Protocol for the PLAY Study: A randomised controlled trial of an intervention to improve infant development by encouraging maternal self-efficacy using behavioural feedback

Introduction The early infant caregiving environment is crucial in the formation of parent-child relationships, neurobehavioural development and thus child outcomes. This protocol describes the Play Love And You (PLAY) Study, a phase 1 trial of an intervention designed to promote infant development through encouraging maternal self-efficacy using behavioural feedback, and supportive interventions. Methods and analysis 210 mother-infant pairs will be recruited at delivery from community clinics in Soweto, South Africa, and individually randomised (1:1) into two groups. The trial will consist of a standard of care arm and an intervention arm. The intervention will start at birth and end at 12 months, and outcome assessments will be made when the infants are 0, 6 and 12 months of age. The intervention will be delivered by community health helpers using an app with resource material, telephone calls, in person visits and behavioural feedback with individualised support. Every 4 months, mothers in the intervention group will receive rapid feedback via the app and in person on their infant’s movement behaviours and on their interaction styles with their infant. At recruitment, and again at 4 months, mothers will be screened for mental health risk and women who score in the high-risk category will receive an individual counselling session from a licensed psychologist, followed by referral and continued support as necessary. The primary outcome is efficacy of the intervention in improving maternal self-efficacy, and the secondary outcomes are infant development at 12 months, and feasibility and acceptability of each component of the intervention. Ethics and dissemination The PLAY Study has received ethical approval from the Human Research Ethics Committee of the University of the Witwatersrand (M220217). Participants will be provided with an information sheet and required to provide written consent prior to being enrolled. Study results will be shared via publication in peer-reviewed journals, conference presentation and media engagement. Trial registration number This trial was registered with the Pan African Clinical Trials Registry (https://pactr.samrc.ac.za) on 10 February 2022 (identifier: PACTR202202747620052).

The Play Love And You (PLAY) Study was developed with the overriding theme that all an infant really needs are opportunities for movement and development (play), responsive and interactive caregiving (love) and the mother herself (you). The PLAY Study was designed following the Standard Protocol Items: Recommendations for Interventional Trials 2013 statement.49 Mother–infant pairs (n=210) will be recruited starting in the third quarter of 2022 until the second quarter of 2023, and individually randomised into two groups (total n after expected attrition=150, n=75 in each group) using simple 1:1 randomisation generated by Stata V.17. The intervention will last from birth to 12 months, with main assessments made when the infants are 0, 6 and 12 months of age. The trial will consist of a standard of care arm and an intervention arm—see figure 1. Data collection teams will be blinded to randomisation at baseline and during primary outcome assessments, and separate data collection teams will collect baseline compared with primary outcome data. Blinding will be maintained by the data manager and broken only for a priori defined analyses (figure 2). Randomisation into either group will be generated by the data manager only after recruitment and baseline assessments are completed, whereafter participants will be assigned to the intervention delivery teams accordingly (intervention delivery teams will thus not be blinded). Intervention delivery teams will work separately and independently of each other, and of data collection teams. The trial is due to be completed in the second quarter of 2024. Participant flow chart. NBO, Newborn Behavioural Assessment. Intervention themes and assessments. We will recruit 210 mother–infant pairs within 3–10 days following delivery at community clinics in Soweto. Soweto is an urban-poor area in the city of Johannesburg covering 200 km2 with over 1.3 million people (6400/km2). Mother–infant pairs will be eligible to participate if: Mothers assigned to the standard of care arm will receive the Road to Health booklet as well as the Side-by-Side services provided by the National Department of Health as routine postpartum care. This content is divided into five knowledge pillars (nutrition, love, protection, healthcare and extra care) and functions to record child’s growth and health interventions, provide information for caregivers and encourage collaboration between healthcare workers and caregivers (https://sidebyside.co.za). In addition to the standard of care, they will be contacted telephonically or in person every 2 months to check in, and to provide referrals to healthcare facilities and community services if necessary; and will have anthropometric, development, movement and interaction measurements taken every 4 months. The standard of care arm will therefore include: Mothers assigned to the intervention arm will receive everything as per the standard of care arm. Additionally, they will receive various microinterventions according to the following key intervention themes: (1) exclusive