HIV prevention in Uganda through 20 years: A social ecological analysis using mixed methods

Introduction: In the late 1990s and early 2000s, Uganda was acknowledged as the first country to reverse a rising HIV epidemic. After years of decline, Uganda’s HIV prevalence has mostly stagnated, but the reasons for this stagnation are less researched. This study contributes to narrowing this knowledge gap, using concepts from sociological theory, behavioural sociology, and sociology of health and illness. It explored perceptions about Uganda’s unsustained HIV prevalence decline, analysed changes in ABC Plus prevention behaviours (abstinence, faithfulness, condom use and male circumcision) and predictors for those behaviours, and elicited suggestions for improving Uganda’s HIV prevention results.

Methodology: Inference-based convergent mixed methods, grounded in social ecology theory, were used. Quantitative method investigated whether and how HIV prevention behaviours changed between 2000/1 and 2016 and factors associated with prevention behaviours. Data from four Uganda Demographic and Health Surveys (2000/1 to 2016) were used. Proportions of the population practising each prevention behaviour were estimated using logistic regression and tested for statistical difference across survey years using two-tailed Fisher’s exact test. Factors associated with prevention behaviours were determined using generalised linear mixed effects models, which were estimated using Laplace approximation for maximum likelihood and compared using Akaike Information Criterion. Qualitative method used focus groups and key informant interviews to explore perceptions about Uganda’s earlier HIV decline, changes in and influences of prevention strategies, and how prevention outcomes could be improved.

Findings: ABC behaviours within a supportive policy environment were the cornerstone of Uganda’s early HIV decline. Knowledge about HIV, its transmission, and ABC Plus behaviours (including male circumcision) has increased, but high-risk practices are re-emerging, suggesting a limited impact of HIV knowledge on prevention behaviours today. Risk behaviours are concentrated among widowed, divorced, and separated individuals; sexually active young adults; and married men. Gaps exist in abstinence and faithfulness, and unprotected sex is common, especially during high-risk encounters. Condom use has hardly increased since 2000/1, despite improved availability. Male circumcision increased significantly since 2006, especially in urban locations and ethnic groups that traditionally circumcise males.

At socio-structural levels, decreased HIV stigma, improved treatment availability, and treatment access may be linked to HIV prevention complacency. Poverty among females exacerbates vulnerability to HIV infection. Unbalanced investments in biomedical versus social science initiatives, weakened local-response leadership, and frequent policy and guideline changes have also contributed to Uganda’s unsustained HIV decline. Prevention programmes can improve results by reviving adherence to ABC Plus behaviours, streamlining and empowering local leadership, improving institutional coordination, and implementing more initiatives to address socio-cultural and poverty-related transmission drivers.

Discussion and Conclusions: A potential HIV resurgence and its devastating effects is imminent in Uganda unless prevention is further improved. This improvement could be realised by addressing socio-cultural factors and gender inequality, which perpetuate HIV risk behaviours. Strategies for improving condom acceptability and use are critical for HIV prevention, particularly in women, men having extramarital relations, commercial sex workers, and young people with a likely or recent sexual debut. Prevention programmes should also address high-risk behaviours of widowed, divorced, and separated men and women.