Challenges and opportunities in evaluating programmes incorporating human-centred design: lessons learnt from the evaluation of Adolescents 360

Adolescents 360 (A360) is a four-year initiative (2016-2020) to increase 15-19-year-old girls' use of modern contraception in Nigeria, Ethiopia and Tanzania. The innovative A360 approach is led by human-centred design (HCD), combined with social marketing, developmental neuroscience, public health, sociocultural anthropology and youth engagement 'lenses', and aims to create context-specific, youth-driven solutions that respond to the needs of adolescent girls. The A360 external evaluation includes a process evaluation, quasi-experimental outcome evaluation, and a cost-effectiveness study. We reflect on evaluation opportunities and challenges associated with measuring the application and impact of this novel HCD-led design approach. For the process evaluation, participant observations were key to capturing the depth of the fast-paced, highly-iterative HCD process, and to understand decision-making within the design process. The evaluation team had to be flexible and align closely with the work plan of the implementers. The HCD process meant that key information such as intervention components, settings, and eligible populations were unclear and changed over outcome evaluation and cost-effectiveness protocol development. This resulted in a more time-consuming and resource-intensive study design process. As much time and resources went into the creation of a new design approach, separating one-off "creation" costs versus those costs associated with actually implementing the programme was challenging. Opportunities included the potential to inform programmatic decision-making in real-time to ensure that interventions adequately met the contextualized needs in targeted areas. Robust evaluation of interventions designed using HCD, a promising and increasingly popular approach, is warranted yet challenging. Future HCD-based initiatives should consider a phased evaluation, focusing initially on programme theory refinement and process evaluation, and then, when the intervention program details are clearer, following with outcome evaluation and cost-effectiveness analysis. A phased approach would delay the availability of evaluation findings but would allow for a more appropriate and tailored evaluation design.

How Can Programs Better Support Female Sex Workers to Avoid HIV Infection in Zimbabwe? A Prevention Cascade Analysis

BACKGROUND

"HIV prevention cascades" have been proposed to support programs by identifying gaps in demand for, access to, and capability to adhere to HIV prevention tools, but there are few empirical examples to guide development. We apply a prevention cascade framework to examine prevention coverage and factors associated with condoms and/or PrEP adherence among female sex workers.

SETTING

Seven sites across Zimbabwe.

METHODS

Seven respondent-driven sampling surveys from the intervention sites of a pragmatic cluster-randomized trial in Zimbabwe in 2016 were analyzed, and 611/1439 women testing HIV-negative included. We operationalized key components of an HIV prevention cascade including demand, supply, and capability to adhere to 2 tools for HIV prevention: condoms and pre-exposure prophylaxis (PrEP). We used adjusted logistic regression to identify determinants of adherence to condoms and PrEP in turn, examining the effect of adherence to one tool on adherence to the other.

RESULTS

There were 343/611, 54.7%, women reporting adherence to condoms and/or PrEP, leaving almost half uncovered. Although women were aware that condoms prevented HIV and reported good access to them, only 45·5% reported full adherence to condom use. For PrEP, a new technology, there were gaps along all 3 domains of demand, supply, and adherence. Alcohol use decreased adherence to PrEP and condoms. Younger and newer entrants to sex work were less likely to take PrEP every day.

CONCLUSIONS

HIV prevention programming among female sex workers in Zimbabwe could consider increasing awareness of PrEP alongside supply, alcohol use interventions, and approaches to engaging younger women.

Applying user preferences to optimize the contribution of HIV self-testing to reaching the "first 90" target of UNAIDS Fast-track strategy: results from discrete choice experiments in Zimbabwe

INTRODUCTION

New HIV testing strategies are needed to reach the United Nations' 90-90-90 target. HIV self-testing (HIVST) can increase uptake, but users' perspectives on optimal models of distribution and post-test services are uncertain. We used discrete choice experiments (DCEs) to explore the impact of service characteristics on uptake along the testing cascade.

METHODS

DCEs are a quantitative survey method that present respondents with repeated choices between packages of service characteristics, and estimate relative strengths of preferences for service characteristics. From June to October 2016, we embedded DCEs within a population-based survey following door-to-door HIVST distribution by community volunteers in two rural Zimbabwean districts: one DCE addressed HIVST distribution preferences; and the other preferences for linkage to confirmatory testing (LCT) following self-testing. Using preference coefficients/utilities, we identified key drivers of uptake for each service and simulated the effect of changes of outreach and static/public clinics' characteristics on LCT.

RESULTS

Distribution and LCT DCEs surveyed 296/329 (90.0%) and 496/594 (83.5%) participants; 81.8% and 84.9% had ever-tested, respectively. The strongest distribution preferences were for: (1) free kits - a $1 increase in the kit price was associated with a disutility (U) of -2.017; (2) door-to-door kit delivery (U = +1.029) relative to collection from public/outreach clinic; (3) telephone helpline for pretest support relative to in-person or no support (U = +0.415); (4) distributors from own/local village (U = +0.145) versus those from external communities. Participants who had never HIV tested valued phone helplines more than those previously tested. The strongest LCT preferences were: (1) immediate antiretroviral therapy (ART) availability: U = +0.614 and U = +1.052 for public and outreach clinics, respectively; (2) free services: a $1 user fee increase decreased utility at public (U = -0.381) and outreach clinics (U = -0.761); (3) proximity of clinic (U = -0.38 per hour walking). Participants reported willingness to link to either location; but never-testers were more averse to LCT. Simulations showed the importance of availability of ART: ART unavailability at public clinics would reduce LCT by 24%.

