1.1 Setting
The VIBRA trial will be conducted in the districts of Butha-Buthe and Mokhotlong, in northern Lesotho, in the catchment areas of 22 health facilities. Both districts are characterized by mostly rural settings with an estimated population of 220,000, mainly living in villages scattered over a mountainous area of 5,842 km2.
This trial utilizes the longstanding VHW country program. VHWs are members of and appointed by the community to provide a package of basic services at the household level, although they have no formal professional health education. They are elected by the village members, complete a 2-weeks training followed by periodic refresher courses, and are supported and supervised by the health center staff of the corresponding catchment area. Most are employed by the Ministry of Health, and receive a monthly stipend of USD 20.
1.2 Design
The VIBRA trial is a cluster-randomized superiority trial. The trial is linked to another trial, the HOSENG (HOme-based SElf-testiNG) trial. Together, HOSENG and VIBRA constitute the GET ON (GETing tOwards Ninety) research project. Reasons for this interlinked design are that both trials rely on interventions involving VHWs, who need to be randomized and specifically trained, and that the HOSENG trial provides one of the recruitment platforms for VIBRA. Thus, the two trials are based on the same cluster-randomization and run in parallel. This design allows us to assess the entire HIV care cascade in one larger project.
The rationale for a cluster-randomized design are the reliance on the VHWs and the high risk of cross-contamination between the study arms if randomization would be done at individual level.
1.3 Cluster-randomization, screening of study participants, eligibility and interventions
Before trial start, the eligible clusters (i.e. villages) are randomized into VIBRA control or VIBRA intervention clusters, stratified by district, village size (≥30 versus <30 households) and access to the nearest health facility (easy versus hard to reach, defined by needing to cross a mountain or river, or >10 km away from health facility). Details about cluster eligibility, cluster sampling, cluster-randomization and the HIV testing campaign are described in detail in the HOSENG study protocol published separately. In short, different campain teams consisting of counsellors and one study nurse visit rural villages in the two study districts. The teams will propose HIV testing and counselling and multi-disease screening and prevention. Household members who are eligible for and consent to testing undergo HIV testing by the counselors according to national HIV testing guidelines.48 All household members with a confirmed HIV-positive result and not taking ART, will be screened by the study nurse for VIBRA eligibility according to the criteria in Table 1.
If a patient is eligible for VIBRA, the study nurse will offer same-day home-based ART initiation and propose follow-up care according to the assigned respective cluster. As successfully implemented through our previous trial and recommended by the national guidelines, same-day home-based ART initiation, using the national standard first-line ART regimen, will be performed in both arms.19,24 If individuals are not eligible for the VIBRA trial and, thus, not eligible for same-day standard first-line ART initiation, they are referred to the health facility. Features of same-day ART initiation are outlined in Table 2.
1.3.1 Intervention clusters
The participants in the intervention clusters are offered the two features of VIBRA model: The first feature is the possibility of village-based ART visit/refill through the VHW with routine clinic visits only at 6 and 12 months after ART intiation. The second is the offer of receiving a tailored short text-message (SMS) intervention. Figure 1 summarizes the VIBRA model.
If participants in the intervention clusters choose the village-based ART visit/refill, they receive an appointment for a first clinical visit at the VHW 12 to 16 days after the home-based ART initiation. At each visit the VHW follows the same paper-based pre-specified checklist (provided as online supplement 1: p34-37) written in the local language (Sesotho). By following the checklist the VHW documents a) patient's symptoms to alert them to a potential drug toxicity, opportunistic infections, immune reconstitution inflammatory syndrome, b) adherence to ART, and c) any visits at other health facilities. In order to ensure safe and high-quality clinical management, participants in intervention clusters will not only be linked to their VHW, but will also be under responsibilty of the Community ART Nurse (CAN) of the corresponding district. CANs are nurses who are experienced in HIV care. One CAN per district has been hired. The VHWs and CANs will have a list of the participants, for whom they are responsible. If any question on the checklist triggers an alert, the VHW informs her/his CAN. Similar to the health facilites, the VHW provides drug supply for 1-3 months at each visit. Participants are, however, encouraged to visit the VHW or the clinic at any time when problems or questions arise. Six months after ART initiation the participant must attend the clinic for the first time for laboratory assessment. VHWs have monthly meetings at the health facility together with a designated facility staff member. The VIBRA model will utilize these existing meetings and the CAN (or representative) will join and provide support. These meetings provide the platform for review of patient files and patients can be up-referred (to the health facility) or down-referred (to the VHW). If a patient misses an ART visit, he/she will be traced by the VHW using a standardized tracking tool (provided as online supplement 1: p39). All VHWs in the intervention clusters will be trained to deliver the VIBRA model: a) dispense ART (and other co-medication such as cotrimoxazole), b) screen for ART-related adverse events and drug-toxicities, c) screen for co-infection (especially tuberculosis), d) assess adherence, e) understand referral algorithm in case of clinical deterioration, f) address disclosure and keep confidentiality, and g) perform basic data entry on the checklists. This training will last for 2-3 days. Every VHW keeps a list of patients he/she is responsible for and will only be allowed to dispense ART to participants on the list.
