Motivate

Brief Description

In order to eliminate new pediatric HIV infections, save maternal lives, and simplify antiretroviral therapy (ART) implementation in settings with generalized HIV epidemics, current WHO guidance recommends lifelong triple ART for all pregnant and breastfeeding women (Option B+). However, despite the promise of Option B+, key challenges including adherence to ART and continuous retention in HIV care exist.

In order to eliminate new pediatric HIV infections, save maternal lives, and simplify antiretroviral therapy (ART) implementation in settings with generalized HIV epidemics, current WHO guidance recommends lifelong triple ART for all pregnant and breastfeeding women (Option B+). However, despite the promise of Option B+, key challenges including adherence to ART and continuous retention in HIV care exist. As Option B+ is scaled up in Kenya, it is essential to identify effective methods to ensure long-term adherence and retention in care for mother-baby pairs, throughout pregnancy, breastfeeding, and beyond. MOTIVATE study being conducted at up to 28 health facilities and associated communities in Kisumu, Homa bay and Migori Counties, Kenya where MTCT rates prior to Option B+ rollout remained near 10%, despite the wide availability of PMTCT services. The study seeks to gain understanding of and address potential barriers at the individual, community, and health facility levels through formative research with HIV-positive pregnant and postpartum women, their male partners, and health care providers.

This information will be used to refine two proposed interventions, which are rigorously being tested; using a cluster randomized 2×2 factorial design. The evidence-based interventions being tested include:

1) Community Mentor Mothers (cMM) who provide support for ART adherence and retention in care for HIV-positive women in the community. They support uptake and retention in PMTCT services through home visits- four visits prenatally and nine visits postnatal- and other community outreach activities. This study is evaluating the use of a cMM approach to facilitate community support for HIV-positive women in order to enhance adherence and retention in care.

• Mentor mothers are HIV-infected women who have been through PMTCT and are tasked with providing peer education and psychosocial support
• They have been shown to increase uptake of services in several settings in sub-Saharan Africa
• In Kenya, mentor mothers have been based out of health facilities rather than in the community.

In order to improve adherence and retention in the new context of Option B+, the acceptability of community mentor mothers (cMMs) was explored between September and November 2014 and involved

•      4 Focus Group Discussions with health care providers
•      40 In-Depth Interviews with HIV infected pregnant women with their male partners
•      Transcripts were transcribed, translated and then coded into major and minor themes.

Majority of interviewed women, their male partners, and health care providers indicated overall acceptability of the planned community-based mentor mother strategy.

They appreciated the fact that cMMs are right there in the community, they already know the participants, and it wouldn’t always be necessary to travel to a clinic.

2) Individually tailored, theory-based mobile phone text messages. Our primary outcomes will include ART adherence at 12 months postpartum and retention in care. Secondary outcomes will include MTCT at 6 weeks, 12 months and 18 months; as well as maternal viral loads and CD4 counts as conducted during routine care. Results from this study will inform the scale-up of Option B+ in Kenya by identifying effective interventions that can maximize the potential of Option B+ with the aims of reaching the elimination of mother to child transmission of HIV and significantly improving maternal health. Mobile phone interventions can help to overcome individual-level barriers to adherence and retention in care. Our study will build on this successful pilot study in order to utilize theory-based behavior text messaging to support adherence and retention in the context of Option B+.

Participants in this arm of the study will receive one text message weekly from enrollments until 12 months postpartum. The text messages are tailored with the participant’s gestation and upon delivery the participants are transited to receive postnatal messages weekly up to 12 months

The study team monitors delivery of the text messages and failed text message delivery is tracked to establish the potential cause.

JUSTIFICATION

There is an urgent need to evaluate service models that maximize ART adherence and retention in care among women and infants in order for Option B+ to achieve its full potential and result in better health outcomes for pregnant women and infants. This study will support the scale-up of Option B+ in Kenya by identifying effective interventions and combinations of interventions that can reduce barriers and increase facilitators of optimal ART adherence and retention in care. We are examining the uptake, feasibility, acceptability of each of the interventions, in order to prepare for further scale-up.

HYPOTHESIS

We anticipate both mobile text messaging and community mentor mothers will increase ART adherence or retention between women and infants receiving Option B+.