breast feeding for 6 months, (2) interactive play and promotion of early learning, (3) responsive caregiving and (4) maternal mental health. Each intervention theme will be prioritised during different phases of the trial and supported by components from the other themes. Figure 2 details when each theme will be prioritised and which supporting interventions will be included. Interventions will be delivered telephonically or in person, and via a mobile app created for the purposes of this trial. Interventions are detailed below. At recruitment, mothers will have the Newborn Behavioural Assessment (NBO)50 administered by trained research staff. The NBO consists of 18 neurobehavioural observations along four dimensions—autonomic, motor, state organisation and attentional–interactional—ultimately yielding a profile of the infant’s behavioural repertoire. Importantly, the NBO allows for provision of infant specific anticipatory guidance, thus engaging parents in discussions around nurturing care, mother–infant interaction, play and development, and breast feeding. The NBO has been shown to improve maternal sensitivity, and to promote positive maternal–infant interactions.50 Every 4 months, mothers will receive rapid feedback via the PLAY Study app and in person on their infant’s and their own movement behaviours and on their interaction styles with their infant. Feedback will be based on objective measurement of infant movement behaviours and headcam video footage of interactive play in the home environment. Feedback on movement behaviours will consist of a graphical representation of the infant’s and the mother’s movement behaviours over the previous week, with personalised information and advice about: sleep routines, play and tummy time, nap times, daily activity routines, peaks of high activity versus lulls in activity, and recommended daily movement guidelines. This feedback is designed to sensitise mothers to their infant’s behaviour patterns and to encourage healthy infant movement and routines, thus improving their self-efficacy to promote healthy routines for development. An example of the graphical feedback presented is included in online supplemental figure 1. Feedback on mother–infant interactions during play is guided by the Video-feedback intervention to promote positive parenting and sensitive discipline (VIPP-SD)51 and will consist of a video segment from a peak interaction moment with mother and infant perspectives presented side-by-side, with personalised information and guidance on: positive interactions at 4 months (baby smiling, eye contact, soothing, directing attention, reinforcing mothers strengths, communication and bonding), perspective taking at 8 months (describing and narrating the interaction, seeing baby as an individual, understanding baby cues) and recognising child feedback at 12 months (sensitivity and responsiveness, following cues, adjusting behaviour, soothing styles, interaction chains). This feedback is designed to promote positive and nurturing interactions and improve maternal self-efficacy. The app will contain the graphical and video feedback and will also allow mothers to observe changes in their infant’s behaviour and development over the trial period. bmjopen-2022-064976supp002.pdf At recruitment, and at 4 months mothers will complete the Edinburgh Postnatal Depression Screening (EPDS) Questionnair, the Post-Traumatic Stress Syndrome Checklist 5 (PCL-5), the Adverse Childhood Experiences (ACE) Scale and the Antenatal Stress Questionnaire (ASQ). Women who score in the high-risk category on one or more of these questionnaires will receive a phone call to assess any immediate risk, and referral to an individual counselling session from a licensed psychologist at their local clinic, followed by continued support as necessary. Women who score at medium risk on one or more of these questionnaires will receive referrals to the licensed psychologist and continued support as necessary, and in both cases uptake of this support will be monitored. All participants will receive a flyer on mental health and available services and support upon completion of the questionnaires. Every week, health theme-specific literacy content will be delivered via the PLAY Study mobile app, which was designed for the purposes of this intervention. The app creates a password-protected profile for each participant and houses a section to deliver the feedback when available, as well as the weekly (and historic) content. This content is loaded to the profile every Monday, and participants can choose to receive push notifications to alert them of new content. Participants can engage with the app at any time, for as long as they wish and as many times as they wish during the intervention. The weekly content will appear in the format of quick and easy tips, short videos, infographics or ideas for home-based activities. Mothers will have continuous access to the app with all relevant resources, activities and tools to support their interactions with their infant. Every 2 months, theme-focused content will be delivered telephonically or in person. Telephonic or in person visits will also follow the key intervention themes schedule, but will allow mothers to guide the conversation by using open questions and providing anticipatory guidance. Referrals to lactation specialists or psychologists or other services will be provided as necessary. Health literacy content themes will be scheduled as