HIV seroprevalence in northern Uganda: The complex relationship between AIDS and conflict

from Associazione Volontari per il Servizio Internazionale
Published on 15 Sep 2004
Acholiland in Northern Uganda is devastated by an 18-year old conflict, which caused, especially since 1996, a massive displacement, involving at present about 90% of the civil population. The dimension of the conflict has caused severe vulnerability in the population, forced in congested Internally Displaced People's (IDPs) Camps, lacking basic services and exposed to brutal disruption of the economic and social fabric of communities and the families.

This situation is matter for a great concern that the national declines of HIV seroprevalence might be reversed in the conflict affected areas. Several reports in the national and international press and information disseminated by international organizations have indicated that in northern Uganda and especially in Acholiland, HIV prevalence is higher than in the rest of the country1. Most of these reports did not mention the methodology of investigation neither provided the sources of data and information. These allegations raise an ethical issue related to the release of information lacking evidence and scientific substance. Moreover there is need to avoid further stigmatization of the population of northern Uganda and particularly of the formerly abducted children.

This paper is an attempt to provide data about prevalence in the region and to discuss the complex relationship between HIV and conflict.


Uganda experienced declines in HIV prevalence during the 1990s. National prevalence peaked at around 15% in the early 90's and fell to 5.1% by 20032. Among pregnant women in Kampala, prevalence declined from a high of approximately 30% to about 10%, while among pregnant women in other areas it fell from more than 10% to less than 5%. Uganda's marked decline in HIV prevalence remains unique worldwide. In other sub-Saharan African countries with epidemics of comparable severity and longevity, similar declines have yet to occur. Accordingly, Uganda's success has been the subject of intense study and analysis.

It appears that Uganda's decline in HIV prevalence was associated with positive changes in all three ABC behaviors: increased abstinence, including delayed and considerably reduced levels of sexual activity by youth since the late 1980s; increased faithfulness and partner reduction behaviors; and increased condom use by casual partners. The most significant of these appear to be faithfulness or partner reduction behaviors by Ugandan men and women, whose reported casual sex encounters declined by well over 50% between 1989 and 1995. This conclusion is supported by comparisons with other African countries3. Actually in Uganda the ABC strategy has been accompanied and supported by a fierce fight against stigma, promotion of education, particularly girl education, empowerment of women and youth, young girl protection, involvement of religious leaders and PLWHAs. Moreover the latest UNAIDS report4 has readjusted the national seroprevalence to a lower (than the expected 5 to 6%) 4.1% (range 2.8 to 6.6%).

The prevalence of HIV infection among pregnant women attending antenatal clinics in selected sites is the main tool utilized in the surveillance system adopted in Uganda by the Aids Control Program (ACP) of the Ministry of Health5. In Acholiland (northern Uganda districts of Pader, Kitgum and Gulu), the ACP surveillance site is located in Lacor hospital in Gulu. Prevalence in Gulu is between 10 and 12 % of mothers attending antenatal clinics services6. In 1993 prevalence in Lacor was 27.1% and in 2001 11.3%. In 2002 it raised to 11.9%.


Conflicts are creating conditions that increase the risk of contracting infections such as HIV and may lead to their spread. Nevertheless evidences from Sierra Leone, Angola and Bosnia Herzegovina indicate that there is a complex relationship between HIV and conflict. In Sierra Leone at the end of the ten-year civil war in 2002, a study indicated low levels of HIV-related knowledge and high levels of sexual violence. At the same time the study showed much lower levels of HIV infection (1 - 4%) than previously documented during the conflict. In Bosnia Herzegovina, where the conflict raged from 1992 to 1995, at the end of the conflict characterized by heavy displacement and high levels of sexual violence, studies registered very low levels of HIV infection (0.0003% of population in 2001). In Angola when the conflict ended in 2002, the country had substantial lower seroprevalence (between 5-10% in Luanda and 1-3% in rural areas) than neighboring countries (not affected by armed conflicts)8.

These facts confirm that conflicts can cause situations where the epidemic can spread as well to act as a brake on the infection's spread. This is also recognized in the work of the Interagency Standing Committee through its "Guidelines for HIV/AIDS Interventions in Emergency Settings"9.


The Acholi Region of North Uganda, where the districts of Kitgum and Pader are located, has been affected by conflict for nearly two decades. The escalation in intensity of the conflict since June 2002 led to a grave humanitarian crisis, displacing almost the entire population of the two districts, disrupting the delivery of health services and reducing the accessibility of rural areas due to insecurity.