CONCLUSIONS

Free HIVST distribution by local volunteers and immediately available ART were the strongest relative preferences identified. Accommodating LCT preferences, notably ensuring efficient provision of ART, could facilitate "resistant testers" to test while maximizing uptake of post-test services.

Sexuality, fertility and family planning characteristics of married women aged 15 to 19 years in Ethiopia, Nigeria and Tanzania: a comparative analysis of cross-sectional data

Background

Adolescents 360 (A360) is an initiative being rolled out across Ethiopia, Nigeria and Tanzania with the aim of increasing uptake of voluntary modern contraception among sexually active women aged 15 to 19 years. Using evaluation baseline survey data, we described key sexuality, fertility and contraceptive use characteristics of married women aged 15 to 19 years living in three sub-national settings.

Methods

Cross-sectional baseline surveys of married women aged 15 to 19 years were conducted in Oromia (Ethiopia), Nasarawa (Northern Nigeria), and Mwanza (Tanzania) between August 2017 and February 2018. We also interviewed the husbands of a sub-group of married respondents to measure spousal acceptance and support for adolescent women to use modern contraception. A clustered sampling design was used in all three countries. We produced descriptive statistics on the socio-demographic and sexual and reproductive health characteristics of married women aged 15 to 19 years by study setting.

Results

In Oromia, Nasarawa and Mwanza, 31.4% (327/1198), 27.4% (1321/4816) and 7.5% (15/201) of married women surveyed had no education, and 68.3, 81.3 and 83.1% had ever been pregnant, respectively. Unmet need for modern contraception was 20.5, 21.9 and 32.0% in married women in Oromia, Nasarawa and Mwanza, made up almost entirely of unmet need for spacing. The vast majority of married women surveyed in Oromia (89.1%) and Mwanza (90.1%) had seen or heard about contraception in the last 12 months, compared to 30.1% of those surveyed in Nasarawa. Modern contraceptive prevalence (mCPR) was highest in married women aged 15 to 19 years in Oromia (47.2%), followed by Mwanza (19.4%) and Nasarawa (8.7%). Of those using a modern method of contraception in Oromia, 93.4% were using injectables or long-acting methods, compared to 49.4% in Nasarawa and 69.6% in Mwanza.

Conclusions

Overall, unmet need for modern contraception is high among married women aged 15 to 19 years across the three settings. mCPR for married women aged 15 to 19 years is low in Nasarawa and Mwanza. Ultimately, no single intervention will suit all situations, but improving the quality, analyses and utilisation of subnational data can help decision-makers design more context specific interventions.

HIV prevention cascades: A unifying framework to replicate the successes of treatment cascades

Many countries are off track to meet targets for reducing new HIV infections. HIV prevention cascades have been proposed to assist in the implementation and monitoring of HIV prevention programmes by identifying gaps in the steps required for effective use of prevention methods, similar to HIV treatment cascades. However, lack of a unifying framework impedes widespread use of prevention cascades. Building on a series of consultations, we propose an HIV prevention cascade consisting of three key domains of motivation, access, and effective use in a priority population. This three-step cascade can be used for routine monitoring and advocacy, particularly by attaching 90-90-90-style targets. Further characterisation of reasons for gaps across motivation, access, or effective use allows for a comprehensive framework, guiding identification of relevant responses and platforms for interventions. Linking the prevention cascade, reasons for gaps, and interventions reconciles the different requirements of prevention cascades, providing a unifying framework.

Mobility and sex work: why, where, when? A typology of female-sex-worker mobility in Zimbabwe

Sex-worker mobility may have implications for health and access to care but has not been described in sub-Saharan Africa. We described sex-worker mobility in Zimbabwe and a mobility typology using data from 2591 and 2839 female sex workers in 14 sites from 2013 and 2016. We used latent class analysis to identify a typology of mobile sex workers. More women travelled for work in 2016 (59%) than in 2013 (27%), usually to find clients with more money (57% of the journeys), spending a median of 21 (2013) and 24 (2016) days away. A five-class mixture model best fitted the data, with 39% women in an infrequent work-mobility class, 21% in a domestic-high-mobility class, 16% in an international-high-mobility class, 16% in an infrequent opportunistic-non-work-mobility class, and 7% who travel with clients. More-mobile classes were better educated; risk behaviours differed by class. Mobility is increasing among sex worker in Zimbabwe, multi-faceted, and not explained by other vulnerabilities.

Trials and tribulations: cross-learning from the practices of epidemiologists and economists in the evaluation of public health interventions

The randomized controlled trial is commonly used by both epidemiologists and economists to test the effectiveness of public health interventions. Yet we have noticed differences in practice between the two disciplines. In this article, we propose that there are some underlying differences between the disciplines in the way trials are used, how they are conducted and how results from trials are reported and disseminated. We hypothesize that evidence-based public health could be strengthened by understanding these differences, harvesting best-practice across the disciplines and breaking down communication barriers between economists and epidemiologists who conduct trials of public health interventions.

Targeted combination prevention to support female sex workers in Zimbabwe accessing and adhering to antiretrovirals for treatment and prevention of HIV (SAPPH-IRe): a cluster-randomised trial

BACKGROUND

Strengthening engagement of female sex workers with health services is needed to eliminate HIV. We assessed the efficacy of a targeted combination intervention for female sex workers in Zimbabwe.