follows and according to figure 2. All content will be guided by a broad range of pre-existing content such as UNICEF resources (https://www.unicef.org) and WHO recommendations6 and other evidence based online resources. Contextually relevant material or adaptations to existing content will also be developed with guidance from a community advisory group (CAG). The intervention arm will thus include the standard of care activities and: The intervention has been informed by the UK Medical Research Council (MRC) Guidelines for Complex Interventions and is grounded in behaviour change techniques developed by Michie et al.42 We used the Template for Intervention Description and Replication Checklist. Figure 3 details the logic model to illustrate the pathways to impact that guided the intervention approach. The agents of change are community health helpers (HHs), who do not need to have any specific qualification or expertise in health promotion but will be trained by the study team in community health promotion, as well as relevant intervention methodology. The HHs are supported by a supervisor with access to a referral network. Play Love And You Study logic model.SES – socioeconomic status The intervention will be delivered by a combination of phone calls, in person visits, app-based health literacy content, personalised behavioural feedback and guidance, as well as individual counselling sessions for high-risk mothers. The content delivery will be guided by a schedule following key themes according to infant age and developmental stage. Phone calls and in person visits will be guided by a conversation flow chart, which will start with HHs asking mothers how they are doing and gently introducing the theme for discussion using open-ended questions. The flow chart will be designed to allow HHs to either ask follow-up questions, or to provide anticipatory guidance and direct them to relevant resources in the app based on mothers’ individual responses (example provided in online supplemental figure 2). This will allow mothers to guide the conversation, while still following key themes for discussion. HHs will refer mothers as necessary if any concerns are raised during these conversations. bmjopen-2022-064976supp003.pdf Data collection will be conducted by independent, trained research assistants. The primary outcome measure is efficacy of the intervention in improving maternal self-efficacy at 12 months. Secondary outcomes will estimate efficacy of the intervention in improving infant development at 12 months, and will assess the feasibility and acceptability of the intervention at 6 and 12 months, as well as interim outcomes to determine mechanisms of impact (figure 2). Maternal self-efficacy will be assessed using the Perceived Maternal Parenting Self-Efficacy Tool at 6 and 12 months. This questionnaire assesses four subscales of self-efficacy: caretaking procedures, evoking behaviour, reading or signalling behaviours and situational beliefs. While this tool was originally designed for assessing mothers’ perceptions of their ability to understand and care for their hospitalised preterm neonates as well as being sensitive to the various levels and tasks in parenting,52 it has since been successfully used in older, healthy infants in low-income settings.53 It has also been piloted in a sample of mothers from Soweto, providing a Cronbach’s Alpha of 0.95 (data not yet published). Infant development will be assessed at 12 months (and at 6 and 8 months) using the ASQ3. These questionnaires assess development in five domains: communication, gross motor, fine motor, problem-solving and social–personal. The ASQ3 has been successfully applied in community-based, low-resource settings and has been found to be feasible in South Africa.54 55 The acceptability of each phase of the intervention will be assessed based on the Theoretical Framework of Acceptability.56 This framework includes the following criteria: affective attitude, burden, ethicality, intervention coherence, opportunity costs, perceived effectiveness, and self-efficacy. These criteria will be assessed by means of questionnaires specific to each intervention theme, followed by focus group discussions (FGDs) in a subsample of participants to expand on core themes identified. Feasibility of each phase of the intervention will be assessed by determining whether the developed intervention is implementable in this context, has potential to work, as well as how it could be improved for future use and in other settings. We will follow the MRC guidance on process evaluation of complex interventions.57 The evaluation will focus on implementation, mechanisms of impact and context. Maternal breastfeeding self-efficacy will be assessed every 2 months from 0 to 6 months using the Breastfeeding Self-Efficacy Short Form.58 This comprises 14 questions, each with a 5-point Likert response scale ranging from 1=not at all confident to 5=very confident. The higher the final aggregate score (which has a range of 14–70), the higher the level of breastfeeding self-efficacy. Exclusive breast feeding will be assessed every 2 months from 0 to 6 months by asking mothers whether they are currently breast feeding and if not, at what infant age did they stop breast feeding and why they stopped breast feeding. They will also be asked if they are giving their infant anything other than breastmilk, and if so, why. Weight and length will be measured by trained research staff at