With own funds and other resources from Elizabeth Glaser Pediatric AIDS Foundation, USAID, The Republic of San Marino and other donors, AVSI supported the start up and implementation of the PMTCT program in three hospitals in Kitgum and Pader districts. Between May 2002 and April 2004, a total of 14,570 mothers attended Antenatal Clinic at the three sites (St. Joseph's Hospital and Kitgum Government Hospital in Kitgum District, Dr Ambrosoli Memorial Hospital in Kalongo, Pader District).

Acceptance rates for counseling and testing have been very high, above national levels, with the former ranging between 96.9% and 99.3% and the latter ranging between 98.2% and 99.3%. Prevalence of HIV in Kitgum stands at 9.9% at St. Joseph's Hospital and 7.8% at Kitgum Government Hospital respectively10. In Pader, the observed prevalence of HIV from the rural area of Kalongo stands at a much lower level of 4.6%. When compared with those of the National PMTCT Program, the process indicators from Kitgum and Kalongo stand out for being particularly high. In the year 2003, pre-test counseling rates were 20% points above national ones, and testing rates surpassed those of the national program by more than 40% points.

The service coverage of the target population can be estimated utilizing the figures made available by the recent 2002 Uganda Population Census. The estimated population of Kitgum and Pader districts is 286,000 and 294,000 respectively. From these figures, the total number of mothers that are expected to be pregnant in one year can be estimated using a national ratio of 5%, and the number of those pregnant mothers who attend to antenatal clinic services can be approximated to 90% of this latter figure. Coverage rates of the target populations ranged between 20% and 35% of pregnant mothers, while coverage of the estimated number of HIV-positive mothers (obtained using as parameter the observed prevalence rates for each district) ranged between 18% and 31%.

The implementation of the PMTCT program in Kitgum and Pader districts has provided the first data on HIV prevalence in the population. The very high acceptance rates of HIV test among the Ante Natal Care (ANC) population -constantly above 95% of new ANC clients- in fact allows extrapolation of HIV prevalence from the PMTCT program, to represent prevalence among women in fertile age. The observed prevalence indicates an ambiguity in the relation between conflict and HIV prevalence (which was initially considered in the theoretical debate to be two realities positively correlated) and the existence of a rural-urban gap in prevalence rates of HIV (confirmed in type and size). In fact the results show a trend similar to the national results, with an analogous urban-rural difference. Actually the results (from PMTCT sites) in the "peaceful district" of Hoima have the same pattern of Kitgum and Pader districts, with a prevalence of 8.1% in Hoima town and 4.3% in Kigorobya Sub County, where Kitgum town can be considered an "urban setting", while Kalongo is a much more rural context.

Conflict is conducive to a range of factors that increase the risk of infection with HIV (rape, displacement, poverty, sex commercialization, abuse and coercion), but this is not necessarily translated into higher prevalence rates (as the data from Kalongo hospital, with a prevalence below 5%, indicate).

The sensational announcements on catastrophic prevalence rates of HIV, especially among children returning from abduction by the LRA, that every now and then appear on the media do not contribute to the understanding of the dynamics of the HIV epidemic in northern Uganda. Very often, the information utilized to back up such statements come from HIV prevalence data coming from VCT sites (which is notoriously affected by a selection bias in the population surveyed). Other times, the source is represented by a limited number of formerly abducted children who accept HIV counseling and testing as part of their reintegration services. Besides the already recalled selection bias, these populations are of generally very small size, thereby making it scientifically unhelpful to use such estimates (that as a matter of fact vary greatly from one sample to another). The insistence of the use of such information and its modality raises ethical concerns that greatly offset any concrete gain for the understanding of the problem or to better programming for the conflict-affected population.blem or to better programming for the conflict-affected population.


The HIV seroprevalence declines in Uganda are increasingly attributed to changes in sexual behavior particularly a delay in sexual debut among the youth and a decrease in number of sexual partners among sexually active adults. Some evidence suggests that in addition to the prevention and care interventions commonly supported by National AIDS Control programs throughout the continent, there existed in Uganda positive elements of social cohesion11.

It is suggested that these elements served as catalysts to convert AIDS knowledge and information to personal modification of sexual lifestyles. One element of social cohesion, supported in study findings, is that in Uganda information about AIDS and about persons affected by AIDS is more likely to be transferred through personal communication networks, compared to other countries in Eastern and Southern Africa. Another important facilitating element in this process has been a high-level of government and civil society leadership, to openly communicate the facts about AIDS risk and prevention to the population12. It seems that despite the chronic and complex conflict, these elements of social cohesion have served as catalysts to affect changes in northern Ugandan, which appear consistent with the trends of the rest of the country.


