A simulation and experiential learning intervention for labor and delivery providers to address HIV stigma during childbirth in Tanzania: Study protocol for the evaluation of the MAMA intervention

Background: The experience of HIV stigma during intrapartum care can impact women’s trust in the health care system and undermine their long-term commitment to HIV care engagement. Delivery of respectful maternity care (RMC) to WLHIV can improve quality of life and clinical outcomes. The goal of this study is to conduct an evaluation of MAMA (Mradi wa Afya ya Mama Mzazi, Project to Support the Health of Women Giving Birth), a simulation team-training curriculum for labor and delivery providers that addresses providers’ instrumental and attitudinal stigma toward WLHIV and promotes the delivery of evidence-based RMC for WLHIV. Methods: The MAMA intervention will be evaluated among healthcare providers across six clinics in the Kilimanjaro Region of Tanzania. To evaluate the impact of MAMA, we will enroll WLHIV who give birth in the facilities before (n=103 WLHIV) and after (n=103 WLHIV) the intervention. We will examine differences in the primary outcome (perceptions of RMC) and secondary outcomes (postpartum HIV care engagement; perceptions of HIV stigma in the facility; internal HIV stigma; clinical outcomes and evidence-based practices) between women enrolled in the two time periods. Will also assess participating providers (n=60) at baseline, immediate post, 1-month post training, and 2-month post training. We will examine longitudinal changes in the primary outcome (practices of RMC) and secondary outcomes (stigma toward WLHIV; self-efficacy in delivery intrapartum care). Quality assurance data will be collected to assess intervention feasibility and acceptability. Discussion: The implementation findings will be used to finalize the intervention for a train-the-trainer model that is scalable, and the outcomes data will be used to power a multi-site study to detect significant differences in HIV care engagement. Trial Registration: The trial is registered at clinicaltrials.gov, NCT05271903.

Discussion: The implementation ndings will be used to nalize the intervention for a train-the-trainer model that is scalable, and the outcomes data will be used to power a multi-site study to detect signi cant differences in HIV care engagement.
Trial Registration: The trial is registered at clinicaltrials.gov, NCT05271903.
Background HIV stigma during the intrapartum period (labor & delivery) can impact the birth experience of women living with HIV (WLHIV), and, in turn, can in uence women's long-term commitment to HIV care.
Prevention of mother-to-child transmission (PMTCT) programs in Africa have achieved near-universal HIV testing for pregnant women and reduced infant HIV infections; however, overall retention in HIV care, especially during the postpartum period, has been suboptimal (1)(2)(3)(4)(5). Our systematic review of PMTCT programs across Africa found that 27% of postpartum WLHIV were no longer in care at six months, rates substantially lower than other groups of people living with HIV (5). Multiple studies have noted the impact of HIV-related stigma on care linkage and retention in PMTCT programs (4,(6)(7)(8). Although data is limited, there is evidence that labor and delivery (L&D) providers may deliver suboptimal and stigmatizing care to WLHIV. Despite this, HIV stigma reduction interventions have not targeted the intrapartum period. In a review of interventions addressing stigma among healthcare providers, none were found to focus on L&D providers (9,10). Lack of engagement with L&D providers is a missed opportunity to change the trajectory of care for WLHIV.
Data from health facilities around the globe suggest that mistreatment and neglect are common during childbirth, which not only undermines women's dignity but also contributes to poor maternal and child health outcomes (11,12). The World Health Organization has proposed a framework of respectful maternity care (RMC), which includes principles of dignity, privacy, con dentiality, freedom from harm, and informed choice (13). RMC not only improves maternal and child health outcomes, but also reduces experiences of stigma, disrespect and abuse, which results in greater engagement and trust in the healthcare system (14,15).
When HIV stigma is present during the intrapartum period, it undermines the delivery of RMC(16). HIV stigma manifests both in the delivery of care and in the receipt of care by WLHIV. Stigma on the part of healthcare providers includes both instrumental stigma (avoidance and neglect, often driven by fear of occupational exposure) and attitudinal stigma (blame and judgment of people living with HIV) (17). L&D providers who deliver respectful and intentionally non-stigmatizing care to WLHIV can help women to overcome both internalized HIV stigma (how they feel about themselves as a WLHIV) and anticipated HIV stigma (how they expect others will treat them). A respectful birth can also deepen women's trust in the health care system, and motivate them to remain in HIV care during the postpartum period(18). While this is important for all women, it is essential for WLHIV, for whom postpartum care engagement is necessary to ensure long-term health outcomes, including newborn HIV testing and long-term adherence to antiretroviral therapy (ART).
Healthcare providers are not immune to the cognitive biases that are deeply embedded in societal norms, and these biases directly impact the delivery of evidence-based practices for stigmatized populations (19). Providers' unconscious attitudes or implicit bias can manifest as subtle in isolation, yet the cumulative effects create mistrust between patients and providers, erode patients' faith in the health care system, and signi cantly impede positive health outcomes (20)(21)(22).
Simulation training is a team-based experiential learning approach that helps providers re ne clinical skills and improve inter-personal competencies, such as patient-provider communication. It integrates cognitive, technical, and behavioral objectives, building on shared experiences, and facilitating re ection and debrie ng. Simulation learning can change providers' attitudes and behaviors and improve the delivery of RMC (9).Simulation training has the potential to reduce instrumental stigma by helping L&D providers develop the skills and self-e cacy to manage routine births and obstetric emergencies in WLHIV while protecting themselves from occupational exposure. It can also reduce attitudinal stigma by helping L&D providers re ect on their implicit biases and how these might be expressed and received by WLHIV.
The goal of this study is to conduct an evaluation of MAMA (Mradi wa Afya ya Mama Mzazi, Project to Support the Health of Women Giving Birth), a simulation team-training curriculum for L&D providers that addresses providers' instrumental and attitudinal stigma toward WLHIV and promotes the delivery of evidence-based, respectful maternity care for WLHIV.