METHODS

We did a cluster-randomised trial from 2014 to 2016. Clusters were areas surrounding female sex worker clinics and were enrolled in matched pairs. Sites were randomly assigned (1:1) to receive usual care (free sexual-health services supported by peer educators, including HIV testing on demand, referral for antiretroviral therapy [ART], and health education) or an intervention that supported additional regular HIV testing, on-site initiation of ART, pre-exposure prophylaxis, adherence, and intensified community mobilisation. The primary outcome was the proportion of all female sex workers with HIV viral load 1000 copies per mL or greater, assessed through respondent-driven sampling surveys. We used an adapted cluster-summary approach to estimate risk differences. This trial is registered with Pan African Clinical Trials Registry, number PACTR201312000722390.

RESULTS

We randomly assigned 14 clusters to usual care or the intervention (seven in each group). 3612 female sex workers attended clinics in the usual-care clusters and 4619 in the intervention clusters during the study. Half as many were tested (1151 vs 2606) and diagnosed as being HIV positive (546 vs 1052) in the usual-care clusters. The proportion of all female sex workers with viral loads of 1000 copies per mL or greater fell in both study groups (from 421 [30%] of 1363 to 279 [19%] of 1443 in the usual-care group and from 399 [30%] of 1303 to 240 [16%] of 1439 in the intervention group), but with a risk difference at the end of the assessment period of only -2·8% (95% CI -8·1 to 2·5, p=0·23). Among HIV-positive women, the proportions with viral loads less than 1000 copies per mL were 590 (68%) of 869 in the usual-care group and 588 (72%) of 828 in the intervention group at the end of the assessment period, adjusted risk difference of 5·3% (95% CI -4·0 to 14·6, p=0·20). There were no adverse events.

INTERPRETATION

Our intervention of a dedicated programme for female sex workers led to high levels of HIV diagnosis and treatment. Further research is needed to optimise programme content and intensity for the broader population.

FUNDING

UN Population Fund (through Zimbabwe's Integrated Support Fund funded by UK Department for International Development, Irish Aid, and Swedish International Development Cooperation Agency).

Evaluating the impact of the DREAMS partnership to reduce HIV incidence among adolescent girls and young women in four settings: a study protocol

Background

HIV risk remains unacceptably high among adolescent girls and young women (AGYW) in southern and eastern Africa, reflecting structural and social inequities that drive new infections. In 2015, PEPFAR (the United States President’s Emergency Plan for AIDS Relief) with private-sector partners launched the DREAMS Partnership, an ambitious package of interventions in 10 sub-Saharan African countries. DREAMS aims to reduce HIV incidence by 40% among AGYW over two years by addressing multiple causes of AGYW vulnerability. This protocol outlines an impact evaluation of DREAMS in four settings.

Methods

To achieve an impact evaluation that is credible and timely, we describe a mix of methods that build on longitudinal data available in existing surveillance sites prior to DREAMS roll-out. In three long-running surveillance sites (in rural and urban Kenya and rural South Africa), the evaluation will measure: (1) population-level changes over time in HIV incidence and socio-economic, behavioural and health outcomes among AGYW and young men (before, during, after DREAMS); and (2) causal pathways linking uptake of DREAMS interventions to ‘mediators’ of change such as empowerment, through to behavioural and health outcomes, using nested cohort studies with samples of ~ 1000–1500 AGYW selected randomly from the general population and followed for two years. In Zimbabwe, where DREAMS includes an offer of pre-exposure HIV prophylaxis (PrEP), cohorts of young women who sell sex will be followed for two years to measure the impact of ‘DREAMS+PrEP’ on HIV incidence among young women at highest risk of HIV. In all four settings, process evaluation and qualitative studies will monitor the delivery and context of DREAMS implementation. The primary evaluation outcome is HIV incidence, and secondary outcomes include indicators of sexual behavior change, and social and biological protection.

Discussion

DREAMS is, to date, the most ambitious effort to scale-up combinations or ‘packages’ of multi-sectoral interventions for HIV prevention. Evidence of its effectiveness in reducing HIV incidence among AGYW, and demonstrating which aspects of the lives of AGYW were changed, will offer valuable lessons for replication.

Electronic supplementary material

The online version of this article (10.1186/s12889-018-5789-7) contains supplementary material, which is available to authorized users.

Changes in engagement in HIV prevention and care services among female sex workers during intensified community mobilization in 3 sites in Zimbabwe, 2011 to 2015

INTRODUCTION

'Sisters with a Voice', Zimbabwe's nationally scaled comprehensive programme for female sex workers (FSWs), intensified community mobilization activities in three sites to increase protective behaviours and utilization of clinical services. We compare indicators among FSWs at the beginning and after implementation.

METHODS

We used mixed methods to collect data at three sites: in-depth interviews (n = 22) in 2015, routine clinical data from 2010 to 2015, and two respondent driven sampling surveys in 2011 and 2015, in which participants completed an interviewer-administered questionnaire and provided a finger prick blood sample for HIV antibody testing. Estimates were weighted using RDS-1 and estimate convergence assessed in both years. We assessed differences in six indicators between 2011 and 2015 using logistic regression adjusted for age, duration in sex work and education.

