Lafiyan Jikin Mata "Excellent Health for Mothers"

Methods/ Study design/ Population/ Interventions Study Design: A cluster randomized controlled trial in which 32 public primary and secondary health care facilities are randomized into 16 intervention and 16 control sites. The control sites will receive routine CIHP support, while the intervention sites will receive structured CQI-BTS interventions.

Study sites Based on pre-set site selection criteria, 32 sites were selected from a total of 241 eligible sites supported by the Centre for Integrated Healthcare Programs (CIHP), to provide comprehensive HIV care and treatment programs. These sites are at higher levels of standardized care compared to Government of Nigeria sites where there is minimal standardization of care. The following criteria were used to select health facilities for the study;

• Facility supports ANC/PMTCT services and has capacity to provide ART until 12-18 months post-partum
• Onsite delivery of HIV+ women by the facility
• Provision of postpartum care follow-up for HIV+ women
• ANC attendance with at least 2 HIV+ women per month for general hospitals and at least 1 per month for primary health care (PHCs)
• Availability of, at least, two trained community health extension workers
• Facilities providing PMTCT services for more than six months prior to onset of study

Study Participants:

Participants are HIV-infected pregnant women presenting at study sites for 1st booking at antenatal clinics with gestational age of 34 weeks or less, ART naive, accept ARV and agree to remain in care for at least six months post-partum.

Control and Intervention arms:

Individual pregnant women and mothers attending facilities in the control and intervention arms receive the same routine care that includes a "test and treat" protocol i.e. same day HIV testing and initiation of triple ARVs (option B) of HIV+ pregnant women. Clients eligible for lifelong ART based on CD4 count, and/or WHO staging, continue ARVs indefinitely for their own health, while those not eligible continue ARV prophylaxis according to the National guidelines. For the purpose of this study, antenatal clients are seen for their ARV drug refills until delivery according to the National ANC schedule of appointments. Postpartum, visits follow the immunization schedule i.e. 6 weeks, 10 weeks, and 14 weeks and thereafter 2 monthly until 12 months. Facilities in the control arm continue to receive routine health system support from CIHP

Intervention package:

CQI interventions employ rapid cycles of data collection, testing of solutions and review of changes will be implemented. At each site a Quality Improvement Team (QIT) will be established from among the local staff. Local Government and State level QI teams will also be set up to provide oversight function of the health facilities' QI initiatives. The BTS is a collaborative learning approach conducted quarterly in each study state at a central location with participants from the intervention sites. The sessions provide opportunity for teams to learn from each other and to adapt and implement changes using the Plan-Do-Study-Act (PDSA) model. Driver diagrams will be developed to understand the primary and secondary drivers for low retention of HIV positive mothers and formulate action plans and indicator sets to address these. Between the collaborative sessions, specific designated improvement coaches will follow up site teams at regular intervals via physical site visits and telephone calls to guide implementation of change ideas. Different site teams will also be encouraged to communicate with each other through telephone calls and cross-site visits to strengthen learning and exchange of ideas.

Sample size estimation:

Sample size was calculated using the windows programme for epidemiologists. The CQI-BTS approach tested within the study would be deemed worthwhile if there is a 20% absolute increase in retention of mothers at 6 months (e.g. expected retention rate in the control arm: 40%; expected retention rate in the control arm: 60%). First, the minimum sample size needed under individual randomization was calculated to detect the desirable difference of proportions with a level of significance set at 95% (two-sided) and at 80% power. We found that 94 mother-infant pairs would be needed in each arm under individual randomization. Second, bearing in mind that this trial will randomize the intervention package under several clusters, the sample size from a non-cluster randomized trial was inflated by an Inflation Factor (IF). The IF is calculated as follows: IF = 1 + (m - 1) ro, where m is the size of each cluster, and ro the internal variability which represents how strongly individuals within clusters are related to each other. In our case, it is difficult to estimate a priori this internal variability due to lack of primary data: it is possible that the standardized structure of the intervention tested within the trial will lead to a relatively low cluster internal variability; alternatively, because the intervention package depends on the individual capacity/will of health-care personnel, this may increase the internal variability across clusters. Hence, we believe it is reasonable to assume a ro value of 0.10, as recommended by most authors. We set the size of cluster at 20, which gave an inflation factor of 2.9. Additionally, since this research is being carried out in an operational context, we allowed for up to 20% of routine data to be incomplete or maternal mortality in pregnancy and therefore data on retention in care not available for analysis. As a conclusion, a total of 327 mother-infant pairs are needed in each arm. There will be 16 clusters in each arm with a minimum number of 20 HIV-infected pregnant women per site. A total of 640 pregnant HIV-infected women will be recruited.

Data management plan:

Primary outcome data is sourced from the local health registers -Data collection and entry occur weekly in both study arms by research assistants. A monthly review of these data is conducted by the LJM study team. In addition, process measures are routinely collected to assess the integrity of implementation of the CQI-BTS intervention.

Primary Outcome: Proportion of HIV+ pregnant women in care 6 months postpartum Source Document: PMTCT ARV register

Secondary Outcome 1: Proportion of HIV+ pregnant women initiating ARV prophylaxis or treatment within 2 weeks of enrolment Source Document: PMTCT HCT register, PMTCT ARV register and Mother Child Care Form

Secondary Outcome 2: Proportion of HIV+ pregnant women care 12 months postpartum Source Document: PMTCT ARV register, Mother and Child Care Form

Secondary Outcome 3: Proportion of HIV exposed infants with Dry Blood Spot (DBS) samples taken at 8 weeks of life Source Document: Child follow up register and Mother Child Care Form

Analysis plan:

Descriptive analyses will be conducted to outline participants' and clusters' baseline characteristics and outcomes by study arms using frequency tabulations and summary statistics. Bi-variate analysis will be done to identify potential prognostic factors confounding the distribution of baseline characteristics among the arms that require adjustment.