Implementation Research for Vulnerable Women in South Africa

This five-year implementation science study used a cluster-randomized stepped-wedge design to evaluate the implementation, service, and patient outcomes associated with the WHC for alcohol and other drug (AOD)-using HIV positive women in usual care settings.

A list of four substance use treatment clinics and four healthcare clinics, identified and approved by the City of Cape Town, South Africa were randomized to begin the intervention during one of the four cycles. Each healthcare clinic was paired with a substance use treatment clinic based on geographic proximity, and each pair was randomized by computer into four succeeding 6-month implementation cycles where implementation of the WHC took place simultaneously at the paired sites. Approximately 120 HIV positive participants were recruited in each cycle (approximately 60 from each clinic) for the patient level outcomes.

Both qualitative and quantitative data were collected to assess the appropriateness of marketing plans developed through formative methods, as well as the acceptability, adoption, feasibility, fidelity, and sustainability of the WHC intervention implementation as well as service outcomes (comprehensive services and timely service linkages) during each implementation cycle.

Each implementation cycle included a pre-implementation period, implementation period (6-months), and post-implementation period. In the pre-implementation period, focus groups and questionnaires surveys were conducted with clinic staff to assess the readiness of each site to implement the WHC. During the implementation period, employees at each site were trained to facilitate the WHC and the WHC was integrated into site operations. Subsequently, in the post-implementation period the researchers collected data related to challenges, benefits, and sustainability from each site. The process was repeated for each cycle and these formative periods between cycles were used to inform backward- and forward- implementation strategies, make modifications to the WHC, and leave time for site-specific training for the next cycle. Consequently, the sites randomized to the first cycle had the longest post-intervention observation period whose implementation sustainability was checked through fidelity forms and sustainability questionnaires, and sites in the fourth cycle benefited the most because of lessons learned and information shared from previous cycles.

The intervention was implemented in a group, however there were instances when only one participant was available and therefore the intervention was implemented one-on-one. The WHC has previously been tested in group and one-on-one formats and both have demonstrated consistent significant intervention effects. Research staff trained clinic staff members to deliver the intervention. The intervention was delivered by clinic staff and was not part of the research. The research questions were related to the feasibility of implementing the intervention in clinics and its acceptability to clinic staff and patients. The researchers collected information on patient-level outcomes to determine if the intervention was effective when it was delivered by clinic staff to patients in the clinic. Also, to assess acceptability of the intervention workshops among patients, the researchers conducted post-implementation focus groups with a randomly selected sub-sample of participants who participated in the intervention and completed their final 6-month appointment in each implementation cycle.