RESULTS

870 FSWs were recruited from the three sites in 2011 and 915 in 2015. Using logistic regression to adjust for socio-demographic differences, we found higher estimates of the proportion of HIV-positive FSWs and HIV-positive FSWs who knew their status and reported being on ART in Mutare and Victoria Falls in 2015 compared to 2011. Reported condom use with clients did not differ by year; however, condom use with regular partners was higher in 2015 in Mutare and Hwange. Reported HIV testing in the last six months among HIV-negative FSWs was higher in 2015 across sites: for instance, in Victoria Falls it was 13.4% (95% CI 8.7% to 19.9%) in 2011 and 80.8% (95% CI 74.0 to 87.7) in 2015. FSWs described positive perceptions of the Sisters programme, ease of engaging with health services, and improved solidarity among peers. Programme data showed increases in service use by 2015 across all sites.

CONCLUSIONS

Improvements in key HIV care engagement indicators were observed among FSWs in two sites and in testing and prevention indicators across the three sites after implementation of an intensified community mobilization intervention. Engagement with services for FSWs is critical for countries to reach 90-90-90 targets.

The association between being currently in school and HIV prevalence among young women in nine eastern and southern African countries

INTRODUCTION

Interventions to keep adolescent girls and young women in school, or support their return to school, are hypothesised to also reduce HIV risk. Such interventions are included in the DREAMS combination package of evidence-based interventions. Although there is evidence of reduced risky sexual behaviours, the impact on HIV incidence is unclear. We used nationally representative surveys to investigate the association between being in school and HIV prevalence.

METHODS

We analysed Demographic and Health Survey data from nine DREAMS countries in sub-Saharan Africa restricted to young women aged 15-19 (n = 20,429 in total). We used logistic regression to assess cross-sectional associations between being in school and HIV status and present odds ratios adjusted for age, socio-economic status, residence, marital status, educational attainment and birth history (aOR). We investigated whether associations seen differed across countries and by age.

RESULTS

HIV prevalence (1.0%-9.8%), being currently in school (50.0%-72.6%) and the strength of association between the two, varied between countries. We found strong evidence that being currently in school was associated with a reduced odds of being HIV positive in Lesotho (aOR: 0.37; 95%CI: 0.17-0.79), Swaziland (aOR: 0.32; 95%CI: 0.17-0.59), and Uganda (aOR: 0.48: 95%CI: 0.29-0.80) and no statistically significant evidence for this in Kenya, Malawi, Mozambique, Tanzania, Zambia or Zimbabwe.

CONCLUSIONS

Although the relationship is not uniform across countries or over time, these data are supportive of the hypothesis that young women in school are at lower risk of being HIV positive than those who leave school in some sub-Saharan African settings. There is a possibility of reverse causality, with pre-existing HIV infection leading to school drop-out. Further investigation of the contextual factors behind this variation will be important in interpreting the results of HIV prevention interventions promoting retention in school.

Secondary analysis of tuberculosis stigma data from a cluster randomised trial in Zambia and South Africa (ZAMSTAR)

SETTING

Zambian and South African TB and HIV Reduction (ZAMSTAR) cluster-randomised trial (CRT) communities, 2006-2009.

OBJECTIVES

To develop TB stigma items, and evaluate changes in them in response to a household intervention aimed at reducing TB transmission and prevalence but not tailored to reduce stigma.

DESIGN

TB stigma was measured at baseline and 18 months later among 1826 recently diagnosed TB patients and 1235 adult members of their households across 24 communities; 12 of 24 communities were randomised to receive the household intervention. We estimated the impact of the household intervention on TB stigma using standard CRT analytical methods.

RESULTS

Among household members, prevalence of blame and belief in transmission myths fell in both study arms over time: adjusted prevalence ratios (aPRs) comparing the household intervention with the non-household intervention arm were respectively 0.61 (95%CI 0.26-1.44) and 0.77 (95%CI 0.48-1.25) at 18-month follow-up. Among TB patients, at baseline a low percentage experienced social exclusion and poor treatment by health staff and a relatively high percentage reported 'being made fun of', with little change over time. Disclosure of TB status increased over time in both study arms. Internalised stigma was less prevalent in the household arm at both baseline and follow-up, with an aPR of 0.85 (95%CI 0.41-1.76). Variability in stigma levels between countries and across communities was large.

CONCLUSION

Robust TB stigma items were developed. TB stigma was not significantly reduced by the household intervention, although confidence intervals for estimated intervention effects were wide. We suggest that stigma-specific interventions are required to effectively address TB stigma.

Evaluating the impact of an intervention to increase uptake of modern contraceptives among adolescent girls (15-19 years) in Nigeria, Ethiopia and Tanzania: the Adolescents 360 quasi-experimental study protocol

INTRODUCTION

Nigeria, Ethiopia and Tanzania have some of the highest teenage pregnancy rates and lowest rates of modern contraceptive use among adolescents. The transdisciplinary Adolescents 360 (A360) initiative being rolled out across these three countries uses human-centred design to create context-specific multicomponent interventions with the aim of increasing voluntary modern contraceptive use among girls aged 15-19 years.