GENERAL OBJECTIVE

Our overall goal is to determine which intervention (or combination of interventions) maximizes ART adherence and retention in care in the context of Option B+ and thus improves maternal and infant health outcomes. This goal will be achieved through the following specific objectives:

Aim 1: To evaluate the acceptability of lifelong triple ARV therapy given to HIV-infected pregnant women both for their own health and for PMTCT (Option B+), as well as facilitators, barriers, and potential interventions for ART adherence and retention in care, using qualitative research methods.

The target population for this aim of the study included HIV-infected pregnant women aged 18 years or older, their male partners, and health care workers. Who must be willing to take part in the study.

Aim 2: To compare service utilization outcomes (pregnant women’s adherence to ART, women’s retention in HIV care, and uptake of early infant diagnosis) in four study conditions (community MM intervention only, text message intervention only, both interventions, and control) using 2X2 factorial design.

Aim 3: To examine effects of the individual and combined interventions on maternal and infant health outcomes, including maternal CD4 counts/viral loads, and MTCT at 6 weeks, 12 months, and 18 months. These aims include 1416 (1338 study participants and 80 pilot participants) HIV-infected pregnant women identified in the antenatal care clinics or within the community at the study sites.

A total of 1338 participants were enrolled into the study between June 2015 and August 2017 from all the four study arms for aims 2 and 3.

Control- 334

Community mentor mother arm-334

Community mentor mother / text message arm -334

Text message arm-336.

For this study we are using an exploratory sequential mixed methods research design in which qualitative findings from the first phase of the research will be used to inform the subsequent quantitative phase. For Aim 1, Formative focus group discussions and in-depth interviews with women, male partners, and health workers to evaluate the acceptability of lifelong triple ARV therapy given to HIV-infected pregnant women both for their own health and for PMTCT, as well as facilitators, barriers, and the acceptability of our proposed interventions for ART adherence and retention.  The findings from this initial phase will be used to refine the interventions and data collection instruments to be utilized in the second phase of the study.For Aims 2 and 3, we will assess both service utilization and maternal and child health outcome indicators using a clustered 2×2 factorial design—7 sites will implement both interventions, 7 sites will implement community-based MM intervention only, 7 sites will implement the mobile phone text message intervention only, and 7 sites will implement the existing standard of care only. The interventions will be added to fully integrated high-quality HIV and antenatal, maternal, neonatal, and child health (ANC/MNCH) services already offered at these sites, which will be enhanced with stigma-reduction/sensitivity training for health workers

Methods for Aim 1: To evaluate the acceptability of lifelong triple ARV therapy given to HIV-infected pregnant women both for their own health and for PMTCT, as well as facilitators, barriers, and the acceptability of potential interventions for ART adherence and retention in care.

Recruitment for Aim 1: To collect qualitative data from three groups.  For this portion of the research we selected four communities to represent different types of communities involved in the study.

HIV-positive pregnant or postpartum women were invited to participate in an in-depth interview. We recruited a purposive sample of 5 women at each of the 4 selected sites, for a total of 20 interviews.

Male partners of HIV positive pregnant/postpartum women were asked to participate in an in-depth interview. We recruited 5 male partners at each of the 4 selected sites, for a total of 20 interviews. Half of these were HIV-positive partners (concordant relationship) and half were HIV-negative partners (discordant relationship).

Service providers at study sites were invited to participate in focus-group discussions. The focus-group discussions which included health professionals, lay health workers, community mobilizers, and community leaders in four selected study communities. Four focus groups (6-8 participants each) was conducted with this participant group, for a total of 24-32 represented service provider views. Potential service provider participants will be contacted in person or by phone and asked to participate in a focus group.

Study procedures for Aim 1: In-depth interviews and focus groups were led by experienced qualitative researchers in the local language, using discussion guides developed by the research team. The topics  explored in these interviews and focus groups included:

1. Perceptions of Option B+ compared to other PMTCT regimens,
2.  Barriers to adoption of Option B+, 3) facilitators to acceptance of lifelong ART in setting of Option B+
3. Barriers and facilitators to adherence to lifelong ART
4. Barriers and facilitators to retention in care
5. Ideas for possible ways to help pregnant/postpartum women with adherence and retention in care
6. Acceptability of community Mentor Mothers, and
7. Acceptability and content of text messages for adherence and retention, including timing and duration of intervention.

