Treating hypertension in rural South Africa: Comparative effectiveness of two different patient outreach models.

Lead Research Organisation: University of Warwick
Department Name: Warwick Medical School


The South African Department of Health is currently reorganising the health system to improve care of chronic conditions. The Department of Health has recognised that chronic conditions, including hypertension, are poorly managed in rural, low-resource areas of South Africa. Hypertension is a very common chronic condition, affecting as much as 46% of the adult population, but many people with hypertension remain undiagnosed, are not taking treatment, or do not take their medication consistently.

Part of the plan of the Department of Health is for a new cadre of Community Health Workers, six per team with one nurse, making visits to households and responsible for around 270 households. In some areas, including the area where this research will take place, they will also be responsible for delivering medication to people's homes, monitoring clinical indicators, encouraging people with chronic conditions, including hypertension, to take their medication. However, there is concern that this model may prove both too impracticable in rural areas because of the large distances, among other things.

We propose to develop and test the effectiveness of a lower resource model in which there is more emphasis on educating and empowering patients to manage their condition. Two lay health workers will be based in a clinic and will support patients with chronic conditions, providing patient education, support to encourage adherence to medication, and texted reminders about clinic appointments and repeat prescriptions. In addition the lay health workers will encourage and support patients to set up self-help groups, which will provide social support but will also provide practical support in fetching medication from the clinics, which are often a long distance from homes.

We will carry out a randomised trial comparing the two models of care in 8 clinics in north eastern South Africa, which is a low resource rural area with widely scattered villages. Because we are basing this research in a Health and Demographic Surveillance System site where an annual census is carried out and there is an extensive database of information on the people living on the site, we will be able to identify the people using the clinics so that we can see patterns of use of the clinics and how they change during the trial. This will also mean that we can examine the effects of the two interventions by factors such as how far the household is from the nearest clinic. The census database will also provide a sampling frame for two surveys which will measure the prevalence and control of hypertension in the community before and after the trial.

During the trial we will also be using in-depth interviews to study patients, health workers and community members perceptions and experience of the interventions and what are the factors that encourage or discourage either attendance at clinics or adherence to medication.

Technical Summary

South Africa has a high prevalence of hypertension, but it is often undiagnosed, or, if diagnosed, untreated. Deficiencies in hypertension management are mostly due to poor functioning of primary care services, which focus on acute conditions. Aware of these problems, South Africa is in the process of re-orientating primary health services to provide integrated care for chronic disease. The intention is to provide a new cadre of community health workers (CHW) delivering outreach integrated chronic care.

There is concern that the proposed model is resource intensive and not sustainable in rural areas. We aim to compare the cost-effectiveness of the proposed model with a lower resource intervention using lay health workers based in the clinics and supporting patient self-help groups. Final plans for the intervention will be developed with local communities.

The research will be conducted across 8 clinical facilities in the Agincourt Health and Demographic Surveillance site, using a cluster randomised trial, where the unit of randomisation is a facility and its catchment population. The intervention will include lay support workers in the clinic to raise awareness, increase patient health literacy, provide adherence counselling and encourage six monthly attendances at the clinic, as well as developing patient self-help groups. Outcomes will be measured at patient-, facility- and population-level (primary outcome: proportion with successfully controlled hypertension).

A cross-disciplinary approach will capture patients' and health care workers' experiences using qualitative methods and exit interviews. This proposal aims to influence local and national policy, providing decision makers and practitioners with information to make an informed choice of an outreach model for providing integrated care, and specifically for improving the awareness, treatment and maintenance of hypertension.

Planned Impact

The economic and societal impacts from the proposed study are likely to be extensive and wide-ranging. Firstly, new evidence will be provided for South African policy makers and practitioners at national, provincial and district level as to which model of primary health care outreach might be more suitable to improve the management of hypertension and other chronic illnesses in a rural setting amongst a poor population. Further, selection of a more cost-effective outreach model (in appropriate contexts) would enhance the financial sustainability of the South African public health sector.

Moreover, the new knowledge as to which strategies contributed towards to (un)successful implementation and why, as well as system barriers and to how to remove such barriers, will, if used, enhance the effectiveness of the South African health services. Furthermore, the collaborative approaches used in the implementation (combining top-down and bottom-up processes) will provide practitioners with experience of contributing positively to policy implementation from the outset, and so strengthen organisational culture and practice in the health facilities.

If shown to be effective, the intervention will improve health and well-being of communities in the study site during the study, and, if rolled out, other communities in South Africa. The findings of the study will also contribute to international knowledge in this field, and may influence health policy (and so its beneficiaries) in other low and middle income countries. Lastly, researchers and research users (Department of Health, involved from project design to end stages) will be able to enhance their research capacity, knowledge and skills through participation in the project.

The beneficiaries of the research, specific to the project, will be
1. Individual policy makers, managers, and practitioners at facility, district, provincial and
national level in South Africa;
2. Organisations such as health clinics, district and provincial departments of health;
3. Individuals suffering from hypertension, diabetes or HIV and their families;
4. Researchers and research users who participate in or engage with the project.

Likely time scales of impact differ according to beneficiary. Researchers and research users participating in the project will benefit from the beginning of the project, as well relevant organisations through their ongoing participation in project implementation. Individuals with hypertension, diabetes or HIV who live in Agincourt are likely to benefit directly through/with the establishment of the intervention. Decision makers (whether policy makers, managers or practitioners) will benefit as the project proceeds through involvement in stakeholder workshops, and particularly when the finding are available. More distant decision makers are likely to benefit when the findings become widely available through press releases, policy briefs, presentations and finally journal articles.
Title Clinic Link Database 
Description As part of the trial, we funded a system of data entry clerks with access to the Agincourt census database on laptops. The clerks were based in all the clinics participating in the trial and collected data on all patients attending the clinics, including information on diagnosis and treatment. These data can be matched with the data from the census. providing, for example, the first information on the age ,sex and diagnosis of clinic attendees, as well as on the volume of clinic activity. This has provided an important source of information for describing the outcomes of the trial and understanding what was happening in the clinics during the trial. The database has been maintained since the end of the trial by other funding. Kabudula CW, Clark BD, Gomez-Olive FX, Tollman S, Menken J, Reniers G. The promise of record linkage for assessing the uptake of health services in resource constrained settings: a pilot study from South Africa. BMC Medical Research Methodology 2014; 14.71 
Type Of Material Improvements to research infrastructure 
Year Produced 2015 
Provided To Others? Yes  
Impact This database has potential but it is too early to describe impacts 
Title census-clinic-link 
Description A unique system has been developed that allows patients attending clinics to be linked to their demographic surveillance record immediately as they attend the clinic. We are building a database which contains both the socio-demographic data from the annual socio-demographic surveillance updates (up to 20 years of individual- and household-level data) and detailed information on all of an individual's visits to the clinic. This database will be similar to, but more detailed and powerful than, the HMO systems in the USA or the general practice databases in the UK. The database will have huge potential both for health service management and for research. 
Type Of Technology Software 
Year Produced 2014 
Impact too early for outputs 
Description Report to Province July 2014 
Form Of Engagement Activity A magazine, newsletter or online publication
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Policymakers/politicians
Results and Impact positive feedback from District and Provincial staff

Year(s) Of Engagement Activity 2014