The Video intervention to Inspire Treatment Adherence for Life (VITAL Start): protocol for a multisite randomized controlled trial of a brief video-based intervention to improve antiretroviral adherence and retention among HIV-infected pregnant women in Malawi

The trial will take place in Malawi, a landlocked country of 17 million people in sub-Saharan Africa. Malawi’s per capita gross national income is $340 USD, and it has an adult literacy rate of 62% [58, 59]. Almost all pregnant women (95%) attend an antenatal care (ANC) clinic at least once in pregnancy, and HIV status is determined in more than 90% of ANC attendees through a routine opt-out approach [3, 60]. Adult HIV prevalence in Malawi is approximately 11%, and combination ART is free for all patients with a confirmed diagnosis (defined as two serial positive results from rapid antibody-based tests on whole blood). The most commonly used first-line regimen for all patients (including pregnant women) is a once-daily fixed dose combination of efavirenz, lamivudine, and tenofovir disoproxil fumarate.

The trial will take place at three sites: two urban health facilities in the capital of Lilongwe (Area 25 and Kawale Health Centers) and one rural district hospital in the southern district of Mangochi (see Table 1). All sites are high-volume public Ministry of Health (MOH) facilities that were selected in consultation with the MOH.

We will recruit 892 pregnant women living with HIV presenting to ANC clinics. Research assistants (RAs) assigned to each facility will deliver health talks to all patients attending ANC before the clinic begins, describing the trial and inviting interested patients to participate. RAs will also work with facility staff to identify eligible participants. To be eligible for trial participation, individuals must meet all of the following inclusion criteria:

Exclusion criteria include any of the following:

Participants will be randomized in a 1:1 fashion to the SOC or VITAL Start arms. In terms of HIV/ART care received, the only difference between the arms will be the type of pre-ART counseling received.

Participants randomized to SOC will receive the SOC pre-ART initiation education from a clinic health worker using the National ART Counseling Flip Chart [61], an educational tool with pictures on one side and key message prompts on the other. Topics include the basics of ART and the importance of adherence. Complete review of the flip chart typically takes one hour. To reduce contamination, VITAL Start and SOC will be delivered in separate areas.

Participants in the VITAL Start arm will receive a 37-min intervention (27-min video and 10-min scripted RA-delivered individual counseling session). The video, titled Chiyembekezo (“Hope” in Chichewa), was developed through an iterative multistep process involving a team of persons living with HIV, researchers, clinicians, MOH staff, and other stakeholders. In conceptualizing the video, three behavior-determinant models were selected based on their demonstrated ability to promote behavior change in HIV/STD interventions:

Development of video content was through a community participatory approach to ensure that the content was acceptable and engaging. Focus groups with people living with HIV/AIDS and local clinicians identified three key subject areas the video should target: (1) the importance of starting lifelong ART while feeling healthy; (2) management of ART side effects; (3) partner disclosure.

An advisory group of HCWs, ART providers, and MOH staff consulted at all stages of video development. Content was generated through a partnership with In Tune for Life, a nonprofit organization with expertise producing health promotion videos. The finished video depicts an urban Malawi setting with a female protagonist who is pregnant and newly diagnosed with HIV at ANC clinic. The film depicts her anxieties about her health, protecting her baby, disclosing to her husband, and remaining on ART for life. A nurse and close friend encourage her to disclose her status to her partner and remain adherent. Through their reassurances, the protagonist is empowered in her relationship with her husband and prepared to take ART for life.

Interventions for both arms (VITAL Start and SOC) occur upon trial enrollment at the ART initiation visit. Due to the one-off nature of the trial intervention, there are no criteria for discontinuing or modifying allocated interventions for a given trial participant after randomization. Participants will not be allowed to switch arms if they desire a different intervention after allocation. All other clinical care will be carried out according to MOH guidelines.

