The studyis a pilot randomized control trial aimed at determining the feasibility, acceptability, and potential efficacy of an HIV stigma counseling intervention for individuals entering antenatal care. The study has two parallel groups: control and intervention. The control group will receive standard of care from the clinical sites; the intervention group will receive standard of care paired with up to three Maisha counseling sessions. The allocation ratio for these parallel groups is 1:1. Table 1 summarizes key elements of the study, and Figure 1 illustrates participant flow through the study.
Ethical approval and registration
The study has been approved by the ethical review committees at Duke University, Kilimanjaro Christian Medical Center, and the National Institute for Medical Research in Tanzania. It is registered at ClinicalTrials.gov (NCT03600142).
The study will be conducted in two government health facilities in Moshi municipality, Tanzania. The Majengo and Pasua Health Centers together see approximately 2,500 pregnant women per year; an estimated 4.8% of pregnant women seen at the clinics are living with HIV. All patients are required to have an HIV test at entry to ANC, unless they present a clinic card confirming that they have previously tested positive for HIV. Pregnant patients are strongly encouraged to bring their male partner to the first ANC visit for pregnancy education as well as partner HIV testing.
The study will enroll 1000 women into the study. Women will be enrolled at entry into ANC, prior to receiving a routine HIV text; of the 1000 female participants, we expect that approximately 50 will be established or newly diagnosed as HIV-positive. Based on record review at the study clinics, we estimate that 70% of female participants will attend the ANC visit with their male partner, allowing for an enrollment of up to 700 men. The sample size was selected in order to have adequate power to detect differences in our outcome of HIV stigma attitudes for our HIV-negative clients, and to have pilot feasibility data related to outcomes for HIV-positive clients (12,13).
Participants will be enrolled prior to attending their ANC appointment, which is where routine HIV testing occurs. Following written consent to participate in the study, participants will complete the baseline survey and be randomized to a condition. Following the survey and Maisha 1 (if randomized to intervention arm), they will return to the clinic for standard of care services, including HIV testing.
Participants will be eligible for the study if they meet the following criteria:
- Minimum 18 years of age
- Women: Pregnant and attending first ANC appointment for current pregnancy
- Men: Accompanying a partner to her first ANC appointment (Note: the partner must be enrolled for the man to be eligible to participate)
- Able to understand and speak Kiswahili
- Able to provide consent
Screening and recruitment
First ANC attendees and their partners will be identified in the clinic waiting room by the clinic nurses. The nurses, in cooperation with research staff, will provide a brief description of the research activities, including the time commitment. Individuals who are interested in learning more will meet with the research team in a private research office, either alone or with their partner. The research staff will confirm the individual(s) meet the eligibility requirements, will clearly explain the study, and will obtain written informed consent. No biological specimens will be collected in this trial; therefore, consent for specimen storage is not applicable.
Contact information will be gathered from participants in order to facilitate the scheduling of follow-up assessments; this information will be securely stored in a locked file cabinet, separate from other participant information and data, and accessed by authorized study staff only.
Baseline data collection
After providing consent, all participants will complete a structured survey using audio computer-assisted self-interview (ACASI) technology on tablets running Questionnaire Development System (QDS) software. ACASI modality can ensure participant privacy and improve data validity by minimizing social desirability bias (14). Participants will complete the assessment on individual tablets where they can read (on the screen) and listen (through recorded audio) to the questions and response options in Kiswahili. As response options are read aloud, the corresponding text will light up on the screen. Participants can select their response using the tablet touch screen. The computerized assessments are programmed to skip questions that are not applicable based on previous question responses. Validity check items (e.g., “For this question, choose strongly agree”)) are included throughout to assess for data quality and participant attentiveness. A research staff member will be present in the room to answer any questions and provide assistance as needed. The data files will be securely transferred each day to the local data manager, who will store the files in a centralized location on a secured drive and review the files on a weekly basis for any quality issues. Stored data will be identified by a unique ID only, with access limited to authorized investigators and staff.
Upon completion of the baseline assessment, the participant will be randomized to receive either the standard of care HIV testing and counseling, or the standard of care plus the Maisha intervention. Female participants will be randomized at a 1:1 ratio using a block randomization method (10 per block) to ensure equal sample sizes by condition and to manage the flow of participants to the intervention condition. The allocation sequences will be prepared ahead of time by a statistician using an online randomization program (www.sealedenvelope.com). Sequences will be generated separately for each of the two study clinics. Study staff who are not involved in participant enrollment, assignment, assessment, or delivery of the intervention will prepare sealed, opaque envelopes for each study ID containing the randomized condition assignment. After a female participant completes the baseline assessment, she will open the envelope marked with her corresponding study ID and will learn her assignment; male participants will be assigned to the same condition as their partners. Participants will not be blinded to their allocation, as all participants will be aware of the additional time and activity commitments required as part of participation in the intervention condition. Since the trial is unblinded, emergency unblinding is not applicable. The research staff who give the participants their assignment envelopes will not know the randomization sequence until the conditions are assigned. Health care personnel at the study clinics will not be informed of participants’ study conditions, in order to prevent interference in standard of care delivery.