Study Overview
This study will evaluate the MAMA training intervention among healthcare providers across six clinics in the Kilimanjaro Region in northern Tanzania. The study will evaluate the impact of the intervention on patient outcomes using a quasi-experimental pre-post design (comparing patient-reported outcomes in the pre-and post-intervention periods), and the impact on provider outcomes using pre-and post-test assessment. The collection of robust quality assurance data will support the re nement of the intervention and prepare us to implement a future clinic-randomized trial. Table 1 summarizes key elements of the study.

Study Setting
The study will be conducted in six primary health centers, including 4 government hospitals and 2 Designated District Hospitals (DDH), which are owned and operated by a faith-based organization and supported by the government. The study will be located in Moshi (urban) and Rombo (rural), both in the Kilimanjaro Region. In 2019, the six clinics together provided intrapartum services for 258 WLHIV.

Participants
The study will enroll approximately 400 birthing women and 60 labor and delivery providers. Birthing women will be eligible if they delivered a baby in a study facility during either the pre-intervention enrollment period or post-intervention enrollment period. Given our primary focus on WLHIV, we will oversample this population, with a goal of 103 WLHIV recruited in each period. For each WLHIV enrolled in the study, we will recruit up to two women who are not living with HIV who gave birth the same day, in the same facility, and are similar in age (+/-5 years) and parity (nulliparous, primiparous, or multiparous). Providers will be eligible if their clinical duties include attending and supporting childbirth in any of the six study facilities.

Screening and Recruitment
Birthing Women. Participants will be recruited from the maternity wards of the six participating facilities. When a WLHIV is in the postpartum ward following childbirth, a member of the clinic staff will introduce the study; if the woman is interested, the facility staff will send a text message to the study coordinator about a potential participant. The study staff will meet with the woman to introduce the study and answer questions. The informed consent process will be conducted orally; to ensure that HIV status is not involuntarily disclosed to others, the consent form does not include any information about HIV status. The same process is repeated with HIV-negative women who meet eligibility criteria.
Providers. Facility leadership will help to identify facility providers who can be released from duty to attend the two-day MAMA training. Research staff will meet with providers as a group to describe research activities, including the time commitment. When providers report to the training, they will be given the informed consent form to review; prior to signing, they will have an opportunity to ask questions.

