Review of targeting methods in HIV programmes

By Josh Colston

Josh Colston studies demography and public health at the London School of Hygiene and Tropical Medicine. In the past he has worked as lecturer in British culture at the China University of Geosciences in Beijing, for a health and development NGO in India and as operations intern at Action Against Hunger UK.

Action Against Hunger (ACF) recently conducted a review of methods of targeting within HIV programmes. The review is intended to facilitate wider discussion. It is based on the premise that there is currently little agreement regarding criteria to be used to determine a person's eligibility for assistance when mainstreaming HIV/AIDS into food security programming. Four types of targeting approach are identified and discussed.

Targeting based on HIV status (Direct targeting)

Arguments against this approach include the fact that HIV positive status does not automatically denote food insecurity or vulnerability. Conversely, many who are vulnerable or food insecure are not HIV positive. Also, many who are HIV positive are unaware of their status and when testing services are available, refuse these services for fear of the stigma attached. Furthermore, in societies where stigma prevails, singling people out for assistance may cancel out the benefits of inclusion in a programme or lead to increased social disharmony.

A child is nutritionally assessed while attending for HIV testing of both mother and child

There are, however, some advantages to such an approach. It is relatively unambiguous with little room for misinterpretation or exploitation by beneficiaries. Furthermore, it may encourage people to find out and be open about their status and therefore help tackle stigma head on.

From ACF's perspective, the main problem with the approach is that by the time a person is diagnosed with HIV, it's likely to be fairly late on in the disease's progression. An effective targeting tool should be able to select households which are vulnerable to the effects of the pandemic before they get to this late stage. Ideally, targeting should be 'predictive' so as to include households of people who may have been infected with HIV but are as yet unaware, and those of people who are not yet infected but whose circumstances put them at high risk of infection.

Targeting based on proxy indicators and livelihoods (Indirect targeting)

The rationale here is that, because of the strong interconnection between food insecurity and the disease, certain types of household are vulnerable to the effects of both (e.g. households affected by death due to chronic illness, child-, female-, or older person-headed households, etc). This realisation has led many organisations to move their focus away from HIV sero-status towards broader vulnerabilities. This leads to selection of a large section of the population. Agencies like the World Food Programme (WFP) target their HIV/AIDS assistance on the basis of food security status, although they will select geographic areas where prevalence of HIV/AIDS is high. WFP targeting aims to catch families before they become destitute and thus even more vulnerable to the risk of infection.

Proxy indicators can, however, be misunderstood or manipulated. One study found that in Malawi, households described as femaleheaded had an active male of about the same age as the household head. In addition, many people described as orphans were in fact the children of broken marriages rather than the children of dead parents.

Participatory approaches to targeting

This approach involves the community in targeting and may be implemented to create greater transparency and accountability to beneficiaries. One ACF project used this approach in Kasungu district in Malawi in 2005/6. Here, existing health committees were recruited with other community members to form a selection committee and to make a list of households that fitted certain selection criteria. They were told to include households that found it hard to undertake casual labour as they did not have an adult male fit enough to work. The explicit aim was to assist those households that were vulnerable with respect to food security, not solely those affected by HIV/AIDS. Evaluation of the project found that only 14.5% of non-beneficiaries thought the selection process was unfair.

One problem with the community based approach is that committees could be tempted to discriminate against those who they suspect of being HIV positive or there may be other vested interests in other regards where some are favoured over others.

There may also be cultural resistance to ideas like gradations of poverty and asking people to distinguish between the poor and very poor might meet opposition.

Poverty, dependency and resilience

HIV affects productive members of households more than dependents. ACF Malawi devised a discriminating version of the dependency ration, the 'food-labour ratio' in order to target resources (see Field Exchange issue 25¹). The theoretical daily food energy requirements were summed for all members of the household and divided by its labour force with sliding contribution from older family members. This method required extensive interviewing to determine the age and sex of each member of a household. Assets also determine resilience to shocks like HIV/AIDS so in Malawi households were given asset scores, which represented the total value of their assets divided by the number of members. For each household, this second score was subtracted from the food labour score plus ten. Households were then targeted based on their final score.

Evaluation of the method concluded that while the methodology did succeed in selecting a higher proportion of highly dependent and poor families than present in the baseline population, this extra precision was not enough to justify the considerable effort in interviewing and data processing.

Conclusions

It is necessary to distinguish between responding to HIV/AIDS (the clinical symptoms that affect an individual) and responding to the effects of HIV/AIDS at household or community level (e.g. money spent on household health care, decrease in household income, etc). In the case of the former, HIV status would be the appropriate targeting criteria with the intervention geared towards nutritional support of individuals, while for the latter, targeting criteria should make reference to the households resilience as outlined above. In other words, if a household were to suffer the effects of HIV/AIDS, how resilient would they be to these effects? The intervention would be more food security than nutrition focussed.

There is a need to take into account the changing level of vulnerability of the HIVinfected and affected. Vulnerabilities increase as the disease progresses but the rate at which it increases depends on various other factors. Where standards of HIV services and access to a holistic package of care are high, the level of vulnerability will be more stable. Where this is not the case, it becomes even more important to identify vulnerable people as early as possible, before the subsequent sudden increase in vulnerability.

The distinction between those who are HIV positive, aware of their status and showing symptoms of the disease and those who are just at risk, whether actually infected or not, is useful because the needs of the two groups are different and the former group can, in principle be targeted (using sero-status with all the problems entailed). To reach people as early as possible in the disease progression, then the targeting tool needs to have some concept of level of risk built into it, i.e. an individual's vulnerability to HIV infection (based on a questionnaire about lifestyle, work, sexual health awareness, etc) and assign to them a score based on this. This score could then be totalled for the household in which they live and balanced against that household's level of resilience to the potential effects of HIV/AIDS. This approach would, however, be extremely labour intensive and rely rather too heavily on the veracity of individual testimony.

For further information, contact: Josh Colston, ACF, email: joshcolston@googlemail.com

¹Field Exchange 25. May 2005. Targeting vulnerable households within the context of HIV/AIDS in Malawi. p46-p47