Population-level impact of Zimbabwe's National Behavioural Change Programme

To assess the impact of Zimbabwe's National Behavioural Change Programme (NBCP) on biological and behavioral outcomes.

Representative household biobehavioral surveys of 18- to 44-year-olds were conducted in randomly selected enumeration areas in 2007 and 2011 to 2012. We examined program impact on HIV prevalence among young women, nonregular partnerships, condom use with nonregular partners, and HIV testing, distinguishing between highly exposed and low-exposed communities and individuals. We conducted (1) difference-in-differences analyses with communities as unit of analysis and (2) analyses of key outcomes by individual-level program exposure.

Four thousand seven hundred seventy-six people were recruited in 2007 and 10,059 in 2011 to 2012. We found high exposure to NBCP in 2011. Prevalence of HIV and reported risky behaviors declined between 2007 and 2011. Community-level analyses showed a smaller decline in HIV prevalence among young women in highly exposed areas (11.0%-10.1%) than low-exposed areas (16.9%-10.3%, P = 0.078). Among young men, uptake of nonregular partners declined more in highly exposed areas (25%-16.8%) than low-exposed areas (21.9%-20.7%, P = 0.055) and HIV testing increased (27.2%-46.1% vs. 31.0%-34.4%, P = 0.004). Individual-level analyses showed higher reported condom use with nonregular partners among highly exposed young women (53% vs. 21% of unexposed counterparts, P = 0.037).

We conducted the first impact evaluation of a NBCP and found positive effects of program exposure on key behaviors among certain gender and age groups. HIV prevalence among young women declined but could not be attributed to program exposure. These findings suggest substantial program effects regarding demand creation and justify program expansion.