Procedures
Birthing Women. After providing consent, patient participants will complete a structured survey using audio computer-assisted self-interview (ACASI) technology on tablets running Questionnaire Development System (QDS) software. The ACASI modality ensures patient privacy and minimizes social desirability bias to improve data validity (23). Participants will complete the assessment on individual tablets where they can read the screen and listen to recorded audio for the questions and response options in Kiswahili. The survey will take approximately 30-45 minutes to complete. Research staff will be present in the room to aid participants if needed. The study staff will review the patient's medical records to collect clinical information about the birth (e.g., patient history, clinical management, complications, birth outcomes). A locator form will be completed to document information about how the participant can best be reached for a brief follow-up survey. Approximately three weeks after the birth, the research staff will call the participant to administer a brief survey focused on the health and well-being of the woman and her child, and engagement in post-partum clinical care. The researcher will attempt follow-up for a two week period before identifying a participant as lost to follow up. The follow-up survey will take approximately 10-15 minutes and take place over the phone.
Providers: After providing informed consent, providers will self-complete a paper-based survey evaluating RMC practices, stigma towards WLHIV, and self-e cacy and clinical knowledge of obstetric emergency care. The providers will then participate in the MAMA intervention, which is a two-day workshop in a central location, followed by a one-day in situ refresher training 1-2 months later at each of the 6 facilities. Follow-up surveys will be conducted at three time points: immediate post training, pre in situ refresher, and two months post in situ refresher. The research team will make an appointment for the follow-up surveys by phone, and then visit the participant on site during a working shift in order to administer the survey. If the participant has been transferred to a new clinical site within the Kilimanjaro Region, they will visit the new site.

Data management
For the patient ACASI data collection, data les will be transferred and uploaded daily by the local data manager in Tanzania. The les will be stored on a secured drive (Box.com) and les will be reviewed weekly for quality assurance. For the provider data collection, paper surveys will be entered into a RedCap database. Limited access to stored data, only identi able by a unique ID, will be provided to authorized research staff.

Intervention condition
The MAMA intervention is a two-day simulation training workshop for labor and delivery providers, followed by a one-day in situ refresher course 1-2 months later. The MAMA intervention is based on PRONTO International's simulation training program to improve obstetric care. PRONTO International is a non-governmental organization whose simulation and team-training curriculum has been implemented in a variety of clinical settings, from primary health centers to district and regional hospitals (24)(25)(26)(27). The PRONTO model has been applied and evaluated in multiple settings, including East Africa, and has been shown to improve provider skills (25,27,28), reduce morbidity and mortality(26, [29][30][31], and improve the use of RMC practices (9). The interactive curriculum is based on simulation and debrie ng of clinical scenarios, case-based learning, skills stations, and teamwork activities.
Adaptation of the PRONTO model to create the MAMA intervention followed the ADAPT-ITT model (32). In order to understand the experiences of RMC for WLHIV, we conducted qualitative in-depth interviews with pregnant and postpartum women (n = 36), focus group discussions with L&D providers in the study clinics (n = 6) and focus group discussions with nurse-midwifery students (n = 2). This data allowed us to best target the key actionable drivers of HIV stigma among health care providers, and to address the clinical needs of the provider population. The curriculum was drafted by the inter-disciplinary international study team (including nurse midwives, an obstetrician, a pediatrician, and a behavioral scientist), and in consultation with the PRONTO leadership. We received two rounds of input from a stakeholder advisory board, which included representation from the municipal and regional Departments of Health, HIV advocates, maternal and child health advocates, community representative and nurse and midwife educators and experts. We conducted "theater testing"(33) of the simulation models in Tanzania, in order to re ne the training materials.
The nal two-day training curriculum includes three simulation scenarios, knowledge reviews (interactive case-based learning activities), teamwork and communication activities, and interactive activities that address bias and stigma. Activities are intentionally ordered so that the learning is scaffolded and the opportunities and challenges require increasing integration and application of new skills and knowledge on clinical practices, bias, and RMC. The in-situ follow-up training includes a structured review of training materials, one simulation scenario, and exibility to focus on areas the trainers identify as needing additional work. The in-situ training will highlight facility readiness and systems level weaknesses that can be addressed by the participating team.
The training curriculum will be delivered by four PRONTO Master Trainers who are based in Kenya. The 2day training will be offered twice (with approximately 30 providers at each training). The training will be hosted at the KCMC Hospital, which is the tertiary referral hospital for all the study facilities.