If participants in the intervention clusters choose the SMS intervention, they will receive monthly reminder SMS in Sesotho to adhere to ART (“Take your medication regularly as prescribed and don’t run out of medication”) and a viral load (VL) result-triggered SMS after the 6- and 12-months follow-up visit:
- If undetectable VL (<20 copies/mL): “Congratulations, your lab test was good. Keep it up!”
- If detectable VL (≥20 copies/mL): “Your lab test results are back. Make sure to come to the health facility as soon as possible and remind the nurse about your lab test.”
- If technical failure of VL measurement: “The lab test was unsuccessful. Make sure to come to the health facility as soon as possible and remind the nurse about your lab test.”
In order to maintain participant confidentiality, messages will not explicitely mention HIV or HIV care. Participants are not asked to confirm receipt of messages or to reply and can at any time choose to opt out from receiving messages.
1.3.2 Control clusters
Participants in the control clusters are offered standard of care, i.e. ART visits/refill at the health facility and no SMS intervention. They receive an appointment for a first clinic visit within 12 to 16 days of the home-based ART initiation. The health facility staff fills in the same pre-specified checklist as the VHWs at every visit. Study participants will not be offered any other differentiated delivery models.
1.4 Endpoints
The primary endpoint is viral suppression (<20 copies/mL) at 12 months, defined as the proportion of all participants with a suppressed VL 12 months (range: 10 – 15 months) after enrolment. VL will be measured in plasma using COBAS TaqMan® HIV-1 Test, v2.0, Roche Diagnostics. Secondary and exploratory endpoints as well as the long-term follow-up are outlined in Table 3.
1.5 Additional research within the project
We will conduct biomolecular research within this project. We will assess prevalence of major drug resistance mutations (DRM) in baseline samples and on all samples with unsuppressed VL at 12 months. Participants who start ART at home during the testing campaign but subsequently never link to care will be specifically traced to assess development of DRM.
Qualitative research is planned alongside the project to provide important contextual data and an in-depth exploration of community response to the intervention. For a qualitative case-control study a random sample from the VIBRA intervention clusters will be chosen. Cases will be participants who refuse village-based ART refill through the VHW, controls will be participants who accept village-based ART refill through the VHW. Moreover, we will conduct standardized interviews with a random sample of all stakeholders involved in delivering this new ART care/delivery model.
We will perform a system impact evaluation and cost-effectiveness analysis, in order to estimate the impact of the VIBRA intervention on health benefits and costs. First, we will assess the direct costs of the interventions. Secondly, we will assess the cost-effectiveness of the VIBRA model. Thirdly, we will assess the economic burden of the interventions to the participants, i.e. including both direct costs and the opportunity costs of their time. The assessment of the direct costs evaluated includes staff costs (campaign staff, clinical staff, laboratory staff, VHWs, CANs), personnel training costs, the cost of equipment needed (costs of HIV tests, ART and other used drugs, laboratory costs including the point-of-care tests at enrolment), and non-medical costs to the participant. The VIBRA model is expected to reduce the number of clinic visits, due to VHW-based ART refill and fewer unscheduled visits because intervention leads to better sustained clinical outcome. This would decrease costs for the health system and the participants (i.e. time required to access care, lost working time while accessing care, additional expenses while accessing care).
1.6 Data collection and management, biologic material, and follow-up
The VHWs and the health care staff at the health facilities collect data from scheduled and unscheduled ART visits on standardized paper study forms (Case Report Forms [CRF]), that act as source documents. CRFs will be collected regularly by the study team and entered into a password-protected database (MACRO, Elsevier). Similarly, relevant data for the SMS intervention will be entered and stored in a separate encrypted and password-protected online database, that offers the possibility to send out SMS automatically and is connected to the district laboratory database containing the VL results. The platform and data are stored on a dedicated server in a data center in Switzerland (Interxion, managed by Hostpoint AG), which meets FINMA-RS 08/07 requirements, is ISO-27001-certified, encrypts data in-transit with SSL and all patient names at-rest using OpenSSL with AES-256-CTR cipher method. Access to both platforms is strictly limited and regulated through personal user profiles. SMS are dispatched using the trusted third-party provider Twilio, headquartered in the United States and certified with the EU-U.S. and Swiss-U.S. Privacy Shield Framework. The consent forms will be stored in a secure way in the headquarter of the study center (SolidarMed Office in Butha-Buthe, Lesotho). Participant files will be maintained in storage for a period of at least 10 years after completion of the study.