The topics explored in the interviews with community mentor mothers included:

1. Experience working as a mentor mother and acceptability of mentor mothers,
2. Option B+ compared to PMTCT prophylaxis regimens,
3. Suggestions for improving the community mentor mother program.

Expected outcomes and alternative strategies for Aim 1: The data collected in Aim 1 of the proposed study will complement the qualitative data we have already collected in our preliminary studies and provide us with comprehensive knowledge of the context, challenges, and best strategies for implementing the cMM and text messaging interventions in the context of Option B+ implementation in Nyanza. These findings were used to refine the data collection instruments and intervention plans for Aims 2 & 3.

Methods for Aims 2 and 3: To compare service utilization outcomes (Aim 2) and maternal and child health outcomes (Aim 3) in four study conditions using a clustered 2X2 factorial design. In the 2×2 factorial design (Figure 4), study sites will be randomized into one of four groups:  (community MM intervention only, text messaging intervention only, both interventions, and control).

The Interventions

Our interventions build on high-quality fully integrated HIV and ANC/MNCH services at the clinic level, adding evidence-based approaches in the community and at the individual level to support adherence to and retention in the setting of Option B+. New components to being tested will include: 1) community-based mentor mothers (cMMs) who conduct home visits and other community outreach activities to support PMTCT services and 2) theory-based mobile phone text messages to relay health messages and help retain women, men, and infants in existing clinical services.

Community Mentor Mothers (cMMs):  community mentor mothers who are living with HIV themselves.  The cMMs conduct home visits for HIV-positive women and male partners in their respective communities in order to assist with safe disclosure, support safe infant feeding, promote safer sex and family planning, encourage early infant testing and follow up, and promote ART adherence and return for HIV care visits. There is evidence to suggest that as women have more contacts and more time in contact with the mentor mothers program, especially in the postpartum period, there is greater impact.40

cMM selection criteria: Criteria for becoming a cMM will be modeled on Kenya Mentor Mother Program Guidelines.19 These include: 1) HIV-positive mother, 2) recent PMTCT experience (6 months – 2 years), 3) minimum of Standard 8 education, 3) has disclosed HIV status to at least one person within her household, 4) lives within the local community, and 5) good ART adherence for at least six months.

Linkage of pregnant women with cMMs:  All pregnant women attending routine ANC at a health facility are asked to provide locator information, which is entered into their mother-baby booklet, as well as the facility ANC register. Each health facility has a predefined non-overlapping catchment area. The locations provided by the women will be grouped into zones. Each zone is comprised of approximately 10 villages and will correspond to the geographic sub-locations in the community unit. Each zone will be given a unique identifying number and will have 2-3 resident cMMs who will also be identified by that number. The healthcare provider will give each HIV-positive mother the name and phone contact of the specific cMM for her zone and also obtain informed consent from the mother concerning a home visit by the cMM. Each cMM will be initially assigned around 25-30 women, each of whom she will visit bi-weekly.

Roles and responsibilities: cMMs primarily play two roles.  1. They support pregnant and postpartum women living with HIV who have been identified at the health facility, including

regular home visits with these women and their families. During home visits, cMMs support HIV disclosure, ART adherence, and facilitate retention in care. They also facilitate support groups for families living with HIV, which meet once a month and provide each other with psychosocial support. 2. They also identify additional pregnant women in the community during door-to-door visits and deliver messages to address prevention of both sexual and vertical transmission of HIV. Using illustrative job aids created by the Kenya Ministry of Health, they  explain how HIV is transmitted to the baby and what mothers and fathers can do to prevent transmission.

Monitoring & Linkages: cMMs also be expected to refer additional women from their zones who are not coming for any or all their ANC visits. To assist with their duties, cMMs will each be given a bicycle, identifying shirts, basic supplies

Small gifts to women during postpartum cMM home visits: The cMMs are given a small gift to take to the women during the first postnatal home visit of approximately Ksh 300-500 in value. The culture and norms of western Kenya require that a neonate be presented with a gift in order for a visitor to enter the home. This enables the cMMs to conduct home visits without any difficulty and also remain acceptable within the community. We do not anticipate this will have a significant impact on 12-month ART adherence or retention in care, as the women being given gifts for their babies for the immediate post-natal visit only.