Outcomes
Primary outcome for birthing women Perceptions of respectful maternity care will be measured with a 31-item measure of person-centered maternity care developed by Afulani and colleagues. (34,35) The measure asks a woman to report on the quality of care she received as it relates to domains of RMC. Items will be summed and standardized on a scale of 0-100. The scale includes three sub-scales: dignity and respect (6 items), communication and autonomy (9 items), and supportive care (15 items).

Secondary outcomes for birthing women
Postpartum HIV care engagement will be measured in the phone follow-up interview after participants have been discharged from the hospital. Participants will be asked a single question (yes/no) about whether they have attended any postpartum visits to the prevention of mother to child transmission (PMTCT) clinic. Participants will also be asked questions about adherence to antiretroviral (ARV) medication in the prior four days. If participants have not attended any PMTCT visits in the postpartum period, or if they have missed any ARV medication in the prior four days, they will be considered to have poor care engagement.
Perceptions of HIV stigma in the facility will be assessed among WLHIV using an 11-item measure that was adapted from two measures of HIV stigmatizing behaviors of health care workers(36, 37). The measure asks women to rate their perception of stigma by health care providers (e.g., Do you feel like providers avoided touching you because of your HIV status? Did a provider say something negative to you about your HIV status?). Item responses are on a Likert scale (0-3), and will be summed for a possible range 0-33.
Internalized HIV stigma will be assessed among WLHIV using Scale A of the HIV and Abuse Related Shame Inventory (HARSI)(38). The measure includes 13 items that ask about experiences of internalized stigma (e.g., It is hard to tell other people about my HIV infection. I am ashamed that I have HIV.). Items are on a Likert scale (0-4), and will be summed for a possible range 0-52.
Clinical outcomes and evidence-based birth practices will be assessed via medical record reviews and include mode of delivery (cesarean vs. vaginal), neonatal ART administration, completion of the 3 steps of Active Management of the Third Stage of Labor (AMTSL), use and timing of oxytocin, episiotomy, indication for cesarean section, and whether the patient was transferred in or out of the facility.

Primary outcome for providers
Practices of respectful maternity care will be assessed using a measure of person-centered maternity care among providers, which was validated in Kenya and Ghana (39). The 9-item measure asks providers how often they have done certain practices in the preceding one month (e.g., How often did you ask patients if they had any questions?). Items are on a Likert scale (0-3) and will be summed for a possible range 0-27.  (40), with additional items added speci cally for the childbirth setting. The 6item measure of stigmatizing attitudes towards WLHIV was developed by Nyblade based on eld testing in six countries (37). All items are measured on a 0-3 scale. A mean score will be calculated to assess overall stigma toward WLHIV, and separate domain scores will capture fear of HIV acquisition, extra precautions, and attitudes toward WLHIV.

Secondary outcomes for providers
Self-e cacy in delivering intrapartum care will be assessed using measures that PRONTO has used in previous evaluations (10). The measure includes nine items, and participants will be asked to rate their con dence in various clinical scenarios on a scale of 0-100 (e.g., How con dent are you in handling a postpartum hemorrhage? How con dent are you in talking to a woman and her family about her condition during an emergency?). Seven of the nine items will be repeated and asked speci cally in the context of caring for WLHIV (e.g., How con dent are you in handling a postpartum hemorrhage in a woman with HIV?). We will calculate mean scores (ranging from 0-100), separately for the general measure and for the measure speci c to caring for WLHIV.
Provider burnout will be assessed using a 2-item measure (41) that were adapted from the Maslach Burnout Inventory (42). The rst question de nes burnout as chronic emotional exhaustion in one's job, and asks how often the participant experiences this. The second question asks how often the participant has felt more callous toward people since they took their job. Both questions have response options on a scale of 1 ("Never") to 7 ("Every Day"). Participants will be considered to have symptoms of burnout if they score a minimum of 4 ("a few times a month") on either the rst question (emotional exhaustion) or the second question (depersonalization). This measure will not be evaluated as an outcome, but rather will rather be used as a confounding factor in providers' ability to provide RMC.