To help ensure treatment fidelity and reduce contamination, we have adopted practices recommended by the National Institutes of Health (NIH) Behavior Change Consortium Treatment Fidelity Workgroup in five areas: trial design, training, delivery of treatment, and receipt of and enactment of treatment skills [62]. Multiple strategies were developed, tested, and revised; key strategies are listed in Table 2. In brief, we will use two fidelity checklists to monitor, measure, and ensure fidelity to the VITAL Start intervention. The two checklists are an Observer Intervention Fidelity Checklist (completed by an RA observing the VITAL Start intervention) and a Self-Administered Intervention Fidelity Checklist (completed by the RA who is administering the intervention). An overall fidelity score will be determined to assess the level of fidelity maintained based upon four key measurement criteria as outlined in Dusenbury, et al. [65]: adherence, dose, quality of delivery, and participant responsiveness. Other fidelity measures are outlined in Table 2.

After enrollment, participants will attend trial visits at months 1, 3, 6, and 12 for a total of five trial visits. Participants will be provided with the equivalent of $5 USD for transport reimbursement for each trial visit. Follow-up surveys will be completed at a time and place separate from routine ART refill visits.

A full list of measures and a schedule of assessments are outlined in Table 3. See also the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) checklist, provided as Additional file 1. The primary outcome is a composite measure of retention and viral suppression at 6 and 12 months. Participants will be considered retained in care if they are verified by clinic records as active on ART within 2 months of the last trial visit. Viral suppression will be measured as viral load < 1000 copies/mL. Retention was chosen as a primary outcome due to its strong association with adherence [81]. Viral suppression as a marker of adherence was included in the primary outcome because viral suppression is the clinical goal of ART and leads to decreased HIV-related mortality and morbidity [82].

We will also examine several secondary outcomes. Adherence will be further measured by self-reported adherence, pharmacy refill data, and tenofovir diphosphate concentration at 12 months. Tenofovir diphosphate (TFV-DP) is a metabolite of tenofovir that allows estimation of average adherence over the preceding 2 weeks. Use of TFV-DP will allow detection of small, clinically significant differences in adherence between groups that may be missed by viral load testing alone [83].

To critically examine the implementation of VITAL Start, we will also examine participant satisfaction, intervention fidelity, contamination, and time-motion data. Guided by social action theory [84], we will also assess a series of behavioral, mental health, and psychosocial measures during the trial as potential mediators or modifiers of the intervention’s effect as well as potential independent factors, which may influence adherence behaviors. These include assessments of HIV knowledge, whether a patient is a new ART initiator or re-initiating ART, measures of depression, alcohol and substance use, intimate partner violence, and availability of social support. Finally, we will collect financial data alongside clinical outcomes to assess the cost-effectiveness of VITAL Start compared to SOC.

The issue of distinguishing trial retention from retention in clinical care deserves special mention in this trial, where retention in clinical services is a primary outcome, and efforts to promote trial retention may have unintended consequences. Trial staff will collect telephone numbers and/or physical address details from participants willing to provide this information. When a participant is absent for a trial visit, staff will trace her first by telephone, and will make three valid attempts (defined as speaking with the participant) or will attempt daily for a week, whichever comes first. After phone tracing, staff will attempt to visit the participant at home, and will make up to three valid attempts. Participants will be allowed to complete trial questionnaires at home during these tracing attempts if they desire and a secure location is available for the interaction. All tracing attempts will be recorded to allow analysis of tracing attempts according to trial arm and controlling for tracing attempts, if necessary.

Participants who transfer to a different facility will not be terminated from the trial unless they withdraw consent or are found to have died. This decision reflects the highly mobile nature of the trial population that we have observed in previous studies. By keeping participants enrolled, they can easily transition back into the trial if they move back to the catchment area or are willing to continue attending trial visits. If a subject is discontinued at any time after entering the trial, the trial team will try and ensure that the subject completes the Final Visit survey and procedures.

The sample size calculation is based on a comparison of the proportions of those in the intervention versus SOC arms who reach the combined primary endpoint at 12 months after maternal ART initiation. A combination of various scenarios was considered for retention in care (50–90%), maternal viral suppression (50–90%) among those retained, and contamination rate (0–15%). A difference of 15% was selected as the smallest difference that would be of clinical significance.

