Decline in reported AIDS cases in Brazil after implementation of the test and treat initiative

BACKGROUND: To evaluate the trends in AIDS detection and the AIDS-attributed death rate in Brazil between 2007 and 2015 and to determine the population characteristics associated with AIDS detection. METHODS: Surveillance data including demographics, mode of HIV transmission, AIDS cases, and AIDS-attributed deaths were analysed. A logistic regression model was fitted to assess the trends in AIDS cases by mode of HIV transmission throughout the study period. A segmented Poisson regression model was used to assess changes in the trends of AIDS detection and mortality rates before (2007–2014) and after (2015) the introduction of the Test and Treat Initiative. RESULTS: In the segmented Poisson regression, the decline in the AIDS detection rate from 2007 to 2014 was 2.0% annually (i.e., the odds ratio (OR) for annual decline was 0.98 (95% [CI: 0.98–1.00, p-value < 0.001). The observed AIDS detection rate in 2015 was 7.7 cases per 100,000, which was 60% lower than the regression estimate of 19.8 for the hypothetical absence of the Test and Treat Initiative. The reduction in AIDS-attributed deaths from 2007 to 2014 was 8.0% annually (i.e., the OR for annual decline was 0.92 (95% [CI: 0.91–0.95, p-value < 0.001). The observed AIDS mortality rate in 2015 was 0.49 cases per 100,000, which was 73% lower than the regression estimate of 2.1 for the hypothetical absence of the Test and Treat Initiative in 2015. CONCLUSIONS: Our results reveal a fairly stable AIDS detection rate from 2007 to 2014, with a sudden significant drop in 2015. This drop may be related to the increasing trends in rapid testing, the number of new HIV diagnoses, the number of patients on antiretroviral therapy (ART), and a median CD4 count at ART enrolment. Further evaluation of the effects of the Test and Treat Initiative on AIDS diagnosis and mortality is needed and must be strengthened with additional data from subsequent years.