METHODS

The primary objective of the outcome evaluation is to assess the impact of A360 on the modern contraceptive prevalence rate (mCPR) among sexually active girls aged 15-19 years. A360 targets different subpopulations of adolescent girls in the three countries. In Northern Nigeria and Ethiopia, the study population is married girls aged 15-19 years. In Southern Nigeria, the study population is unmarried girls aged 15-19 years. In Tanzania, both married and unmarried girls aged 15-19 years will be included in the study. In all settings, we will use a prepopulation and postpopulation-based cross-sectional survey design. In Nigeria, the study design will also include a comparison group. A one-stage sampling design will be used in Nigeria and Ethiopia. A two-stage sampling design will be used in Tanzania. Questionnaires will be administered face-to-face by female interviewers aged between 18 and 26 years. Study outcomes will be assessed before the start of A360 implementation in late 2017 and approximately 24 months after implementation in late 2019.

ETHICS AND DISSEMINATION

Findings of this study will be widely disseminated through workshops, conference presentations, reports, briefings, factsheets and academic publications.

Robust analysis of stepped wedge trials using cluster-level summaries within periods

In stepped‐wedge trials (SWTs), the intervention is rolled out in a random order over more than 1 time‐period. SWTs are often analysed using mixed‐effects models that require strong assumptions and may be inappropriate when the number of clusters is small.

We propose a non‐parametric within‐period method to analyse SWTs. This method estimates the intervention effect by comparing intervention and control conditions in a given period using cluster‐level data corresponding to exposure. The within‐period intervention effects are combined with an inverse‐variance‐weighted average, and permutation tests are used. We present an example and, using simulated data, compared the method to (1) a parametric cluster‐level within‐period method, (2) the most commonly used mixed‐effects model, and (3) a more flexible mixed‐effects model. We simulated scenarios where period effects were common to all clusters, and when they varied according to a distribution informed by routinely collected health data.

The non‐parametric within‐period method provided unbiased intervention effect estimates with correct confidence‐interval coverage for all scenarios. The parametric within‐period method produced confidence intervals with low coverage for most scenarios. The mixed‐effects models' confidence intervals had low coverage when period effects varied between clusters but had greater power than the non‐parametric within‐period method when period effects were common to all clusters.

The non‐parametric within‐period method is a robust method for analysing SWT. The method could be used by trial statisticians who want to emphasise that the SWT is a randomised trial, in the common position of being uncertain about whether data will meet the assumptions necessary for mixed‐effect models.

Reaching young women who sell sex: Methods and results of social mapping to describe and identify young women for DREAMS impact evaluation in Zimbabwe

Young women (aged 15–24) who exchange sex for money or other support are among the highest risk groups for HIV acquisition, particularly in high prevalence settings. To prepare for introduction and evaluation of the DREAMS programme in Zimbabwe, which provides biomedical and social interventions to reduce adolescent girls’ and young women’s HIV vulnerability, we conducted a rapid needs assessment in 6 towns using a “social mapping” approach. In each site, we talked to adult sex workers and other key informants to identify locations where young women sell sex, followed by direct observation, group discussions and interviews. We collected data on socio-demographic characteristics of young women who sell sex, the structure and organisation of their sexual exchanges, interactions with each other and adult sex workers, and engagement with health services. Over a two-week period, we developed a “social map” for each study site, identifying similarities and differences across contexts and their implications for programming and research. Similarities include the concentration of younger women in street-based venues in town centres, their conflict with older sex workers due to competition for clients and acceptance of lower payments, and reluctance to attend existing services. Key differences were found in the 4 university towns included in our sample, where female students participate in diverse forms of sexual exchange but do not identify themselves as selling sex. In smaller towns where illegal gold panning or trucking routes were found, young women migrated in from surrounding rural areas specifically to sell sex. Young women who sell sex are different from each other, and do not work with or attend the same services as adult sex workers. Our findings are being used to inform appropriate intervention activities targeting these vulnerable young women, and to identify effective strategies for recruiting them into the DREAMS process and impact evaluations.

Evaluating the impact of DREAMS on HIV incidence among young women who sell sex: protocol for a non-randomised study in Zimbabwe

Background

“Determined, Resilient, AIDS-free, Mentored and Safe” (DREAMS) is a package of biomedical, social and economic interventions offered to adolescent girls and young women aged 10–24 years with the aim of reducing HIV incidence. In four of the six DREAMS districts in Zimbabwe, DREAMS includes an offer of oral pre-exposure prophylaxis (DREAMS+PrEP), alongside interventions to support demand and adherence, to women aged 18–24 who are at highest risk of HIV infection, including young women who sell sex (YWSS). This evaluation study addresses the question: does the delivery of DREAMS+PrEP through various providers reduce HIV incidence among YWSS Zimbabwe? We describe our approach to designing a rigorous study to assess whether DREAMS+PrEP had an impact on HIV incidence.

Methods

The study design needed to account for the fact that: 1) DREAMS+PrEP was non-randomly allocated; 2) there is no sampling frame for the target population for the evaluation; 3) there are a small number of DREAMS districts (N = 6), and 4) DREAMS+PrEP is being implemented by various providers. The study will use a cohort analysis approach to compare HIV incidence among YWSS in two DREAMS+PrEP districts to HIV incidence among YWSS in non-DREAMS comparison sites. YWSS will be referred to services and recruited into the cohort through a network-based (respondent-driven) recruitment strategy, and followed-up 12- and 24-months after enrolment. Women will be asked to complete a questionnaire and offered HIV testing. Additional complications of this study include identifying comparable populations of YWSS in the DREAMS+PrEP and non-DREAMS comparison sites, and retention of YWSS over the 24-month period. The primary outcome is HIV incidence among YWSS HIV-negative at study enrolment measured by repeat, rapid HIV testing over 24-months. Inference will be based on plausibility that DREAMS+PrEP had an impact on HIV incidence. A process evaluation will be conducted to understand intervention implementation, and document any contextual factors determining the success or failure of intervention delivery.