Control: Standard of care HIV testing and counseling
Participants randomized to the control condition will receive the standard HIV testing and counseling protocol in the clinic, which is administered by clinic nurses. The standard of care was chosen as the comparator in order to evaluate whether the Maisha intervention has an impact above-and-beyond the standard clinic procedures for HIV counseling and testing. According to the Tanzania PMTCT guidelines, HIV pre-test counseling should provide education about HIV and prepare a woman (and her partner, if present) for HIV testing (15). For anyone who tests positive for HIV, counseling should help the woman/couple to accept an HIV test result and discuss implications for treatment. HIV-infected women should be registered for PMTCT and immediately initiated on ARVs. HIV-infected men should be referred to the HIV care and treatment clinic (CTC) for same-day initiation of ARVs.
Intervention: Standard of care + Maisha
Participants randomized to the intervention condition will attend the standard of care HIV testing and counseling and will also receive the Maisha intervention. Maisha is a brief, scalable, theory-based counseling intervention that addresses HIV stigma at entry into antenatal care and includes up to three counseling sessions. The intervention model combines a stigma framework with principles of cognitive-behavioral therapy to address and mitigate the impact of stigma on health outcomes.In Earnshaw and Chaudoir’s HIV Stigma Framework, internalized, anticipated, and enacted stigma all intersect to undermine health-seeking behaviors (4,16). In developing the intervention framework, we observed that these components of stigma track onto the CBT ‘cognitive triad’ of negative beliefs about oneself, the future, and others/the world (17). Thus, the Maisha intervention addresses these three forms of stigma using principles of cognitive-behavioral therapy and formative work by Tshabalala and Visser (16,18).
Upon entry to ANC (for most, prior to learning their HIV status), we will deliver information, present the lived experiences of PLWH using a video, and promote self-reflection about community attitudes related to HIV. Participants will examine how community perceptions influence their beliefs about PLWH, including prejudicial beliefs that contribute to self-stigma in the event of a positive test. In sessions two and three, with participants who are HIV-positive (either presenting to ANC knowing their status or getting a new diagnosis of HIV), we will review the video content and provide additional structured counseling to address the difficult emotions and cognitions often associated with a positive status. Through linkage to the video, we will use Beck’s interventions for cognitive bias (19,20) as well as Third Wave behavioral concepts (21,22) to address one’s automatic negative thoughts about the self (internalized HIV stigma), the future (anticipated HIV stigma) and the world (enacted stigma). Intervention techniques include recognizing and reframing cognitive distortions, helping individuals to develop a positive self-schema, and promoting personal acceptance and value-driven behavior to reduce stigma and encourage positive HIV care engagement (Figure 2). Table 2 provides an overview of the Maisha intervention.
The development of the Maisha intervention was informed by our team’s previous research (2,3,23), qualitative interviews with patients and healthcare providers, and input from a study advisory board. In order to develop the video content, we engaged the local community advisory group, comprised of individuals who were living with or affected by HIV. These “expert patients” helped us to develop the video, which tells a story of a pregnant woman and her husband who both test positive for HIV and take steps to cope with the diagnosis. The actors for the video were selected from the community advisory group, and participated in the iterative refinement of the script during filming. Following the production of the video and the development of the intervention scripts, we conducted a trial run of the intervention with eight participants recruited from the ANC. Their input was elicited, which contributed to further modifications in the final intervention.
The Maisha intervention will be delivered by Bachelor’s-level counselors who have social work or counseling backgrounds. Counselors will receive a minimum of two weeks of training on counseling techniques and the intervention content, and thereafter will receive at least one hour per week of clinical supervision, with opportunity for additional supervision as needed. The counseling sessions will take place in a private research office at the clinic site.
The intervention counselor will complete a quality assurance (QA) and process rating form after each session. This form records issues raised in the session, coverage of session content, and feasibility of delivery. All intervention sessions will be recorded (with participant consent), and each week one session from each counselor will be reviewed during a group supervision session. Using a structured form (24), the counselors and supervising staff will assess the recorded sessions for intervention fidelity and presence of core components of counseling. During the group supervision session, the team will review the recording to discuss challenges and provide additional feedback and training.
