**Background:** The global burden of type 2 diabetes (T2D) is rising rapidly, especially in low- and middle-income countries (LMICs). In many LMICs, health systems struggle to provide adequate care, with data from 40 LMICs showing only 38% of T2D patients on treatment and 23% achieving glycemic control. The SMART2D project aimed to develop and evaluate contextually appropriate self-management strategies for T2D prevention and control in Uganda (low-income, rural) and South Africa (middle-income, semi-urban). This paper reports the effectiveness of integrated facility-plus-community care versus facility care only on T2D outcomes.
**Methods:** A pragmatic cluster randomized trial was conducted. Clusters were primary health facilities and their catchment areas. In Uganda, nine facilities were randomized to three arms (three per arm): usual care, facility care, and integrated care. In South Africa, two facilities were randomized to facility care or integrated care. Participants were adults aged 30–75 years with T2D (diagnosed ≤12 months) or at high risk (defined by fasting plasma glucose 6.1–6.9 mmol/L in Uganda; random glucose ≤11 mmol/L, BMI ≥25, plus ≥1 risk factor in South Africa). Enrollment occurred between January 2017 and November 2018. The facility care arm included organization of care and strengthening patient self-management. The integrated care arm added community mobilization, peer support groups, care companions, and facility-community linkage. Primary outcomes were glycemic control (HbA1c <53 mmol/mol) at 12 months for T2D participants, and reduction in HbA1c ≥3 mmol/mol for high-risk participants. Secondary outcomes included retention into care and incident T2D. Multilevel modified Poisson regression was used, with propensity score weighting to adjust for baseline imbalances. Missing data for primary outcomes were handled via multiple imputation.
**Key Results:** In Uganda, 801 participants were enrolled (424 with T2D, 377 at high risk); in South Africa, 566 participants (281 with T2D, 285 at high risk). Among T2D participants, glycemic control was achieved in 29.2% in Uganda vs. 16.4% in South Africa (p<0.001). Adjusted analysis showed no significant differences between arms, except facility care vs. usual care in Uganda (IRR 0.71 [0.52–0.96]). Among high-risk participants, reduction in HbA1c ≥3 mmol/mol occurred in 36.6% in Uganda vs. 5.6% in South Africa (p<0.001). Crude percentages were higher in integrated care (54.8%) and facility care (37.3%) vs. usual care (17.6%) in Uganda, but adjusted analysis showed no significant differences. Retention into care among T2D participants was 75.0% in Uganda vs. 48.0% in South Africa (p<0.001). In Uganda, adjusted retention was significantly higher in facility care (IRR 1.41 [1.08–1.84]) and integrated care (IRR 1.41 [1.09–1.83]) vs. usual care. In South Africa, integrated care had higher retention than facility care (IRR 1.15 [1.10–1.19]). Among high-risk participants, retention was 57.6% in Uganda vs. 49.5% in South Africa (p=0.039). In Uganda, adjusted retention was significantly higher in facility care (IRR 2.46 [1.33–4.53]) and integrated care (IRR 3.52 [2.13–5.80]) vs. usual care. In South Africa, integrated care had slightly higher retention (IRR 1.02 [1.02–1.03]). Incident T2D occurred in 4.6% overall (4.1% Uganda, 4.2% South Africa), with no significant differences between arms.
**Clinical Implications:** The study demonstrates that even minimal improvements in facility care—such as ensuring medication availability, training nurses, and supporting appointment-keeping—can significantly improve retention into care in low-resource settings. The added value of community-based peer support was more evident for high-risk individuals, suggesting a role for community engagement in T2D prevention. However, the lack of effect on glycemic control and HbA1c reduction, compounded by high and unbalanced loss to follow-up, limits conclusions about clinical efficacy. The findings underscore the need for adequately functioning primary care systems as a foundation for T2D management, with community strategies potentially enhancing prevention efforts. In South Africa, implementation challenges (e.g., low CHW capacity, staff turnover, security concerns) may have diluted intervention effects. The study highlights that in settings with very low baseline care quality, basic improvements can yield substantial retention benefits, but more research is needed to identify strategies that also improve glycemic outcomes.