For the primary analysis in all subjects, accounting for up to a 15% contamination (pilot data suggested 0%) and 5% attrition rate (based on transfer-out rates from MOH data at trial clinics in 2016) [3], and assuming a 1:1 allocation ratio to achieve 90% power to detect a minimum difference of 15% between the control and intervention groups, 796 subjects are required (398/398 per arm) at a significance level of 0.025 to account for multiple comparisons. We further increased this by 12% to 892 total (446 per arm) to account for a potential change to national guidelines to recommend dolutegravir-based regimens as first-line ART. The higher potency of dolutegravir-based regimens may allow higher rates of viral suppression at lower adherence levels, therefore improving overall viral suppression and making a difference between arms more difficult to detect.

Participants will be randomly assigned in a 1:1 fashion stratified by clinic proportional to the number of potential eligible participants at that site. Smith’s randomization algorithm will be applied for random assignment to VITAL Start or SOC to reduce imbalance throughout the randomization process [85]. The primary trial biostatistician will generate serial random codes for each clinic and place them on a sequential list that will be maintained by the trial coordinator. Assignments will be provided to RAs in opaque sealed numbered envelopes to be opened sequentially prior to administration of the baseline survey. If a participant refuses to participate in the trial after randomization, the next subject will not replace her. The randomized participant who refused participation will be included in the intention-to-treat (ITT) analysis. Due to the nature of the intervention, RAs, clinic staff, and participants will not be blinded. Separate research staff (outcome assessors) will remain blinded and conduct the outcome assessments including follow-up surveys. They will not be informed of the participants’ group assignment. All trial investigators will also remain blinded. After data collection is complete and the database cleaned and locked, a research team member will break the randomization code to input the group allocation within the pre-existing data set and enable between-group analyses.

Strategies for ensuring fidelity to the VITAL Start intervention are described in detail in Table 2. Fidelity will be quantitatively assessed by both the RA performing the counseling and an RA observing the counseling session. Contamination may occur if participants who receive the VITAL Start intervention discuss elements of the video with participants who received standard of care (SOC). Pilot data suggested little to no risk of contamination with the contamination reduction procedures that will be used in this trial (Table 2). However, in case contamination does occur, we will assess the prevalence, magnitude, and source of contamination. Participants will complete a contamination assessment at months 1, 6, and 12, and we will account for contamination via a contamination-adjusted ITT analysis [64].

All data will be collected on electronic case report forms (CRFs) by trained trial staff using electronic CRFs on tablets and will be monitored weekly via data checks by the study team. We elected to use electronic CRFs to improve both efficiency and data quality through the use of internal checks and automated skip patterns [86]. Backup paper CRFs will be maintained at all sites. The use of tablets also allows us to use an audio computer-assisted self-interview approach to assess self-reported adherence to medication, theoretically improving participants’ comfort with responding truthfully to these questions [87]. RAs will administer surveys at baseline; all follow-up surveys will be conducted by blinded outcome assessors at a location separate from the ART clinic where they receive their ART medications.

All tablets will be encrypted and password-protected and will use a secure connection to transfer data to a central server on a daily basis. A full list of data sources and schedule of assessments can be found in Table 3.

The primary analysis will be conducted as ITT. If there are exceptional circumstances which could lead to excluding randomized subjects from the ITT population without the possibility of introducing bias (e.g., an ineligible patient is mistakenly enrolled), a modified ITT (mITT) analysis will be used as the primary analysis. A statistical analysis will be performed at the end of the trial; no statistical interim analysis is planned.

Baseline characteristics by arm will be displayed using descriptive statistics. The primary outcome measure (maternal ART retention and viral suppression) is a binary variable, and it will be compared by chi-square tests at 12 months, and followed by logistic regression to study the association between intervention and outcome while controlling the baseline prognostic factors of the main outcome. The main exposure variable and any factor deemed to be clinically relevant will be forced into the model regardless of the statistical significance. Secondary outcomes include components of the composite outcome, self-reported adherence, pharmacy refill data, ART knowledge and self-efficacy, motivation and behavior skills, partner disclosure, and social support, etc. For the 12 months outcome, chi-square tests and two-sample t tests will be used. For repeated measures, the generalized estimating equations (GEE) method and linear mixed models will be used to examine differences in trajectories over time between the intervention and control groups while accounting for the correlation among repeated measures from the same subject. Additionally, for time-to-event outcomes, such as time to partner disclosure, we will conduct survival analysis using the Kaplan-Meier (KM) method with log-rank test and Cox proportional hazard models including the covariates. A p value of 0.05 will be deemed statistically significant and a p value of 0.1 regarded as a statistical trend. All analyses will be performed using SAS software (SAS Institute, Inc., Cary, NC, USA).