For participants who are assigned to receive Maisha 2 and 3, the counselor will aim to deliver the Maisha 2 session on the same day as enrollment and Maisha 1 (i.e., the day of the first ANC appointment). If this is not possible (e.g., if the participant does not return after testing, or the participant does not have time to complete the session that day), the counselor will attempt to schedule the session for another day, ideally within 72 hours. The Maisha 3 session will be scheduled approximately two weeks after completing Maisha 2. In order to reduce the burden on participants and improve attendance, the counselor will try to schedule the Maisha 3 session on the same day as the participant’s next clinic appointment. The counselor will call participants to remind them of upcoming sessions. If a participant misses a scheduled session, the counselor will call to follow up and reschedule if possible. Participants may choose to stop attending Maisha sessions or withdraw from the study at any time.
During the study period, participants will continue with the standard antenatal care, which includes routine HIV pre-test and post-test counselling for unknown status. The intervention is intended to supplement, not replace, existing clinical care, and therefore does not impact use of or access to routine medical care.
A subset of participants (approximately 300 women and 200 men) will be selected to complete a follow-up assessment 3 months after completing the baseline assessment. All HIV-positive participants and their partners (if enrolled) will be contacted for follow-up. A subset of HIV-negative participants will also be eligible for follow-up. In order to observe changes in stigmatizing attitude scores among HIV-uninfected participants, individuals with stigmatizing attitude scores greater than 14 will be eligible for follow-up assessment; of those who meet criteria, a random 60% will be invited for follow-up.
Participants selected for follow-up will complete a structured post assessment, following the same procedures for ACASI-based data collection used for the baseline assessment. At the end of the assessment, all Maisha intervention participants will also respond to a short series of open-ended feedback questions about the intervention; this section of the assessment will be orally administered and audio recorded to fully capture participant responses. Responses to the open-ended questions will be directly translated into English from the audio, and entered into a REDCap database. For all HIV-positive study participants, data on HIV care engagement will be abstracted from their medical records and entered into REDCap using double data entry, allowing for data quality checks and secure data storage and transfer.
Participant tracking and retention
Research staff will conduct a daily review of the clinic ANC logs to record the number of potentially eligible participants who were not enrolled in the study. For enrolled participants, the HIV test results and estimated date of delivery will be recorded on the day of enrollment. The research staff will maintain tracking logs to record the dates when assessments and Maisha sessions are scheduled and completed. In preparation for the scheduled sessions, research staff will contact participants to remind them of upcoming appointments in order to ensure retention. If participants are unreachable or did not provide a contact number, research staff will consult the participant’s medical record for upcoming appointments and try to speak to them in person to schedule a follow-up visit. All participants will receive a transport allowance (5,000 TSh = approx. 2 USD) to facilitate their return to the clinic. For participants who fail to return for a scheduled follow-up appointment, the reasons will be documented and collated across participants.
Measures were selected based on previous research in East Africa, including measure validation when available, and evaluation of face validity by the Tanzanian researchers on our team. All measures were translated from English into Swahili and then back-translated and discussed to reach consensus on best translation. Table 3 summarizes the outcome measures that were assessed at baseline and 3-month follow-up, specific to the participant’s known HIV status.
Primary outcomes for HIV-positive participants
HIV care retention (women). Among female HIV-positive participants, retention in care at 3-month follow-up will be assessed via medical record review, with retention defined as having no more than a 60 day gap between PMTCT visits at the study clinic, or having record of an official transfer to another clinic (25). We will assess differences between conditions in the proportion retained in care at 3 months.
Internalized HIV stigma: Among HIV-infected participants, internalized stigma will be self-reported, measured by Scale A of the HIV and Abuse Related Shame Inventory (HARSI) (26), plus one added item. Internalized stigma will be measured at baseline (among individuals with established HIV diagnoses) and 3 months (among individuals with new and established HIV diagnoses). We will assess differences between conditions in mean scores of internalized HIV stigma scores at 3 months. For individuals with established HIV diagnosis, we will control for baseline scores.
Primary outcomes for HIV-negative participants
Attitudes toward people living with HIV (PLWH): Attitudes will be measured by self-report, using a modified version of Personal and Attributed Stigma Scale (PASS), which includes two sub-scales: blame/judgement and interpersonal distancing (27,28). The scale was adapted to the local context based on formative qualitative data collection and revised after a pilot of the measure with 88 individuals. Attitudes will be measured at baseline and 3 months; individuals with a known HIV diagnosis will not be assessed on their attitudes toward PLWH. We will assess differences between conditions in mean attitudes scores at 3 months, controlling for baseline scores.
