To our knowledge, this is the first study to investigate the cascades of care for hypertension and diabetes using population-based data in Lesotho. Our results show insufficient rates of awareness, treatment, and control for both conditions. With regards to hypertension, one in three adults did not know they had hypertension, nor had initiated treatment, and only half of all the adults with hypertension had adequate BP control. The diabetes cascade shows even larger gaps across all the steps. Only half of the adults that screened positive for diabetes had a previous diagnosis or had initiated treatment and glycemic control was achieved by less than half of the participants. Young adults (18–30 years) with hypertension or diabetes had lower odds of treatment initiation and control.
Curbing the morbidity and mortality of hypertension and diabetes in sub-Saharan Africa will require identification of all those who remain undiagnosed, untreated, and uncontrolled27. Routine monitoring of care cascades is a fundamental tool to achieve these objectives in health programs. In the case of diabetes, in 2022 the World Health Organization (WHO) set programmatic targets of 80% for each cascade step28. With regards to hypertension, WHO standard recommendations to monitor cascade indicators were released in 2022 without specific targets29,30. Once established, CVDs programs must attain high rates of BP and glycemic control and integrate comprehensive CVDs risks management. Health systems strengthening approaches are needed to ensure no cascade gaps, including a functional supply chain for drugs and diagnostic commodities, capacity building for task shifted care, and availability of routine BP and HbA1c measurements (including point-of-care devices). Much can be learned from the HIV and tuberculosis programmes31–35.
Our study shows that the performance of both cascades, although better than the regional average27, remains suboptimal in Lesotho. There are significant gaps across all steps, and BP and glycemic control are insufficient to substantially decrease CVD risk in this population. Adults with hypertension appeared to have a higher level of awareness (69.6%, 95%CI 67.2–72.2%), than those with diabetes (48.4%, 95%CI 42.0–55.0%), indicating a more frequent routine measurement of BP than glycemia.
In 2014, the Lesotho Demographic and Health Survey36 reported that 68.5% of the participants who were found to have hypertension were on treatment, and 70.0% of participants who were found to have diabetes had started treatment, representing similar findings to this study. A recent analysis of the continuum of care in Tanzania37 revealed important gaps in both cascades. In this setting only 21% and 11% of participants with hypertension were on treatment and had achieved BP controlled, respectively, whereas 66% and 48% of participants with diabetes had started treatment and had achieved glycemic control, respectively. The authors suggested that these gaps were heavily influenced by out-of-pocket expenses, i.e., patient fees for services. In Lesotho, services for hypertension and diabetes are predominantly free in health centers, and patients are only required to pay a minimal fee when attending hospital services. Our study found no significant association between household wealth and awareness, treatment initiation, or control. The 2012 South Africa National Health and Nutrition Examination Survey38 reported that half of the adults who screened positive for hypertension were undiagnosed, 78% of those with hypertension were not on treatment and, overall, only 9% of participants with hypertension had achieved the threshold of treatment control.
In our study, we did not find a significant association between participants living with HIV and outcomes across the hypertension or diabetes care cascades. A recent study reviewing hypertension and diabetes outcomes along the HIV care cascade in rural South Africa39 found that participants living with HIV had lower systolic BP and blood glucose than participants not living with the virus. Today, a growing body of evidence fosters the idea that HIV treatment programs can successfully integrate, and thus strengthen, care for CVDs in settings with a HIV prevalence such as Lesotho31,39–41. Nonetheless, maintaining the quality of CVD care in the long-term remains programmatically challenging, especially in fragile health systems. This is the case in Mozambique, where a recent report revealed worsening performance of the diabetes cascade, with a decrease of 3% in diagnosis and 50% of treatment initiation between 2005 and 201542.
Improvement in coverage and quality of services for CVDs will require active identification and successful engagement in care of all the adults who live with these conditions43. With regards to this, our study found that participants with hypertension who were male, below 30 years, and lived in urban areas had lower odds of taking treatment or being controlled than women, elder age groups or those in rural areas. These findings are consistent with regards to women and elder groups engaging in care better37,44–46, however it is contrary to previous studies which have suggested that awareness and control tend to be higher in urban areas47,48.
Our study has several limitations. The information for HIV status, and the first step (awareness) in the cascades are collected from self-reported information.49,50 This is the reason why in the diabetes care cascade (Fig. 2B) the proportion of participants being aware or recalling a previous diagnosis is lower than those who reported being on diabetes treatment. Thus, this step needs to be interpreted with caution and we decided not to include this step in the multivariable analyses. Second, as this was a cross-sectional household-based survey, hypertension diagnosis and control relied on single-day measurements. However, this is the standard approach recommended by WHO for household-based surveys, such as STEPwise approach to NCD risk factor surveillance surveys, in these settings8. Third, from approximately one third of the participants we did not have information on their HIV status. This was due to procedural changes during the survey conduct, and there is no reason to believe the survey population was different after the procedural change than before and thus we consider this data being missing completely at random.