Providing a conceptual framework for HIV prevention cascades and assessing feasibility of empirical measurement with data from east Zimbabwe: a case study

Abstract

Background: The HIV treatment cascade illustrates the steps required for successful treatment and is a powerful advocacy and monitoring tool. Similar cascades for people susceptible to infection could improve HIV prevention programming. We aim to show the feasibility of using cascade models to monitor prevention programmes.

Methods: Conceptual prevention cascades are described taking intervention-centric and client-centric perspectives to look at supply, demand, and efficacy of interventions. Data from two rounds of a population-based study in east Zimbabwe are used to derive the values of steps for cascades for voluntary medical male circumcision (VMMC) and for partner reduction or condom use driven by HIV testing and counselling (HTC).

Findings: In 2009 to 2011 the availability of circumcision services was negligible, but by 2012 to 2013 about a third of the population had access. However, where it was available only 12% of eligible men sought to be circumcised leading to an increase in circumcision prevalence from 3·1% to 6·9%. Of uninfected men, 85·3% did not perceive themselves to be at risk of acquiring HIV. The proportions of men and women tested for HIV increased from 27·5% to 56·6% and from 61·1% to 79·6%, respectively, with 30·4% of men tested self-reporting reduced sexual partner numbers and 12·8% reporting increased condom use.

Interpretation: Prevention cascades can be populated to inform HIV prevention programmes. In eastern Zimbabwe programmes need to provide greater access to circumcision services and the design and implementation of associated demand creation activities. Whereas, HTC services need to consider how to increase reductions in partner numbers or increased condom use or should not be considered as contributing to prevention services for the HIV-negative adults.

Funding: Wellcome Trust and Bill & Melinda Gates Foundation.

Figure 1

Generic conceptual HIV prevention cascades applied to the population who would otherwise acquire HIV or are at risk of acquiring HIV (A) Steps to prevention taking the clients' perspective and their perception of risk into account. (B) Steps to prevention taking the intervention perspective and whether there is supply of the product available. (C) Including the possibility that those at risk can move into a not at risk population and vice versa. (D) Combining two prevention cascades where the population at risk for the start of the second intervention is the population that remain at risk after the application of the first intervention.

Figure 2

HIV prevention cascades for voluntary medical male circumcision in sexually experienced uninfected men aged 15–54 years in Manicaland, Zimbabwe (A) Cascade based on service availability for 1888 individuals in 2009–11. (B) Cascade based on service availability for 1476 individuals in 2012–13. (C) Cascade based on participant's perception of personal risk of acquiring HIV infection for 1476 individuals in 2012–13.

Figure 3

HIV prevention cascades for HIV testing and counselling and sexual partner reduction in sexually experienced uninfected adults aged 15–54 years in Manicaland, Zimbabwe (A) Cascade for 1888 men based on service availability in 2009–11. (B) Cascade for 1468 men based on service availability in 2012–13. (C) Cascade for 3793 women based on service availability in 2009–11. (D) Cascade for 2743 women based on service availability in 2012–13. (E) Cascade for 1468 men based on participant's perception of personal risk of acquiring HIV infection in 2012–13. (F) Cascade for 2743 women based on participant's perception of personal risk of acquiring HIV infection in 2012–13.

Figure 4

HIV prevention cascades for HIV testing and counselling and increased condom use in sexually experienced uninfected adults aged 15–54 years in Manicaland, Zimbabwe (A) Cascade for 1888 men based on service availability in 2009–11. (B) Cascade for 1468 men based on service availability in 2012–13. (C) Cascade for 3793 women based on service availability in 2009–11. (D) Cascade for 2743 women based on service availability in 2012–13. (E) Cascade for 1468 men based on participant's perception of personal risk of acquiring HIV infection in 2012–13. (F) Cascade for 2743 women based on participant's perception of personal risk of acquiring HIV infection in 2012–13.