Secondary outcomes for HIV-positive participants
ART adherence. Adherence to antiretroviral therapy at the 3 month follow-up will be measured by self-reported medication adherence; missing two or more pills in the past 30 days (<94% adherence) will be considered poor adherence. We will assess differences between conditions in the proportion with poor adherence at 3 months.
Depression: Depression will be measured by the Edinburgh Postnatal Depression Scale (EPDS) (29) for women and the Patient Health Questionnaire (PHQ–9) for male partners (30). Depression will be measured at baseline and 3 months. We will assess differences between conditions in mean attitudes scores at 3 months, controlling for baseline scores.
HIV disclosure: HIV disclosure will be measured by self-report of whether or not participants have ever disclosed their HIV status to a person outside of the heath care workers directly involved in their antenatal and PMTCT care. HIV disclosure will be measured at baseline (among individuals with established HIV diagnoses) and 3 months (among individuals with new and established HIV diagnoses). We will assess differences between conditions in the proportion with poor adherence at 3 months. For individuals with an established HIV diagnosis, we will assess differences in the change in proportion of participants who have disclosed their HIV status in each condition between baseline and 3 months.
Anticipated HIV stigma: For HIV-positive participants, anticipated HIV stigma will be measured using an adapted scale (31), which includes 16 items assessing the degree to which PLWH expect that they would experience prejudice and discrimination from others if their status were known. Anticipated stigma will be measured at baseline (among individuals with established HIV diagnoses) and 3 months (among individuals with new and established HIV diagnoses). We will assess differences between conditions in mean scores of anticipated HIV stigma scores at 3 months. For individuals with established HIV diagnosis, we will control for baseline scores.
Linkage to care (men): Among HIV-positive male partners, linkage to care at a Care and Treatment Clinic (CTC) will be self-reported at the 3 month follow-up, with linkage to care defined as having attended any CTC appointment. We will measure differences between conditions in the proportion of HIV positive men attending a Care and Treatment Clinic (CTC) for HIV care at 3 months.
Secondary outcomes for HIV-negative participants
Willingness to test for HIV in the future: Willingness to test for HIV in the future will be measured by self-report of whether or not participants intend to test for HIV in the next 12 months. Willingness to test for HIV will be measured at baseline and 3 months. We will assess differences in the change in proportion of participants who have disclosed their HIV status in each condition between baseline and 3 months.
Anticipated HIV stigma: For HIV-negative participants, anticipated HIV stigma will be measured using an adapted scale (31), which includes 16 items assessing the degree to which participants would expect to experience prejudice and discrimination if they were to be told they were living with HIV. Anticipated stigma will be measured at baseline and 3 months. We will assess differences between conditions in mean scores of anticipated HIV stigma scores at 3 months, controlling for baseline scores.
Quality assurance (QA) data
Participant satisfaction with the intervention will be assessed at 3 month follow-up via structured and open-ended questions about satisfaction with the intervention and facilitator, satisfaction with the timing and length of the sessions, ability of the intervention to address issues specific to participant’s experience and context, and suggested changes to the intervention. Intervention sessions will be recorded and a subset of recordings will be reviewed to assess whether core components of the sessions were completed, and to evaluate the effectiveness of the counselor in achieving session objectives.
Feasibility, acceptability, and potential efficacy of the intervention
Data analysis will follow guidelines of the CONSORT 2010 statement, as extended to pilot feasibility trials (32). Feasibility and acceptability of the intervention and the associated trial will be described by: recruitment and retention patterns, participant satisfaction and fidelity of intervention delivery. Retention will be monitored to calculate the percentage of eligible intervention participants who attend the Maisha 2 and 3 sessions and participants who complete the 3-month assessment. The team will document barriers to attendance for participants and will examine differences between participants who attend and those who do not. Participant satisfaction data at the 3-month follow-up will be described, with >80% satisfaction used as a metric of acceptability. Open-ended questions will be thematically coded to summarize participants’ perceptions of the intervention, suggestions for changes, and feasibility moving forward. The fidelity to the intervention will be assessed by examining the percentage of components from the Maisha session guides that are covered in each session.
Potential efficacy will be examined by analyzing separate outcomes for HIV-positive and HIV-negative individuals. For HIV-positive individuals, we are interested in differences between conditions in health outcomes (retention in PMTCT and medication adherence for women, linkage to CTC for men, depression) and stigma constructs (anticipated stigma, internalized stigma, and HIV disclosure). For HIV-negative individuals, we are interested in differences between conditions in stigma constructs (attitudes to PLWH and anticipated stigma), as well as willingness to test for HIV in the future. Additional analyses will examine changes in stigma attitudes by sub-scales (i.e., moral judgment and social distancing) and will include stratified analysis by gender.
