Tracking Risk Compensation Over Time in a National MC Roll-Out in Zimbabwe

描述（由申请人提供）：男性包皮环切术（MC）是目前可在艾滋病毒高流行国家实施的最有效的生物医学艾滋病毒预防策略之一（Auvert 等人，2005 年；Gray 等人，2007 年；Bailey 等人，2007 年）。模型显示，年轻男性人群必须大量摄入 MC，才能对 HIV/艾滋病流行产生最大影响，但这可能会被风险补偿 (RC) 所抵消（Bollinger 等人，2009 年；Gray 等人，2007 年；UNAIDS，2009 年）；也就是说，接受割礼的男性可能会通过增加他们的危险行为来补偿 MC 提供的保护（Templeton，2010）。对肯尼亚、乌干达和南非三个 MC 随机对照试验 (RCT) 队列的随访表明，随着时间的推移，RC 很少。然而，这些随机对照试验中的男性接受了严格的随访（每 3 或 6 个月一次），包括预防咨询、性病治疗，并在每次随访时提供免费避孕套。因此，RCT 队列不太可能很好地表明在 MC 的普通推广中 RC 发生的程度，因为重复强化预防咨询是不可行的。显然需要确定通过在艾滋病毒高流行国家实施的国家 MC 计划，在接受包皮环切术的男性中是否存在风险补偿，并制定预防策略。由于津巴布韦目前正在实施 MC 计划，作为其国家艾滋病毒预防战略的一部分，因此这是进行纵向研究的理想时机和环境，以确定：a）通过该计划进行割礼的男性是否发生 RC，b）自包皮环切以来 RC 的患病率和患病率随时间的变化，以及 c）RC 的行为决定因素。因此，我们建议进行一项队列研究，从那些决定接受包皮环切术的男性中随机选出 1200 名男性，以及由 1200 名决定不接受包皮环切术的男性组成的非等效对照组。我们将对这些男性进行基线（包皮环切术前）以及 6、12、24 和 36 个月的调查，以确定 RC 是否在 MC 愈合后不久发生，或者 RC 是否持续较长时间。综合行为模型（Montano & Kasprzyk，2008）将为设计调查提供理论框架，以衡量 RC 行为的潜在环境（社会文化）和个人（态度、规范、个人代理、性满意度）决定因素。将进行分析以确定接受包皮环切术的男性与未接受包皮环切术的男性相比，风险行为是否存在差异变化，以及 RC 随着时间的推移而变化的模式；并确定最能解释 RC 的特定模型构造和信念。这些结果将指导基于理论的预防信息。该研究将包括津巴布韦两个主要城市哈拉雷和布拉瓦约的男性，这将使我们能够确定 RC 和决定因素是否存在种族（绍纳、马塔贝莱）差异。这项研究将首次提供有关包皮环切国家计划中风险补偿普遍程度的信息，并对这种行为的决定因素进行建模，以便设计有效的风险预防信息。

DESCRIPTION (provided by applicant): Male circumcision (MC) is currently one of the most effective biomedical HIV-prevention strategies that can be implemented in high HIV prevalence countries (Auvert, et al, 2005; Gray, et al., 2007; Bailey, et al., 2007). Modeling shows that high uptake of MC must occur in a young adult male population to have the largest impact on the HIV/AIDS epidemic, but this may be offset by risk compensation (RC) (Bollinger et al., 2009; Gray, et al., 2007; UNAIDS, 2009); that is, men who are circumcised may compensate for the protection MC offers by increasing their risk behavior (Templeton, 2010). Follow-up with the three MC randomized control trial (RCT) cohorts in Kenya, Uganda, and South Africa suggests little RC over time. However, the men in these RCTs were followed intensely (every 3 or 6 months) with prevention counseling, STD treatment, and free condoms offered at each follow-up visit. Thus, the RCT cohorts are unlikely to provide a good indication of the degree to which RC will occur in an ordinary roll out of MC where repeat intensive prevention counseling is not feasible. There is a clear need to determine whether risk compensation occurs among men getting circumcised through national MC programs that are being implemented in high HIV-prevalence countries, and to develop strategies to prevent it. Since Zimbabwe is currently implementing an MC program as part of its national HIV prevention strategy, this is the ideal time and setting to conduct a longitudinal study to determine: a) whether RC occurs among men circumcised through this program, b) the prevalence and change in prevalence in RC over time since circumcision, and c) the behavioral determinants of RC. Thus, we propose to conduct a cohort study where we will accrue 1200 randomly selected men from those who have made a decision to get circumcised and a non-equivalent comparison group of 1200 men who have decided not to get circumcised. We will conduct surveys of these men at baseline (prior to circumcision) and at 6-, 12-, 24-, and 36-months, in order to determine whether RC occurs soon after MC healing or whether RC continues longer term. The Integrated Behavioral Model (Montano & Kasprzyk, 2008) will provide the theoretical framework for designing the survey to measure potential environmental (socio-cultural) and individual (attitudinal, normative, personal agency, sexual satisfaction) determinants of RC behavior. Analyses will be carried out to determine whether there is differential change in risk behavior among men who get circumcised compared with those who do not, and the patterns of change in RC over time; and to identify specific model constructs and beliefs that best explain RC. These results will then govern theory-based prevention messages. The study will include men in Zimbabwe's two main cities, Harare and Bulawayo, which will allow us to determine whether there are ethnic (Shona, Matabele) differences in RC and determinants. This study will be the first to provide information on the prevalence of risk compensation in a national program rolling out circumcision, as well as model the determinants of this behavior in order to design effective risk-prevention messages.