* It is necessary to sustain careful and accurate surveillance systems in the areas affected by the armed conflict, refraining from "ideological conclusions" on seroprevalence in the region, basing instead public health interventions on evidence-based information.

* It is a priority to implement prevention, care and treatment interventions in emergency situations, within the well established and experimented national strategy of the Ugandan Government as reaffirmed in the recent XV International AIDS Conference in Bangkok13.

* PMTCT services provide an important contribution to the enhancement of HIV/AIDS programs in northern Uganda.

* PMTCT improved coverage could be achieved through scaling up the program to additional sites, but necessarily taking into account the specific types of constraints already highlighted. AVSI has partnered with Kitgum Government Hospital to staff and keep open the health centre in Pajimo (Akwang) camp, which serves a population of 11,270. After carrying out specific field-level assessments, AVSI is in the process of supporting the District Health Department in the scaling up of PMTCT sites, and the subsequent extension of services to Akwang, Padibe, Mucwini, Madi Opei and Namokora IDP camps. This will guarantee a marked expansion of program coverage, enhancing the prevention, care, surveillance and treatment in the districts.

* More and well-coordinated operational research is needed to clarify the actual situation of HIV/AIDS epidemic in northern Uganda14. The on-going Ugandan national serosurvey will provide very crucial inputs for a better knowledge of the context and rational and well-guided public health and social preventive and therapeutic actions.


1 a) As Uganda AIDS rate falls, orphan ranks swell - 16 Aug 2004 - By Mercedes Sayagues

b) 50% of freed kids have HIV - By Joyce Namutebi - New Vision Tuesday, 31st August, 2004

c) UN special AIDS envoy in weeklong visit to Uganda - Source: Agence France-Presse - Date: 24 Jul 2004

d) Northern Uganda under siege from AIDS and rebels: UN special AIDS envoy - Source: Agence France-Presse - Date: 31 Jul 2004 -

e) OCHA Regional Support Office for Central and East Africa - MISSION REPORT - UGANDA - 14 to 17 June 2004

2 Adults (15-49 years) rate at the end of 2001 ranged between 3.5% and 7.9% (UNAIDS - 2004 Report on the AIDS Global Epidemic - 4th global report - June 2004).

3 The ABCs of HIV Prevention

4 UNAIDS - 2004 Report on the AIDS Global Epidemic - 4th global report - June 2004.

5 HIV/AIDS Surveillance report (June 2003) - STD/AIDS Control Programme - Ministry of Health - Uganda

6 Nsambya hospital, one of the major surveillance sites in Kampala, the capital city in southern Uganda, reported a prevalence of 8.5 % in 2002. In the same period Lacor hospital, the surveillance site in Gulu reported a seroprevalence of 11.9% - STD/ACP June 2003.

7 Focus - AIDS and conflict: a growing problem worldwide - UNAIDS - 2004 Report on the AIDS Global Epidemic - 4th global report - June 2004, pages 175 to 181.

8 Namibia 21.3% and Zambia 16.5% (without mentioning the catastrophic situation in Botswana and Zimbabwe (though not neighboring Angola).


10 In the period January - June 2004 in the two sites of Kitgum district 2,641 mothers attended PMTCT services with an accepting rate of 99.8% and an HIV prevalence of 7.2 % (in 2003 was 8.0% and in 2002 8.6%). (Preliminary data).

11 Social cohesion is defined as the norms and social relations embedded in the social structures of societies that enable people to coordinate action to achieve desired goals.

12 Stoneburner, RL, Low-Beer, D - Analyses of HIV trend and behavioral data in Uganda, Kenya, and Zambia: prevalence declines in Uganda relate more to reduction in sex partners than condom use. (Abstract ThOrC734). XIII International AIDS Conference, Durban, South Africa July 7-14, 2000 Low-Beer D, Stoneburner R, Whiteside A, Barnett A. - Knowledge diffusion and personalizing risk: key indicators of behavior change in Uganda compared to Southern Africa. (Abstract ThPeD5787) - XII International AIDS Conference, Durban, South Africa. July 7-14, 2000.

13 President Museveni's July 12 speech at the XV IAC in Bangkok, Thailand

Interview: President Yoweri Museveni 14th June 2004

14 M Zucca, F Ciantia, G Azzimonti, L Castelli, C Caracciolo - Challenges of running PMTCT in conflict areas: AVSI's experience in northern Uganda - (Abstract TuPeD5266) - XV International Aids Conference, Bangkok, Thailand. July 11-16, 2004.