Quality Assurance (QA) Data
In order to assess the implementation of MAMA and therefore its potential for scalability, we will focus on intervention feasibility and acceptability, as de ned by Procter et al 12 . Feasibility will be assessed based on the proportion of the eligible providers who attend the intervention, and the proportion of the total intervention time attended. Attendance rosters will track who attends and for how long, and the intervention will be considered feasible if at least 65% of the eligible participants attend a minimum of 75% of the intervention time. During the trainings, we will have an individual dedicated to taking detailed notes of participants' discussion and input. These notes will be thematically coded to capture the quality and content of participation, and to identify areas to improve future iterations of the curriculum.
Acceptability will be assessed in the immediate post surveys with participating L&D providers.
Participants will be asked to rate the usefulness of the intervention for their clinical practice and their satisfaction with the intervention format and facilitators. The intervention will be considered acceptable if the mean scores of items are greater than 4.0 (on a scale of 1 to 5). Participants will also be asked to provide open-ended feedback about their favorite and least favorite aspects of the intervention and their suggestions for modi cation and scale-up. Responses will be thematically coded and summarized.

Statistical Analysis
Primary outcome: women giving birth We will rst compare the RMC full scale scores between pre-and post-intervention periods. We will use ordinary least square regressions (OLS) to model RMC scores with time point (pre-intervention or postintervention) as an indicator variable while controlling for demographic characteristics, clinical characteristics, and facility level factors. We will then repeat the analysis for the three RMC sub-scales (dignity and respect; communication and autonomy; supportive care). To assess whether HIV status moderates the impact of the intervention, an interaction term for HIV status and timepoint will be added in the OLS models. A signi cant interaction term will be evidence that HIV status moderated the effect of the intervention.

Secondary outcomes: Women giving birth
The outcome measures of perceptions of HIV stigma in the facility and internalized HIV stigma will both be analyzed continuously. OLS models will used to examine differences in scores in the samples enrolled in the pre-intervention period and the post-intervention period. The outcomes of postpartum HIV care engagement, clinical outcomes, and presence of evidence-based maternity practices will be analyzed categorically. Chi square tests will be used to assess differences in the proportion of WLHIV who had the presence of characteristics (e.g., poor care engagement, cesarean delivery) in the pre-intervention period, compared with the post-intervention period.

Primary outcome: L&D providers
To assess changes in practices of RMC among providers, we will use a random mixed effect model. We will model RMC scores with timepoint (pre-intervention, immediate post-training, pre in situ refresher, and two months post in situ refresher) as a predictor variable.

Secondary outcomes: L&D providers
The outcome measures of stigma towards WLHIV and self-e cacy in providing intrapartum care will both be analyzed as continuous variables. We will following the same analytic strategy as we have outlined for the primary outcome, using a random mixed effect model and including timepoint as a predictor variable. As a follow-up analysis, we will assess whether providers' reported burnout at baseline is associated with our primary and secondary outcome variables at baseline, and with change in the outcome variables over time.

Sample size considerations
A power calculation was conducted a priori, considering our primary outcome of RMC among WLHIV. Power analysis showed that we will need 103 WLHIV at each time point to be able to detect a moderate intervention effect (d = 0.4) with two-sided signi cance level of 5% and 80% power, hence a total of 206.
To allow comparison with HIV negative women, a similar sample size with some matching characteristics with WLHIV will be enrolled, but low response rate and some operational challenges are expected in his group; therefore, up to 200 HIV negative women will be invited to participate. The sample size for providers was based on convenience; we hope to reach at least 65% of all eligible providers in the six study facilities, which we estimate will provide a sample of 60 providers.