Study Outcomes for Aims 2 & 3:  We hypothesize that more women will adhere to ART and HIV care visits in the intervention arms of the study, resulting in possible increases in maternal CD4 counts, high proportions of viral load suppression, and reductions in vertical transmission.43  ART adherence will be measured via a self-report measure routinely collected in the woman’s medical record, in which a woman’s adherence is classified as good (> 95%), fair (85%-94%), or poor (<85%) assessed from the last clinic visit.  Additionally, viral load measurements will be conducted after 6 months of ART. Those with viral load <1,000 copies/mL will be considered adherent. Those with viral load >1,000 copies/mL will be evaluated for evidence of treatment failure or poor adherence. As has been noted in previous studies, there is no gold standard to measure retention in care.44 Our primary outcome for retention in care in this study will be obtained from women’s electronic medical records, from which we will obtain the proportion of women who have an HIV care visit within 90 days at 12 months after the birth.

Blood sampling and laboratory testing required for primary and secondary study outcomes including HIV testing for pregnant women and exposed infants, maternal CD4 counts and viral loads etc will be conducted as part of routine care per Kenya National ART guidelines and FACES laboratory protocols. Blood will not be stored for later testing and no sampling or laboratory testing will be conducted outside of routine care for research purposes.

Sample size and power for Aims 2 & 3: We have powered this study based on differences in proportions in a cluster randomized trial for our key outcomes of maternal ART adherence (% of women who report at least 85% adherence) and maternal retention in care (proportion of women with visit within 90 days) at 12 months post-partum. Our sample size and power estimates are based on both clinical and logistic factors. Current estimates on required adherence to non-nucleoside based regimens are estimated to be 75-80%.45-47 Additionally, current standard HIV clinic forms in use in Kenya define good adherence as > 95% adherence, fair adherence as 85-94% adherence, and poor adherence as <85% adherence based on number of missed doses. (MOH Comprehensive Care Clinic Patient Card)  Power calculations were made based on estimated differences in these rates in the “intervention” versus the “control” arms of the study, using appropriate methods for clustered data.48 We estimated a baseline ART adherence rate of 55% based on current rates found in the literature for SSA 49 and a baseline retention in care rate of 48% based on data from our recent trial of service integration conducted at FACES sites in Nyanza.³²

We powered our study in order to be able to detect a 25% absolute increase in ART adherence due to each intervention individually.  Given that the actual intra-cluster correlation coefficients (ICC) are not known for these outcomes, and are not easily specified, we calculated the number of communities needed for ICCs ranging from 0.06 – 0.12 for 80% power with a 2-sided alpha of .05, for a fixed average cluster size. (Table 1) To achieve 80% power for detecting a significant difference in the proportion who are at least 85% adherent at 12 months, from 55% in the control arm versus 80% in the intervention arm (2-sided p < 0.05), with a conservative (high) intra-cluster correlation coefficient (ICC) of 0.12 and an average of 70 women sampled per community (estimated minimum of 50 and a maximum of 130 women per cluster), we would need 18 communities, 4.5 in each arm of the trial.  For retention in care, to achieve 80% power to detect a significant difference in the proportion who have an HIV care visit within 90 days of the 12-month postpartum date, from an estimated 48% in the control arm versus 73% in the intervention arm (2-sided p < 0.05), with a conservative (high) intra-cluster correlation coefficients (ICC) of 0.12 and an average of 70 women sampled per community (50-130 women per cluster), we would need 20 communities, 5 in each arm of the trial.  This sample size will power us to evaluate both of our primary outcomes and is conservative due to the high ICC estimate, and thus should also be sufficient if we experience loss-to-follow-up.  This will result in total of 1336 women (334 women per arm).