Discussion

HIV prevention products of known efficacy are available. Innovative studies are needed to provide evidence of how to optimise product use through combination interventions to achieve population impact within different contexts. We describe the design of such a study.

Associations between friendship characteristics and HIV and HSV-2 status amongst young South African women in HPTN-068

INTRODUCTION

Prevalence of HIV among young women in South Africa remains extremely high. Adolescent peer groups have been found to be an important influence on a range of health behaviours. The characteristics of young women's friendships might influence their sexual health and HIV risk via connections to sexual partners, norms around sexual initiation and condom use, or provision of social support. We investigated associations between young women's friendships and their Herpes Simplex Virus Type 2 (HSV-2) and HIV infection status in rural South Africa.

METHODS

Our study is a cross-sectional, egocentric network analysis. In 2011 to 2012, we tested 13- to 20-year-old young women for HIV and HSV-2, and collected descriptions of five friendships for each. We generated summary measures describing friend socio-demographic characteristics and the number of friends perceived to have had sex. We used logistic regression to analyse associations between friend characteristics and participant HIV and HSV-2 infection, excluding likely perinatal HIV infections.

RESULTS

There were 2326 participants included in the study sample, among whom HIV and HSV-2 prevalence were 3.3% and 4.6% respectively. Adjusted for participant and friend socio-demographic characteristics, each additional friend at least one year older than the participant was associated with raised odds of HIV (odds ratio (OR) = 1.37, 95% CI 1.03 to 1.82) and HSV-2 (adjusted OR=1.41, 95% CI 1.18 to 1.69). Each additional friend perceived to have ever had sex also raised the odds of HIV (OR = 1.29, 95% CI 1.03 to 1.63) and HSV-2 (OR=1.18, 95% CI 1.03 to 1.35).

DISCUSSION

We found good evidence that a greater number of older friends and friends perceived to have had sex were associated with increased risk for HSV-2 and HIV infection among young women.

CONCLUSIONS

The characteristics of young women's friendships could contribute to their risk of HIV infection. The extent to which policies or programmes influence age-mixing and young women's normative environments should be considered.

Sample Size Calculations for Population Size Estimation Studies Using Multiplier Methods With Respondent-Driven Sampling Surveys

Background

While guidance exists for obtaining population size estimates using multiplier methods with respondent-driven sampling surveys, we lack specific guidance for making sample size decisions.

Objective

To guide the design of multiplier method population size estimation studies using respondent-driven sampling surveys to reduce the random error around the estimate obtained.

Methods

The population size estimate is obtained by dividing the number of individuals receiving a service or the number of unique objects distributed (M) by the proportion of individuals in a representative survey who report receipt of the service or object (P). We have developed an approach to sample size calculation, interpreting methods to estimate the variance around estimates obtained using multiplier methods in conjunction with research into design effects and respondent-driven sampling. We describe an application to estimate the number of female sex workers in Harare, Zimbabwe.

Results

There is high variance in estimates. Random error around the size estimate reflects uncertainty from M and P, particularly when the estimate of P in the respondent-driven sampling survey is low. As expected, sample size requirements are higher when the design effect of the survey is assumed to be greater.

Conclusions

We suggest a method for investigating the effects of sample size on the precision of a population size estimate obtained using multipler methods and respondent-driven sampling. Uncertainty in the size estimate is high, particularly when P is small, so balancing against other potential sources of bias, we advise researchers to consider longer service attendance reference periods and to distribute more unique objects, which is likely to result in a higher estimate of P in the respondent-driven sampling survey.

Interventions to strengthen the HIV prevention cascade: a systematic review of reviews

BACKGROUND

Much progress has been made in interventions to prevent HIV infection. However, development of evidence-informed prevention programmes that translate the efficacy of these strategies into population effect remain a challenge. In this systematic review, we map current evidence for HIV prevention against a new classification system, the HIV prevention cascade.

METHODS

We searched for systematic reviews on the effectiveness of HIV prevention interventions published in English from Jan 1, 1995, to July, 2015. From eligible reviews, we identified primary studies that assessed at least one of: HIV incidence, HIV prevalence, condom use, and uptake of HIV testing. We categorised interventions as those seeking to increase demand for HIV prevention, improve supply of HIV prevention methods, support adherence to prevention behaviours, or directly prevent HIV. For each specific intervention, we assigned a rating based on the number of randomised trials and the strength of evidence.

FINDINGS

From 88 eligible reviews, we identified 1964 primary studies, of which 292 were eligible for inclusion. Primary studies of direct prevention mechanisms showed strong evidence for the efficacy of pre-exposure prophylaxis (PrEP) and voluntary medical male circumcision. Evidence suggests that interventions to increase supply of prevention methods such as condoms or clean needles can be effective. Evidence arising from demand-side interventions and interventions to promote use of or adherence to prevention tools was less clear, with some strategies likely to be effective and others showing no effect. The quality of the evidence varied across categories.

INTERPRETATION

There is growing evidence to support a number of efficacious HIV prevention behaviours, products, and procedures. Translating this evidence into population impact will require interventions that strengthen demand for HIV prevention, supply of HIV prevention technologies, and use of and adherence to HIV prevention methods.

