective confidence in IAS after education. Both groups identified existing or potential access barriers to their respective educational method. While further modifications should be trialed to ensure equitable access and to maintain long-term engagement, online education is a feasible method to teach complex subspecialty skills to providers working in low-resource settings.