BACKGROUND ON SUB-STUDY

ASSESSMENT ART ADHERENCE

Objective assessment of medication adherence is notoriously difficult, among the methods used have various setbacks. For example:

• • Verbal recall – affected by social desirability bias
  • Electronic pill monitoring devices (MEMS cap) – can be manipulated by patients
  • HIV viral load – can be affected by resistance to HIV medications, poor absorption of medications, and normal fluctuations in virus

98.8% of MOTIVATE participants report GOOD adherence at 6 month visits

Other ways to measure adherence

• Drug levels
  • blood level of drugs taken at specific time (trough)
  • Patient can take medication day before clinic visit and this is likely to give a false impression on their adherence- “white coat”- compliance
• Hair
  • It is more accurate but not widely acceptable across many cultures.
• Dried blood spot

Measurement of tenofovir in red blood cells (RBC) has recently been introduced. It is an objective and quantitative assessment of long-term drug exposure and adherence

Tenofovir Phosphorylation

Tenofovir is ingested in the gut, it passed through plasma into the RBC where it is phosphorylated to TFV-DP.TFV-DP is trapped inside RBC with a prolonged half-life (17-days)25-fold accumulation to steady state

The expected TFV-DBS level at steady state has been quantified to determine the average number of weekly doses of TDF a patient has taken over the last several

MEASURING TFV-DP USING DBS

• Tenofovir (TDF)
  • The accumulation of TFV-DP reflects adherence in the preceding 1-2 months rather than whether single dose was recently ingested Similar to ARV levels in hair, the additional benefit that blood collection is part of routine clinical care

REASONS FOR THE STUDY

• The measurement of TDF using DBS to assess ART adherence in pregnant/postpartum women living with HIV in sub-Saharan Africa will substantially add to the growing body of evidence supporting the method’s use in resource limited settings.
• The findings from this study will help to develop validated ART levels in pregnant/postpartum women living with HIV.

HYPOTHESIS

• We hypothesize adherence to TDF-based ART will vary among women and correlate to viral load and other psychosocial factors.

MAIN OBJECTIVE

• Our overall goal is to determine the cumulative adherence to TFV-DP-based ART among HIV positive pregnant and postpartum women already enrolled in the MOTIVATE study.

SPECIFIC AIMS

• Aim 1: To assess the cumulative adherence to tenofovir- based ART medication by collecting DBS samples among HIV positive pregnant and postpartum women on ART.
• Aim 2: To determine the association between cumulative adherence assessment and viral load, clinical, demographic and psychosocial factors.
• Aim 3: To assess contributing factors related to HIV positive pregnant and postpartum women’s adherence through qualitative semi-structured interviews conducted at 12 months postpartum.

STUDY SITES

• Only current MOTIVATE study sites are implementing this supplemental study.
• Located in Kisumu, Homabay, or Migori Counties
• Providing PMTCT services including Option B+ integrated into ANC/MCH
• Level II-IV health facility

OVERVIEW OF DESIGN & METHODS FOR AIMS 1&2

• We will collect a small amount of blood via venipuncture from 150 HIV-infected pregnant or postpartum women already enrolled in the MOTIVATE study
• Inclusion criteria:
  • HIV infected pregnant women 18 years or older already consented and participating in the MOTIVATE Study and willing to participate in the additional DBS specimen collection
  • On TDF-based ART for a minimum of 10 weeks prior to sub-study enrollment
    • Nearly 90% of the women in the MOTIVATE study are on a TDF-based ART regimen
  • Enrolled to care at urban or peri-urban site logistically able to collect DBS specimen and transport for freezer storage

Completed less than 12 months of MOTIVATE study intervention at time of sub-study enrollment.

Currently enrollment for this phase of the study is under way in all the study sites.

SAMPLING AIMS 1&2

• Current MOTIVATE study participants who meet eligibility criteria will be contacted by study personnel to discuss interest in participating in this sub study either during routine clinic visits or by phone.
• We will continue to contact eligible women until the proposed sample size of 150 women is reached.
• Those interested will be re-consented. Potential participants will be informed, screened and those eligible will be enrolled into the study.
• Sample size is based on a convenience sample to collect pilot data on the use of TDF-DBS in pregnant women in this setting.

STUDY PROCEDURES FOR AIMS 1&2

We will aim to collect blood from sub-study participants at three different time points:

• during pregnancy (or 10 weeks after initiating ART to achieve TFV steady state)
• at 6 months
• at 9-12 months postpartum
• Personnel will use universal precautions for blood borne pathogens. This procedure can be performed safely outside a biosafety cabinet. 