FUNDING

Bill & Melinda Gates Foundation.

The HIV prevention cascade: integrating theories of epidemiological, behavioural, and social science into programme design and monitoring

Theories of epidemiology, health behaviour, and social science have changed the understanding of HIV prevention in the past three decades. The HIV prevention cascade is emerging as a new approach to guide the design and monitoring of HIV prevention programmes in a way that integrates these multiple perspectives. This approach recognises that translating the efficacy of direct mechanisms that mediate HIV prevention (including prevention products, procedures, and risk-reduction behaviours) into population-level effects requires interventions that increase coverage. An HIV prevention cascade approach suggests that high coverage can be achieved by targeting three key components: demand-side interventions that improve risk perception and awareness and acceptability of prevention approaches; supply-side interventions that make prevention products and procedures more accessible and available; and adherence interventions that support ongoing adoption of prevention behaviours, including those that do and do not involve prevention products. Programmes need to develop delivery platforms that ensure these interventions reach target populations, to shape the policy environment so that it facilitates implementation at scale with high quality and intensity, and to monitor the programme with indicators along the cascade.

Bias and inference from misspecified mixed-effect models in stepped wedge trial analysis

Many stepped wedge trials (SWTs) are analysed by using a mixed‐effect model with a random intercept and fixed effects for the intervention and time periods (referred to here as the standard model). However, it is not known whether this model is robust to misspecification.

We simulated SWTs with three groups of clusters and two time periods; one group received the intervention during the first period and two groups in the second period. We simulated period and intervention effects that were either common‐to‐all or varied‐between clusters. Data were analysed with the standard model or with additional random effects for period effect or intervention effect. In a second simulation study, we explored the weight given to within‐cluster comparisons by simulating a larger intervention effect in the group of the trial that experienced both the control and intervention conditions and applying the three analysis models described previously.

Across 500 simulations, we computed bias and confidence interval coverage of the estimated intervention effect.

We found up to 50% bias in intervention effect estimates when period or intervention effects varied between clusters and were treated as fixed effects in the analysis. All misspecified models showed undercoverage of 95% confidence intervals, particularly the standard model. A large weight was given to within‐cluster comparisons in the standard model.

In the SWTs simulated here, mixed‐effect models were highly sensitive to departures from the model assumptions, which can be explained by the high dependence on within‐cluster comparisons. Trialists should consider including a random effect for time period in their SWT analysis model. © 2017 The Authors. Statistics in Medicine published by John Wiley & Sons Ltd.

Measuring Sexual Behavior Stigma to Inform Effective HIV Prevention and Treatment Programs for Key Populations

BACKGROUND

The levels of coverage of human immunodeficiency virus (HIV) treatment and prevention services needed to change the trajectory of the HIV epidemic among key populations, including gay men and other men who have sex with men (MSM) and sex workers, have consistently been shown to be limited by stigma.

OBJECTIVE

The aim of this study was to propose an agenda for the goals and approaches of a sexual behavior stigma surveillance effort for key populations, with a focus on collecting surveillance data from 4 groups: (1) members of key population groups themselves (regardless of HIV status), (2) people living with HIV (PLHIV) who are also members of key populations, (3) members of nonkey populations, and (4) health workers.

METHODS

We discuss strengths and weaknesses of measuring multiple different types of stigma including perceived, anticipated, experienced, perpetrated, internalized, and intersecting stigma as measured among key populations themselves, as well as attitudes or beliefs about key populations as measured among other groups.

RESULTS

With the increasing recognition of the importance of stigma, consistent and validated stigma metrics for key populations are needed to monitor trends and guide immediate action. Evidence-based stigma interventions may ultimately be the key to overcoming the barriers to coverage and retention in life-saving antiretroviral-based HIV prevention and treatment programs for key populations.

CONCLUSIONS

Moving forward necessitates the integration of validated stigma scales in routine HIV surveillance efforts, as well as HIV epidemiologic and intervention studies focused on key populations, as a means of tracking progress toward a more efficient and impactful HIV response.

Overlapping HIV and sex-work stigma among female sex workers recruited to 14 respondent-driven sampling surveys across Zimbabwe, 2013

HIV stigma can inhibit uptake of HIV testing and antiretroviral therapy as well as negatively affect mental health. Efforts to reduce discrimination against people living with HIV (LWH) have contributed to greater acceptance of the infection. Female sex workers (FSW) LWH may experience overlapping stigma due to both their work and HIV status, although this is poorly understood. We examined HIV and sex-work stigma experienced by FSW LWH in Zimbabwe. Using the SAPPH-IRe cluster-randomised trial baseline survey, we analysed the data from 1039 FSW self-reporting HIV. The women were recruited in 14 sites using respondent-driven sampling. We asked five questions to assess internalised and experienced stigma related to working as a sex worker, and the same questions were asked in reference to HIV. Among all FSW, 91% reported some form of sex-work stigma. This was not associated with sociodemographic or sex-work characteristics. Rates of sex-work stigma were higher than those of HIV-related stigma. For example, 38% reported being "talked badly about" for LWH compared with 77% for their involvement in sex work. Those who reported any sex-work stigma also reported experiencing more HIV stigma compared to those who did not report sex-work stigma, suggesting a layering effect. FSW in Zimbabwe experience stigma for their role as "immoral" women and this appears more prevalent than HIV stigma. As HIV stigma attenuates, other forms of social stigma associated with the disease may persist and continue to pose barriers to effective care.