Dissemination
A study advisory board has been established to provide ongoing stakeholder input on the study and share emerging data and ndings. The board will be convened for three half-day workshops during the study: initially, for input on intervention content; mid-way, for feedback on the curriculum and preliminary ndings; and at the conclusion, for interpretation/dissemination of results. Advisory board members will be updated on the study progress.
At the conclusion of the study, we will conduct a feedback forum with a larger audience of stakeholders.
During the forum, the team will share the ndings of the study and facilitate a discussion about the implications of the data for future research and practice. Results will also be published in peer-reviewed journals and presented at appropriate scienti c meetings, including regional, national, and international meetings. Authorship eligibility guidelines will follow the authorship guidelines of the International Committee for Medical Journal Editors (www.icmje.org).
All study investigators, along with the data management team, will have access to the nal trial dataset.
Researchers from outside the team can request access; data can be shared with a data transfer agreement from the respective Institutional Review Boards and within the constraints required for the protection of con dentiality for study subjects.

Study organization
As the principal investigators of the MAMA intervention, Drs. Melissa Watt, Susanna Cohen, and Blandina Mmbaga are charged with co-leading the study. They will ensure the completion and integrity of the study by managing and monitoring study activities and the reporting of study ndings. They will facilitate collaboration between the University of Utah and KCMC by initiating and maintaining communication between these two institutions and the study staff at both locations. Drs. Watt, Cohen, and Mmbaga will monitor the ethical overall conduct of research activities and be responsible for overseeing compliance of nancial expenditures in accordance with sponsoring agency regulations.
The faculty investigator and PhD candidate in the study, Dr. Mlay and Mrs. Barabara, will bring expertise on maternal and infant care delivery and HIV engagement to the MAMA intervention. They will support the scienti c oversight of the study, meeting weekly with study staff and providing on-going supervision and support.
A minimum of one data collection staff member will be the point person for each of the clinical sites and be responsible for recruiting participants and obtaining study data through surveys (using ACASI technology) and qualitative interviews. The data management team, led by statistician Linda Minja at KCMC, will be responsible for storing, analyzing, and interpreting quantitative data. The team will clean data and code measures at each time point in order to ensure that the data is valid and easily interpreted.
To elicit stakeholder input, we have established a study advisory board (see Dissemination section) that includes representatives from the Tanzanian Ministry of Health, leadership in the study clinics, community-based organizations, and members from the KCMC HIV Community Advisory Board.

Discussion
The MAMA intervention is a provider training that was designed to help providers develop interpersonal and clinical skills to support the births of WLHIV. The intervention is based on the PRONTO model of simulation training, which has shown signi cant impacts on clinical outcomes(28, 29,43,44) and RMC practices (9) in multiple countries. This is the rst time that the model will be adapted to support the births of WLHIV. In this study, we will deliver the MAMA intervention to 60 providers and assess outcomes in WLHIV (comparing patient-reported and clinical outcomes in WLHIV in the periods before and after the training), as well as outcomes in participating providers (comparing knowledge, attitudes and practices before and after the intervention). The implementation ndings will be used to nalize the intervention for a train-the-trainer model that is scalable, and the outcomes data will be used to power a study to detect signi cant differences in HIV care engagement.
As we implement this study, we are aware of some key limitations of the study design. First, we do not have preliminary data to predict the mean and variance of our primary outcome; therefore, it is possible that our study is under-powered to detect signi cant differences. Second, the measurement of a key secondary outcome (HIV care engagement) is dependent on follow-up of patients following discharge from the labor ward. Given the time demands of women in the early postpartum period, and the possibility that some women may relocate to have additional support in caring for a new child, we may have challenges in retaining women for the follow-up survey. Similarly, we may face challenges in retaining providers, especially if they are reassigned to other clinical settings. Finally, the impact of the training on patient outcomes is dependent on a critical mass of providers receiving the training. Although we will offer the training twice and we will provide an honorarium for participating providers, we acknowledge that we may face challenges in getting the majority of providers trained in the MAMA intervention. Despite these limitations, this study will provide important data both to modify the intervention for future scalability and also to inform a future clinical trial.
In conclusion, WLHIV who are giving birth are a vulnerable population. While much work has been done to improve the antenatal care continuum for WLHIV through prevention of mother to child transmission programs, the clinical juncture of childbirth for WLHIV has received scant attention (45). This study will ll an important gap by building the capacity and skills of L&D providers to provide evidence-based, respectful intrapartum care to WLHIV. If successful, the MAMA intervention can be scaled-up to be part of the training curriculum for midwifery students and other L&D providers.

Acknowledgements
We are grateful for the support of the