Participants in all clusters undergo HIV testing and phlebotomy at enrolment, and phlebotomy at 6 and 12 months. For each participant, study-ID-coded blood samples will be stored at -80 °C at the laboratory of Butha-Buthe hospital. All samples collected fall under the biobank and material transfer agreement, approved by the ethics committees in Switzerland and Lesotho. Figure 2 displays the SPIRIT flow diagram with the overview of data collection, laboratory assessments and follow-up visits.
1.7 Sample-size
Based on data from the CASCADE trial, we expect the proportion of patients engaged in care with documented viral suppression 12 months after same-day ART initiation in the control arm to be approximately 50%. Assuming a 20% refusal/ineligibility rate, about 400 individuals need to be screened in order to identify 320 eligible individuals and 90% power to detect a 20% increase in the intervention group. We plan to enrol a minimum of 262 patients to ensure a minimum power of 80%. All sample size calculations were done assuming a type 1 error of 0.05 and an intra-cluster correlation coefficient of 0.015. Table 4 provides estimates of the sample size under varying conditions.
1.8 Analyses
Analyses will be performed following CONSORT guidelines for cluster-randomized trials49 and an intention-to-treat principle including all participants as randomized per cluster-randomization. Clusters are as unit of randomization, but individuals are set as unit of analysis. Multi-level statistical models will be used to adjust for the clustered data. The following analysis sets will be used in this trial:
- Intention-to-treat (ITT) set: All study participants will be evaluated according to cluster assignment at randomization
- Cluster per-protocol (CPP) set: This set includes all participants from clusters who completed the study without a major protocol deviation
- Individual per-protocol (IPP) set: This set includes all participants who completed the study without a major protocol deviation
The primary analysis for VIBRA study will be the comparison of viral suppression rates 12 months after offer of same-day ART initiation. The primary analysis will use a multi-level logistic regression model to assess the difference between the arms, adjusted for the pre-specified randomization stratification factors, the clustering according to village and other factors found to be unbalanced between intervention and control clusters.
Baseline characteristics will be presented according to randomized groups, no formal testing will be performed. Categorical variables will be described as absolute and relative frequencies and continuous variables as medians and interquartile ranges. As with the primary analysis, secondary endpoints will be analyzed with multi-level logistic regression model. All results will be presented as absolute differences, risk ratios and their respective 95% confidence intervals. Sensitivity analyses will be conducted as measure for robustness and will center on the deviations in model assumptions. The effect of sociodemographic and clinical determinants (age groups, gender, education status, employment status, WHO stage, CD4-count, TB status, CAGE status, HIV/ART history, HIV knowledge) on key study outcomes will be assessed in subgroup analyses. As the study is not powered for these pre-planned subgroup analyses, these results will be considered exploratory. Where data are missing in important covariates, multiple imputation will be utilized and results compared to models ignoring missing data.
The study protocol foresees a predefined snapshot analysis once all participants have completed 6-months follow-up. The snapshot analysis will entail all secondary endpoints assessed within the 6-month study period. All analyses will be done using Stata (version 14, Stata Corporation, Austin/Texas, USA), using 2-sided p-values and a significancel level of 0.05.
1.9 Monitoring, auditing, and data safety and monitoring board
At least one external monitoring visit will assess adherence to the approved trial protocol, accuracy of completed CRFs, and the electronic dataset. VIBRA trial represents implementation research, safety profiles of all used drugs are well-known, and the intervention does not include any new drugs. Thus, major adverse effects on patients’ health from this intervention are not expected, also given the encouraging results from similar trials in Uganda, Kenya and Tanzania50–52 and participants in the VIBRA model can opt to switch back or be referred to facility-based care at any time during the trial period. Therefore, it is not planned to establish a data safety and monitoring board. However, a separate, detailed safety monitoring plan has been developed to handle (Serious) Adverse Events ([S]AE), in-line with Swiss and Basotho ethics regulations. (S)AE will be graded according to the Division of AIDS Table for Grading the Severity of Adult and Pediatric Adverse Events, Version 2.0., November 201453 and managed according to study sites standard procedure following the national guidelines.19 The study physicians are responsible for all safety procedures. If a participant develops an AE of Grade 2 or higher at last study visit, he/she will remain under observation by the study physicians even after study termination, until the AE is resolved or stabilized.