HIV-related stigma and universal testing and treatment for HIV prevention and care: design of an implementation science evaluation nested in the HPTN 071 (PopART) cluster-randomized trial in Zambia and South Africa

BACKGROUND

Stigma and discrimination related to HIV and key populations at high risk of HIV have the potential to impede the implementation of effective HIV prevention and treatment programmes at scale. Studies measuring the impact of stigma on these programmes are rare. We are conducting an implementation science study of HIV-related stigma in communities and health settings within a large, pragmatic cluster-randomized trial of a universal testing and treatment intervention for HIV prevention in Zambia and South Africa and will assess how stigma affects, and is affected by, implementation of this intervention.

METHODS/DESIGN

A mixed-method evaluation will be nested within HIV prevention trials network (HPTN) 071/PopART (Clinical Trials registration number NCT01900977), a three-arm trial comparing universal door-to-door delivery of HIV testing and referral to prevention and treatment services, accompanied by either an immediate offer of anti-retroviral treatment to people living with HIV regardless of clinical status, or an offer of treatment in-line with national guidelines, with a standard-of-care control arm. The primary outcome of HPTN 071/PopART is HIV incidence measured among a cohort of 52 500 individuals in 21 study clusters. Our evaluation will include integrated quantitative and qualitative data collection and analysis in all trial sites. We will collect quantitative data on indicators of HIV-related stigma over 3 years from large probability samples of community members, health workers and people living with HIV. We will collect qualitative data, including in-depth interviews and observations from members of these same groups sampled purposively. In analysis, we will: (1) compare HIV-related stigma measures between study arms, (2) link data on stigma to measures of the success of implementation of the PopART intervention and (3) explore changes in the dominant drivers and manifestations of stigma in study communities and the health system.

DISCUSSION

HIV-related stigma may impede the successful implementation of HIV prevention and treatment programmes. Using a novel study-design nested within a large, community randomized trial we will evaluate the extent to which HIV-related stigma affects and is affected by the implementation of a comprehensive combination HIV prevention intervention including a universal test and treatment approach.

Proceedings of the First International Conference on Stepped Wedge Trial Design

I1 Introduction

Mona Kanaan, Noreen Dadirai Mdege, Ada Keding

O1 The HiSTORIC trial: a hybrid before-and-after and stepped wedge design

RA Parker, N Mills, A Shah, F Strachan, C Keerie, CJ Weir

O2 Stepped wedge trials with non-uniform correlation structure

Andrew Forbes, Karla Hemming

O3 Challenges and solutions for the operationalisation of the ENHANCE study: a pilot stepped wedge trial within a general practice setting

Sarah A Lawton, Emma Healey, Martyn Lewis, Elaine Nicholls, Clare Jinks, Valerie Tan, Andrew Finney, Christian D Mallen, on behalf of the ENHANCE Study Team

O4 Early lessons from the implementation of a stepped wedge trial design investigating the effectiveness of a training intervention in busy health care settings: the Thistle study

Erik Lenguerrand, Graeme MacLennan, John Norrie, Siladitya Bhattacharya, Tim Draycott, on behalf of the Thistle group

O5 Sample size calculation for longitudinal cluster randomised trials: a unified framework for closed cohort and repeated cross-section designs

Richard Hooper, Steven Teerenstra, Esther de Hoop, Sandra Eldridge

O6 Restricted randomisation schemes for stepped-wedge studies with a cluster-level covariate

Alan Girling, Monica Taljaard

O7 A flexible modelling of the time trend for the analysis of stepped wedge trials: results of a simulation study

Gian Luca Di Tanna, Antonio Gasparrini

P1 Tackling acute kidney injury – a UK stepped wedge clinical trial of hospital-level quality improvement interventions

Anna Casula, Fergus Caskey, Erik Lenguerrand, Shona Methven, Stephanie MacNeill, Margaret May, Nicholas Selby

P2 Sample size considerations for quantifying secondary bacterial transmission in a stepped wedge trial of influenza vaccine

Leon Danon, Hannah Christensen, Adam Finn, Margaret May

P3 Sample size calculation for time-to-event data in stepped wedge cluster randomised trials

Fumihito Takanashi, Ada Keding, Simon Crouch, Mona Kanaan

P4 Sample size calculations for stepped-wedge cluster randomised trials with unequal cluster sizes

Caroline A. Kristunas, Karen L. Smith, Laura J. Gray

P5 The design of stepped wedge trials with unequal cluster sizes

John N.S. Matthews

P6 Promoting Recruitment using Information Management Efficiently (PRIME): a stepped wedge SWAT (study-within-a-trial)

R Al-Shahi Salman, RA Parker, A Maxwell, M Dennis, A Rudd, CJ Weir

P7 Implications of misspecified mixed effect models in stepped wedge trial analysis: how wrong can it be?

Jennifer A Thompson, Katherine L Fielding, Calum Davey, Alexander M Aiken, James R Hargreaves, Richard J Hayes

S1 Stepped Wedge Designs with Multiple Interventions

Vivian H Lyons, Lingyu Li, James Hughes, Ali Rowhani-Rahbar

S2 Analysis of the cross-sectional stepped wedge cluster randomised trial

Karla Hemming, Monica Taljaard, Andrew Forbes