Wilson DP, Blower S. How far will we need to go to reach HIV-infected people in rural South Africa? BMC Med 2007 Jun 19;5:16.
To estimate, through modeling, the accessibility of ART in rural areas in KwaZulu-Natal, based on how far persons would have to travel to reach health care centers.
The study uses a mathematical model to evaluate accessibility of ART treatment.
TKwaZulu-Natal (KZN). The South African government has outlined an operational plan for ART rollout in KZN, using 17 currently available health care facilities (HCF), out of a total of 54. These facilities are a subset of provincial hospitals, and do not include community health centers, mobile clinics, or other primary health care clinics.
The study was modeled for persons living with HIV/AIDS (PLWHA) (estimated n=110,000) in rural or peri-urban KwaZulu-Natal and served by the 17 HCF. The populations of the three largest urban areas (Durban, Pietermaritzburg, and Newcastle) were excluded.
The model was based on determining the "effective demand," or the proportion of people who could access ART, weighted according to travel distance. Accessibility was estimated for various distances from a HCF, from 0 to 80 km. The model assumes that accessibility decreases as distance from an HCF increases, such that treatment access is reduced to 1% at the periphery. The distance-discounting measure used was based upon a modified form of a two-dimensional gravity-type model (Gaussian distribution function).
The primary outcome was treatment accessibility, as a function of distance from HCFs, and the number of facilities.
Based on the model, the authors found a nonlinear relationship between treatment accessibility and catchment size. This means that substantially increasing the size of a catchment area (e.g. from 1 km to 20 km) would only increase treatment accessibility by a few percent. The model estimates that the proportion of PLWHA who can receive ART in rural areas would only reach 50%, even if the travel distance was 50 km. Increasing the number of available HCFs for ART distribution by threefold (to the total of 54) would not lead to a proportionate increase in treatment accessibility. This is because most of the additional 37 HCFs are more rural and serve a smaller population.
The authors conclude that many PLWHA in rural KwaZulu-Natal are unlikely to have access to ART, and that the impact of an additional 37 HCFs on treatment accessibility would be less than expected.
There is no standard quality scale for mathematical model studies. Estimates developed by this model may be biased downward by the assumption that only 1% of PLWHA would access ART at the periphery of a catchment area, regardless of the size of that area. Empirical data on distances that PLWHA would be willing and able to travel were not available.
Several resource-limited settings have developed systems to address the need for improved access to ART in rural sub-Saharan Africa.(1,2,3,4,5) Potential solutions include distributing ART through community-level facilities, using mobile health workers to provide home-based care and follow-up, and providing transportation reimbursement. The results of this study demonstrate the use of mathematical models to predict the impact of increasing the number of health care facilities providing services. This model cannot be applied to other settings, however, without modification of the input variables, which need to be individualized to a country's population distribution, mix of health care facilities, and the type of care each facility can provide. In particular, this model for KZN evaluates ART access only through district hospitals, rather than community health care centers or smaller clinics.
The problem of how to provide HIV/AIDS care and ART management to rural populations needs to be addressed if ART rollout is to be successful. Transportation and roads are often poor, preventing PLWHA from accessing facilities, medication, and follow-up. The model estimates in this paper present a rather gloomy forecast of the ability to access ART in KZN using the current distribution system, or even all 54 health care facilities. However, the estimates may be low because the model assumes that only 1% of persons at the periphery of any catchment area, regardless of size, would access ART. This may not actually be the case. For example, if everyone within a 10Km radius were willing to travel this distance to obtain care, there could be close to 100% coverage within this area, rather than a non-linear decrease to 1% at a 10km distance. Thus, before this or similar models are used to estimate coverage and placement of health care facilities, it would be important to collect empirical data on how far PLWHA are able and willing to travel. Nevertheless, it is likely that a more creative mix of delivery systems will be needed. Tertiary and district hospitals might be best able to provide periodic comprehensive evaluation, while the bulk of ongoing monitoring and medication distribution could take place at a community level or through mobile health care workers. Improved roads and transportation would be an important investment to improve the economies and health of communities, both of which are interdependent.
- Curran, et al. Scaling up treatment for the global AIDS pandemic: challenges and opportunities. BGH 2005. (No abstract available.)
- Rosen S, Sanne I, Collier A, Simon JL. Hard choices: rationing antiretroviral therapy for HIV/AIDS in Africa. Lancet 2005 Jan 22-28;365(9456):354-6. (No abstract available.)
- Médecins Sans Frontières. Surmounting challenges: procurement of antiretroviral medicines in low- and middle-income countries-the experience of Médecins Sans Frontières, MSF/WHO/UNAIDS Secretariat, 2003, 56 pp.
- WHO: Access to antiretroviral treatment and care: the experience of the HIV Equity Initiative, Cage, Haiti 2003. (No abstract available.)
- Katzenstein D, Laga M, Moatti JP. The evaluation of the HIV/AIDS drug access initiatives in Cote d'Ivoire, Senegal and Uganda: how access to antiretroviral treatment can become feasible in Africa. AIDS 2003 Jul;17 Suppl 3:S1-4. (No abstract available.)
- Viboud C, Bjornstad ON, Smith DL, Simonsen L, Miller MA, Grenfell BT. Synchrony, waves, and spatial hierarchies in the spread of influenza. Science 2006 Apr 21;312(5772):447-51.
- Savill NJ, Shaw DJ, Deardon R, Tildesley MJ, Keeling MJ, Woolhouse ME, et al. Topographic determinants of foot and mouth disease transmission in the UK 2001 epidemic. BMC Vet Res 2006 Jan 16;2:3.