Preparation of samples:

25μL of blood will be spotted onto designated circles of the DBS Cards. 

Any remaining blood will be disposed as per blood safety protocols.

Two DBS Cards will be prepared per study participant per blood draw.

Cards will be air-dried in a card holder for at least 2 hours

Storage: 

1) Cards will be transported to Lumumba Health Center (a study site) and stored in bags in a cardboard box at -20°C or -80°C. 

• Shipping of samples: 
  • One DBS card per study participant and blood draw will be shipped to the Colorado Antiviral Pharmacology Laboratory
  • DBS cards will be shipped on dry ice to the Pharmacology Laboratory. 
  • Shipping of samples internationally will occur twice during the one year study period.

Processing of samples:

• • DBS analysis will be performed at the UCD Laboratory under the supervision of Dr. Anderson
  • Results of DBS will not be immediately available and thus will not be used in patient care in Kenya.
  • Cumulative and recent adherence to tenofovir will be estimated by TFV-DP concentration in a 3mm punch from a dried blood spot.
    • The punch will be assayed by validated liquid chromatography tandem mass spectrometry

Data collection

• We will not collect new data except:
  • -lab register/sample register
  • Follow up interviews

STUDY OUTCOMES FOR AIMS 1&2

• Primary study outcome: cumulative adherence to TDF-based ART as measured by TDF-DBS testing describe above.
  • We will describe this pilot data regarding adherence among pregnant and postpartum women on TDF-based ART at different time points during pregnancy and postpartum.
  • We hypothesize that ART adherence as measured by TDF- DBS will vary among women on ART and will not always correlate with self-reported adherence.
  • We will compare TDF-DBS adherence measurements to a woman’s self-reported adherence at clinic visits which is classified as good (> 95%), fair (85%-94%), or poor (<85%) assessed at clinic visit nearest to DBS sampling.

We will also compare TDF-DBS adherence with viral load collected nearest to DBS sampling

OVERVIEW OF DESIGN & METHODS FOR AIM 3

• A sub-sample of women (n=45) from those with DBS tests will be contacted for a follow-up semi-structured interview at 12 months postpartum (to prevent contamination of MOTIVATE intervention results).
• Inclusion Criteria:
  • A subsample of 45 women from Aims 1&2
  • Undergone TDF-DBS testing
  • Completed 12 months of intervention at time of interview
• Topics the interviews will explore include:
  • Women’s perceptions and knowledge of ART (and the importance of good adherence)
  • ART adherence experiences during and after pregnancy
  • Facility-level barriers and facilitators to ART adherence
  • Interpersonal-level barriers and facilitators to ART adherence (including support from male partners)
  • Intervention experiences (text or cMM) influencing ART adherence

Interviews will be led by experienced qualitative researchers in the local language, using a semi-structured interview guide

• Forty-five women who are 12 months postpartum and have participated in the DBS collection will be contacted by study personnel either during routine clinic visits or by phone to discuss interest in participating in a qualitative semi-structured interview.
• We will recruit women who have varying levels of ART adherence and reside in different geographic areas/attend various clinic locations.
• We will continue to contact eligible women until the proposed sample size of 45 women is reached
  • Enough for theoretical saturation of ART adherence-relevant behaviors.
• Those interested will be re-consented

STUDY OUTCOMES FOR AIM 3

• The data collected in Aim 3 will complement the qualitative data we have already collected in the MOTIVATE study and provide us with comprehensive knowledge of the context, challenges, and facilitators for adherence to ART during pregnancy and post-partum periods for women living with HIV.
• We will additionally be able to link the interviews to various measures of ART adherence, including the TDF-DBS.

PROPOSED BENEFITS OF THE STUDY

• Potential to scale up this adherence test in the future in Kenya in conjunction with the Kenyan Ministry of Health.
  • Already the University of Colorado, Denver has trained and supported at least one lab in sub-Saharan Africa to conduct this testing.
  • Potential to benefit other Kenyan and sub-Saharan African HIV positive women and their infants in the future.

IMPORTANCE OF KNOWLEDGE TO BE GAINED

The proposed research will contribute to addressing key knowledge gaps in the implementation of Option B+ and to further understand adherence on ART among positive pregnant and postpartum women