Science - March 9, 2006

Fighting Aids with computers

It sounds a bit strange, using a computer programme to combat the spread of Aids. But administrators in South Africa can use Bas Vanmeulebrouk’s digital maps to help them decide where to invest in sewerage and water supply provision. These are public health measures that can prevent HIV patients from getting Aids.

‘It makes me feel old when I say I’ve been on sabbatical,’ laughs Bas Vanmeulebrouk. The 32-year-old software engineer from the Centre for Geo-Information worked 40 instead of 36 hours for a number of years so that he could take time off to go with his girlfriend to South Africa, where she had a postdoc position. Through the University of Cape Town he came across Cell-Life, an organisation that provides software and data management for Aids clinics.

Much of the work against Aids in South Africa consists of prevention. There are medicines that inhibit Aids, but most people cannot afford them. And without medicines, most people who are HIV positive get Aids sooner or later. Therefore the South African government is trying to extend the time that elapses between HIV infection and developing Aids as much as possible. HIV patients have regular white blood cell-counts to check whether they have developed Aids, as the cells are destroyed by the virus. Half a million South Africans need HIV treatment, which is about one percent of the population.

Townships
A large number of HIV patients live in townships, poor neighbourhoods that arose during the apartheid era. These areas still often lack good housing, sewerage, water supplies and basic health care facilities. But it is precisely these basic provisions that are so important for HIV patients. ‘A person with no access to sewerage or clean drinking water is more likely to become ill,’ says Vanmeulebrouk.

Administrators can use Vanmeulebrouk’s geographical information system (GIS) to determine what provisions an area has. The system is based on data from a census that was carried out in 2001. Users can zoom in on neighbourhoods with an internet browser, and can then input data on aspects such as the distance people have to walk to a drinking water supply, sewerage and health care provision. The data goes into a central computer system, from which digital maps can be generated that give a picture of the extent of basic public health provisions in the various neighbourhoods. If a neighbourhood shows up as red, then something needs to be done quickly. Green means that the facilities are up to standard.

Local administrators can use the maps to back up their arguments for investing in sewerage or drinking water supply when it comes to negotiating with local authorities. And local authorities can also assess the state of basic facilities throughout the whole town, and use this information to decide where and how much to invest. The only disadvantage is that the administrators have to feed the data in themselves, Vanmeulebrouk admits. ‘But if it is to be of real use to them, then they have to fill it in properly.’ Two students will now further develop the GIS.

Privacy
One of the problems Vanmeulebrouk encountered was privacy. At first the idea was that he would work on a GIS that focused far more on the HIV patients. In that system, people who visit HIV patients would be able to use mobile phone text messaging to input patient data into the database. But this did not work out because of the sensitivity of the data.

There was not much money available either. So Vanmeulebrouk worked with open source software, programmes developed cooperatively and for which the source code is available free. Licences for commercial programmes were too expensive and the programme also had to be quickly and easily accessible online. ‘In a country like the Netherlands where broadband is more or less standard, it is difficult to imagine, but there people are still working with ancient, snail-paced modems,’ tells Vanmeulebrouk

His work in South Africa has given Vanmeulebrouk a different perspective on his work in the Netherlands. ‘What I did there was of a very different order from what I do here. At present I’m working on a system that calculates the suitability of grassland for the black-tailed godwit.’

Martin Woestenburg

Re:act