Political and socioeconomic instability

Five years ago, we reported a low and stable HIV prevalence in selected population groups, despite the poor environment and political instability, in the Democratic Republic of Congo (DRC) [1]. Since then, the central government has lost control of almost one third of the country and war is waging across DRC [2,3]. We hereby report to what extent the ongoing conflict could impact on the stability of HIV prevalence reported in our previous study. From March to May 2002, we conducted HIV serosurveys in four major cities: Kinshasa, Mbuji-Mayi, Lubumbashi, located in the government-held areas, and Kisangani, under the control of the rebels, Rwandan and Ugandan armies. Convenient sampling was obtained among pregnant women attending antenatal clinics, tuberculosis patients, clinic outpatients, blood donors, sexually transmitted infection (STI) patients, sex workers and internally displaced people (IDP). After informed consent, 3090 individuals, 2201 women (72.1%) and 889 men (28.8%), were screened for HIV antibody by a rapid assay (Abbott Determine HIV-1/2; Abbott Laboratory, Tokyo, Japan) or a commercial enzyme-linked immunosorbent assay (Organon Technika, Belgium). All reactive specimens were discriminated and confirmed by a line immunoassay (Innolia HIV-1/2; Innogenetics, Ghent, Belgium). Comparing our data with those obtained in the previous study (Table 1), we found that in Kinshasa, the prevalence rates in pregnant women, tuberculosis patients, sex workers, STI patients and blood donors had not changed significantly. Similar trends were observed among pregnant women and tuberculosis patients in Mbuji-Mayi. However, HIV prevalence among outpatients in Kinshasa has decreased significantly (χ2 = 6.3, P = 0.01).Table 1: HIV prevalence per population groups and per city and comparison with the 1997 data in Kinshasa and Mbuji-Mayi.The reasons for these results remain unclear, but one could assume that the pattern of healthcare-seeking behaviour may have changed as a result of increased poverty and insecurity. Impoverished individuals rarely find the money needed to pay for health services. The pattern of service delivery has also changed: medical services have been eroded over decades of mismanagement and in many communities they collapsed completely during the war. These trends could also be associated with the change in risk behaviour caused by a limitation of movements and a decrease in social life. The purchasing power of Congolese civil servants has declined steadily, and public sector wages are not paid regularly [2,4], forcing men to change their behaviour by limiting their access to commercial sex workers and other partners. When stratified by age groups, our data show some decline in HIV prevalence rates (although not statistically significant) in those under 20 years: pregnant women, from 3% in 1997 to 1.7% in 2002; outpatients from 37.5% in 1997 to 16% in 2002; and STI patients from 20% in 1997 to 7.4% in 2002. Recent infections among young people (15–24 years) could have decreased over time, maybe suggesting a change in risk behaviour [5,6]. With respect to sex issues, data from tuberculosis patients showed that women under 25 years constituted 70% of HIV infections (χ2 = 7.5, P = 0.005). Although the rape of women and girls had been reported in Kisangani [7], in our study, the rate of HIV infection in pregnant women was 3.6%, lower than the 4.5% reported in 1997 [8]. In addition, an HIV prevalence of 3.4% among pregnant women in Lubumbashi contrasts with the level of 9% reported in the 1999 sentinel survey [8]. However, the high HIV prevalence observed among IDP (7.1%) is in line with most other findings, showing that collective violence and consequent population movements are likely to increase the vulnerability of this population to the further spread of HIV [9,10]. As reported by Leroy et al. [11], the prevalence of HIV-1 in pregnant women from rural areas in Kigali, Rwanda, was higher than expected (24%) and was attributed to rape and displacement during the genocide. Although our research included two sites used in the 1997 study, no clear trend emerged regarding the HIV epidemic in these population groups. The low and stable HIV prevalence observed in urban area may hide a high and growing problem in vulnerable groups [12,13]. The selection bias, which excluded rural populations who do not have access to healthcare services, could explain the limitations of the study in assessing the overall situation [14,15]. However, the high prevalence of HIV among IDP shows that population groups (used in sentinel surveys), and traditional methods are not appropriate to evaluate the effect of insecurity and political instability on the HIV situation in DRC. A comprehensive study that includes other vulnerable groups should be carried out at national level to determine the reasons behind the low and stable seroprevalence. Persisting insecurity, military movement and civilian displacement seem to be linked to the high seroprevalence observed in some population groups. National authorities are urged to establish an enabling environment for the implementation of effective and sustainable HIV/AIDS prevention and care programmes. Acknowledgements The authors express their gratitude to the Ministry of Health, the National AIDS/STI Committee and the National Ethical Committee for permission to perform this study. The authors also wish to thank the directors and staff of the laboratory in different regions, especially Antoinette Mayamba, Gisèle Bwalungu, Olivier Disasi, Adolphine Kalume, Mama Nicaise, Rigobert Kambembo, for providing logistic and technical support in the field and for their kind cooperation.