Spotlight on SHM data: Mapping HIV prevalence in the Netherlands using geographic information systems
Could you summarise the main findings of your research?
We have identified which municipal areas and districts in large Dutch cities have a high prevalence of HIV, with high prevalence being defined as 2 or more people living with HIV per 1,000 inhabitants (15-60 years of age). Greater insight into these hotspots should make it possible to carry out more targeted detection of HIV, for example by encouraging general practitioners (GPs) in high-prevalence districts to carry out HIV tests more routinely.
The geographical information system (GIS) analyses that we carried out for this study showed that there are ten municipal areas in the Netherlands with an HIV prevalence greater than 2/1000, including Amsterdam (8.1), Rotterdam (3.4) and The Hague (2.7). Within the cities we also saw large variation between the districts, particularly in Amsterdam. Here, HIV was highly concentrated in 2 districts (Central Amsterdam and Southeast Amsterdam), whereas in Rotterdam and The Hague the inter-district differences were less pronounced.
Why did you undertake this research?
Although the number of new HIV diagnoses has been decreasing for a number of years now in the Netherlands, 44% of people living with HIV still enter care at a late stage of infection. We therefore need even greater efforts to identify and treat HIV-positive individuals, which would also help to halt further transmission. We investigated how this could be achieved more efficiently and set out to identify the hotspots in the Netherlands. Worldwide, the HIV epidemic seems to be primarily concentrated in the large cities and this has led to a local approach to reducing HIV, known as the city approach. The first step in such an approach is mapping the high-prevalence districts. In those districts with a prevalence above 2/1000 it can be effective ,and even cost-effective, to employ targeted interventions, such as having general practitioners (GPs) proactively offer HIV testing to patients newly-registering at the practice or to patients who are having blood samples taken for other reasons. The benefit of this approach is that it avoids the need to link the testing to high-risk behaviour or having a migrant background, for example. This makes it easier for the GP to offer an HIV test. Moreover, offering HIV testing on a broader basis is necessary because not everyone who is at risk of HIV seeks HIV testing of their own accord. In addition, a proactive district-based approach through healthcare/primary healthcare may make the difference in reducing the number of late HIV diagnoses. It is now up to the GPs to act on this recommendation.
Could you tell us more about the method you used?
Stichting HIV Monitoring provided the numbers of people living with HIV and in care per municipal area and per district (postal code area) for a number of large cities. Because the SHM database has nationwide coverage, we were able to combine the data with those from Statistics Netherlands on the number of inhabitants per municipality and per postal code area. This allowed us to calculate the HIV prevalences. We then imported this information into software that’s used to create geographical maps, known as ArcGIS. The maps show HIV prevalence by colour, with a darker colour indicating higher HIV prevalence. As a proxy for HIV incidence, we repeated the analyses for the number of new HIV diagnoses in the last 5 years and this yielded more or less the same picture.
What are your plans for further research?
We still need to carry out a lot of additional research. For example, at the implementation level we need to investigate whether more proactive testing is actually feasible for GPs. At the moment, there’s a pilot project running in Amsterdam to provide GPs with interactive training on their diagnostic work-up behaviour for HIV and sexually transmitted diseases (STD), known as the Diagnostic Test Consultation (DTC). During this consultation, GPs receive benchmarking information and feedback on their STD and HIV laboratory workup requests. They subsequently work together to draw up a quality improvement plan based on existing guidelines. Another issue that requires further investigation in terms of feasibility and cost effectiveness is the integration of HIV and hepatitis testing into other blood tests or during a general health check. At present, quality improvement in chronic diseases is an important focus in general practice. However, it remains a challenge for the less common diseases where, although early diagnosis has been proven to be beneficial, a GP may need to carry out up to 1,000 tests just to detect one or two cases.
Combining HIV diagnoses with information about the number of HIV tests at GP laboratories and sexual health centres (SHC) could also provide considerable insight into those districts that require a broader testing strategy. In addition, more in-depth GIS analyses are possible, based on information such as age, gender and migration background. Finally, as part of the H-TEAM initiative (‘HIV Transmission Elimination Amsterdam’), GIS research is now being carried out using testing data from GPs and the SHCs. Various organisations are involved in this project, including AHTI (Amsterdam Health and Technology Institute), the AMC and Stichting HIV Monitoring.
Reference:
Mapping HIV prevalence in the Netherlands with geographic information systems [article in Dutch]
Op de Coul ELM, Joore IK, van Sighem A, Bom BCJ, Hillebregt M, Prins JM, Geerlings SE, van Bergen JEAM. Ned Tijdschr Geneeskd. 2017;161(0):D965.