The MOH will be the main implementer of the VITAL Start intervention as well as SOC pre-ART counseling. Therefore, we will assess costs from the provider (MOH) perspective. We will use both micro- and macro-based approaches as appropriate to identify and quantify intervention resources [88‐90]. Retrospectively, we will extract information on resources expended for video design and production from inventory and other program documents. Unit prices will be obtained from account records or from local retailers. To measure staff time devoted to pre- ART counseling, we will perform time-motion assessments in line with the Suggested Time And Motion Procedures (STAMP) [80, 91, 92]. We will sample sufficiently to account for potential sources of heterogeneity such as trial clinic and provider. The value of staff time will then be estimated as the product of their gross salary and share of time allocated to intervention/pre-ART counseling.

We will calculate the incremental cost-effectiveness ratio (ICER) for the primary outcome of economic interest. The ICER is the difference in costs between two interventions divided by the difference in their effects, and it can be interpreted as the incremental price of a unit health effect from the intervention under study, as compared to the alternative. We will conduct one-way and multiway sensitivity analyses to model cost-effectiveness under various conditions, such as lower pricing of video production, mode of delivery, tablets, and alternate health care worker (HCW) cadres to deliver the intervention. Finally, we will conduct probabilistic sensitivity analyses using bootstrapping methods to ascertain the robustness of our results to extreme assumptions/scenarios. Based on emerging literature [89, 93], we will determine cost-effectiveness by comparisons with published ICERs for other commonly implemented HIV interventions in Malawi and southern Africa, as well as with gross domestic product (GDP)-based thresholds (0.5× and 1× the Malawi per capita GDP of $487) [94].

Blinded data monitoring will be conducted using descriptive statistics for quality assessment purposes, such as monitoring subject accrual and compliance as well as identifying outstanding values, etc., for data quality control. The results will be displayed for follow-up measures for the combined group, and otherwise by blinded treatment groups. We will monitor for overall trends in retention weekly. All unanticipated problems and adverse events will be recorded on designated CRFs and reported to the principal investigator (PI) and appropriate regulatory bodies. All potential adverse events will be reviewed weekly by the PI, trial coordinator, and research supervisors. The data manager will review trial data weekly. A stakeholder advisory group (SAG) comprising trial site supervisors, trial site clinical staff, and community members will meet twice a year to provide input on study implementation. In addition, an external monitor will evaluate trial operations annually.

Measures to ensure intervention implementation and fidelity are detailed in Table 2.

This trial has been approved by the Baylor College of Medicine Institutional Review Board (IRB; protocol H-39785) and the Malawi National Health Sciences Research Committee (NHSRC; protocol 16/05/1593). If the investigators elect to make important modifications to the protocol, amendments will be submitted to both the Baylor IRB and Malawi NHSRC.

Informed consent will be obtained in writing from all participants by trial site staff using forms in Chichewa approved by both the Baylor IRB and Malawi NHSRC. A staff member fluent in Chichewa will be responsible for reviewing the form in detail and eliciting any questions before the participant signs. If the patient is unable to read, staff will read the consent form to the participant in its entirety, the participant will mark her thumbprint to indicate consent, and the process must be witnessed by a third party not involved in the trial.

Confidentiality will be prioritized and will include the following measures specific to our intervention and trial population:

All participants will receive free lifelong ART as per Malawi MOH guidelines. Lab results from the trial (viral load measurements and TFV-DP concentration) will be provided to participants after the final trial visit in writing, along with a brief interpretation of the results for the participant’s clinician.

We will ensure that trained psychosocial counselors will be available to support participants who screen positive for current intimate partner violence, depression, or suicidality.

Results will be reported to the MOH and local stakeholders, as well as at local and international conferences. Results will also be submitted to peer-reviewed journals. Authorship eligibility will be guided by the International Committee of Medical Journal Editors criteria [95].

The protocol is version 4.0 dated 23 August 2018. Trial enrollment started in October 2018, and recruitment will be completed in March 2021.