We conducted a scoping review to describe published data from cohort and intervention studies that assessed community-based models of care for people living with T2DM in SSA where a relevant part of care was provided outside the facility. To our knowledge, this is the first review that addresses this question. Our literature search strategy yielded four eligible articles out of 5,335 distinct records. The four studies were very heterogeneous in design, type of care model described, and outcomes reported. Quality of evidence provided by the studies as well as the relatively low number of T2DM patients enrolled in the four community models of care make it impossible to conclude on the effectiveness of community-based T2DM care in SSA. As such, this scoping review’s main finding is that there is a considerable evidence gap regarding community-based T2DM care in SSA, and larger scale, well designed studies are needed. Such future studies may build on the experience reported in the four studies included in this review.

The Kgatelopele’s programme, a facility initiated community model of care, reported only on blood glucose levels. After treatment initiation at the facility, significant aspects of patient care were moved to the community for 22 selected patients among whom only two achieved blood sugar control. In contrast, 46 of the 168 (26%) remaining in facility-based care achieved glycemic control. In BIGPIC’s community-initiated model, with 10 T2DM patients, there was only a very modest reduction in fasting blood glucose and HbA1c at six months follow-up. Similarly, the HDM Intervention found no clear indication of better T2DM control in the cohort receiving home delivered medication. These results differ from other reports from LMICs¹ which showed significant reductions in fasting blood glucose and HbA1c measurements in favor of community-based models care where the community-aspect was an add-on to and did not replace standard facility care.

In one of the studies assessed in our full text review but excluded from the scoping review⁴⁰, the intervention supported self-monitoring of blood glucose and relied on six approaches: identification of high risk patients by clinicians and enrollment in the intervention by community health workers; sending patients home with a glucometer and cell phone access; weekly follow-up via phone calls for blood glucose results; glucose results and medication dose summaries generated for clinician to review. Although utilizing an eHealth model for follow-up, the study was not included in this review as majority of participants had type 1 diabetes mellitus. However, average HbA1c was reduced from 13.3% at baseline to 9.1% at 6 months.

Unsurprisingly, BIGPIC’s community model increased linkage to care, performing better than conventional care. Although there was a 70% engagement in care at 9 months in this model, comparison data was not reported for conventional care. However, the high linkage to care with the model is similar to other studies⁴¹. Adopting community care approaches has resulted in better care engagement in various diseases⁴². Our finding thus suggests that adopting community models of care has the potential to improve linkage and engagement in care for diabetes management in SSA. The success seen in linkage to care and engagement in care reported within BIGPIC’s community model possibly resulted from their peer-based approach, in line with reports from other studies.^43,44 Further, the micro-finance component which made funds available and promoted income-generating activities likely contributed to increased linkage and engagement in care.

The HDM intervention and Mobile Diab system’s eHealth model reported high acceptability of both models to both patients and health workers. Although there was no comparison of ‘acceptability’ to conventional care, several other studies report high acceptability of eHealth by end-users^45,46. A recent study evaluating the implementation of home delivery of medication for various illness during the COVID-19 pandemic found that patients and providers preferred the continuance of this approach, with overall improvements in patient adherence to medication⁴⁷. The eHealth model observed a clinically relevant reduction of HbA1c⁴⁸. In a systematic review conducted to determine the effectiveness of telemedicine in the delivery of diabetes care in low- and middle-income countries¹⁰, telemedicine yielded significant reductions in HbA1c, with interventions via telephone and short message service yielding the highest treatment effects¹⁰.

Our scoping review has several limitations. First, as we did not target grey literature, we may have missed some models of care. Second, design, patient recruitment and outcome definition and evaluation differed across included studies making comparison of outcomes impossible. For instance, acceptability was assessed differently between the two studies that reported the outcome. Third, the overall number of T2DM patients enrolled in the four community-based models of care was relatively low. Fourth, lack of comparison groups in some of the included studies made it difficult to fully interpret the findings in the studies. Finally, all studies had substantial flaws in design and/or reporting making conclusions on effectiveness of community-based care for T2DM in SSA impossible.