For outcomes where there is a baseline measure, mixed-effects regressions will be used to model pre-post differences within and between arms, using a time by condition model specification (time, condition, and time*condition). Individual-level random intercepts will be used to account for correlation due to repeated measurement. Using a mixed-effects regression approach leaves flexibility to control for baseline outcome values that may not be balanced between groups due to small sample size, and may improve precision of treatment effect estimation. For outcomes where there is no baseline measure, we will examine differences in means or proportions with 95% confidence intervals. If we suspect, a priori, that baseline imbalance in prognostic covariates may be an issue, we will move into a regression framework. Given ACASI data collection methods, we expect low amounts of missing data. In cases of missing data, multiple imputation methods will be used.
Analysis of outcomes for the HIV negative participants will be completed using a subsample of enrolled participants who are selected to complete a follow up survey. Primary analysis will be done on Intention to Treat (ITT) including all cases that were selected for a follow up survey. We also conduct secondary analysis on per protocol (PP) basis considering only participants that received all intervention sessions and completed follow up. Furthermore, sensitivity analysis will be conducted to assess the effect of missing data on results. Some interim analysis will be performed to check study progress and monitor for adverse events. For clients who are HIV seropositive, apart from retention in care, impact on other outcomes (e.g., internalized HIV stigma) will be assessed only among clients who completed the baseline survey knowing their HIV status.
Given the low-risk nature of the counseling intervention, we do not anticipate any adverse events as a result of intervention exposure. Therefore, we will not have an independent data monitoring committee. The study team will monitor adherence to the protocol and will report any potential adverse events to the institutional review boards. To ensure the quality of our science, we will have daily check-ins with the field team to monitor daily enrollment of clients, challenges or issues with the consenting process, and any matters arising during survey administration and intervention delivery. The data manager will do a thorough review of the data on a weekly basis, and queries will be sent to research assistants for clarification. Any pressing issues in need of discussion or decision making will be discussed with the broader team during a weekly team call. Call minutes will be documented and shared with the full team. Due to low likelihood of adverse effect as a consequence of trial procedures, post-trial administration of counselling is not applicable.
The depression measure that is included in the survey includes a question that assesses for suicidal ideation (29,30). If the suicidal ideation question is endorsed, a coded message will be displayed to the research staff at the end of the assessment. The counselor will also make note if a participant says something during a Maisha session that indicates they are experiencing suicidal ideation. In either of these cases where ideation is indicated, the staff will follow a protocol to assess the risk, identify sources of support, make a plan for the participant’s safety, and make referrals to clinic personnel as needed. Research staff have been trained on procedures related to emotional distress and the assessment of personal safety and risk and will report any adverse events to the study coordinator and the local Principal Investigator for further follow-up.
A study advisory board has been established to provide ongoing stakeholder input on the study and share emerging data and findings. 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 findings; and at the conclusion, for interpretation/dissemination of results. Advisory board members will be sent a quarterly newsletter with updates on the study progress.
At the conclusion of the study, we will conduct a feedback forum with a larger audience of stakeholders from a variety of institutions, including clinic staff and patients, regional representatives of the Ministry of Health, HIV advocacy and service organizations, and women’s health organizations. During the forum, the team will share the findings 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 scientific 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); we do not intend to use professional writers in the reporting of study results.
All study investigators, along with the data management team, will have access to the final trial dataset. The observational data may be analyzed to answer research questions beyond our stated objectives, and researchers from outside the team can request access to the data. Data can be shared with a data transfer agreement from the Tanzanian National Health Research Ethics Review Committee and within the constraints required for the protection of confidentiality for study subjects. Shared data will not include identifiable information.
As the principal investigators of the Maisha intervention, Drs. Melissa Watt 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 findings. They will facilitate collaboration between Duke University and KCMC by initiating and maintaining communication between these two institutions and the study staff at both locations. Drs. Watt and Mmbaga will monitor the ethical overall conduct of research activities, and be responsible for overseeing compliance of financial expenditures in accordance with sponsoring agency regulations.
The faculty investigators in the study, Drs. James Ngocho and Jenny Renju, will bring expertise on PMTCT care delivery and mental health to the Maisha intervention. They will support the scientific 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 based at each of the clinical sites, and be responsible for recruiting participants and obtaining study data through surveys (using ACASI technology) and qualitative interviews. One counselor will be based at each clinical site, and will be responsible for delivering the Maisha sessions. 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.