baseline and 12 months (and at 4 and 8 months). Weight-to-length ratio (kg/m) will then be calculated to estimate anthropometric changes in body composition. At 12 months, DXA scans will be performed by a trained technician. Whole-body measurements of fat mass and fat-free mass will be extracted for use in analyses. FMI (kg/m3) will be calculated from these estimates to describe adiposity. Health literacy will be assessed at 6 and 12 months using the parent version of the Health Literacy Questionnaire,59 which covers nine areas of health literacy including functional health literacy and has been validated in high-income and low-to-middle-income countries. Various aspects of maternal mental health will be assessed at baseline, 4 and 12 months. The PCL-560 will be administered whereby rating scale descriptors include: ‘not at all’, ‘a little bit’, ‘moderately’, ‘quite a bit’ and ‘extremely’. A total symptom severity score is calculated by adding up the item scores, which range from 0 to 80. The cut-off for PTSD that will be used in this study was 31. The EPDS61 will be used to detect postnatal depression. Scores for each item range from 0 to 3 on a Likert scale and are added together to obtain an overall total. Scores of 7–13 indicate mild depression, 14–18 moderate and 19–30 severe depression. The ACE Questionnaire62 will assess childhood abuse, neglect and household challenges. A score from 0 to 10 is calculated based on how many types of events were experienced before 18 years old. The ASQ will be administered to assess stressful life events. An additional nine items about social support will be included. The Postpartum Bonding Questionnaire63 will be used to detect early mother–infant bonding. It consists of four subscales ‘impaired mother/infant bonding’, ‘rejection and anger’, ‘anxiety about care’ and the ‘risk of abuse’. These questions are rated on a 6-point Likert scale, ranging from 0 (always) to 5 (never). Infant adherence to movement guidelines will be assessed every 4 months using accelerometery. Seven-day, 24-hour wrist worn accelerometery data (Axivity-AX3 worn on both mother and infant) will be collected, downloaded (omgui, Open Movement, UK), auto-calibrated and processed. Data is summarised to generate average acceleration (mg) per 5 s interval, as well as overall time (hourly and weekly) spent in varying intensities of activity, which will allow for determination of adherence to movement guidelines and changes in activity levels over the intervention period. Mother–infant interactions will be measured during play using first-person observation and behavioural analysis every 4 months. Mother–infant interactions have primarily been measured using questionnaires or third-party observation, yet these measures are limited by their subjectivity and intrusiveness. First-person observation using headcams—small cameras which are attached to headbands on both infants and mothers for a period of time while interacting at home—provide a first-person view of the interaction from both infant and mother perspectives with more naturalistic and detailed measures. These headcams have been piloted in South Africa and were found to be largely acceptable and feasible in this context.64 Headcam measures will be collected during at least five, 3–5 min play interactions over the course of a week. Headcam observations will be coded using Noldus Observer XT software. An event-based coding framework will be used to categorise second-by-second key maternal and infant behaviours, that is, touch, play, facial expressions, hand movement, and gaze. This allows for determination of the probability of certain behaviours preceding or following another, associations between maternal/infant measures and these behaviours, and synchrony between mothers and infants as a potential indicator of sensitivity and responsiveness. Prior to trial implementation, a CAG (n=16) of mothers from Soweto with infants aged 0–3 months will be established to serve as experts on the lived experience of being a mother to an infant in Soweto. For about half of these women, family matriarchs will be recruited to form a separate CAG (n=8). Research from South Africa has indicated that family matriarchs play an important role in the choices that mothers make around infant feeding and nutrition,13 and as such it is pertinent they be included in interventions targeting breast feeding and other child care practices. Intervention content and activities, and behavioural feedback will be tested and/or developed with the CAG, using FGDs and piloting of intervention content and feedback. Thus, the initial acceptability and feasibility of the intervention content will be tested by the CAG in order to understand any changes required, and thus refine the intervention accordingly. The CAG will continue to advise throughout the duration of the intervention, via regular (every 4 months) FGDs where the research team will raise queries about intervention content, delivery and uptake, retention strategies, and feedback mechanisms. Additionally, the CAG will provide important insight into their lived experience of being a mother in Soweto over time and as their infant grows. This will provide useful insight to guide intervention conversations and to better understand the context within which the intervention is being delivered. Monitoring of intervention activities will be done throughout the trial. The delivery of each contact (duration, date, content) will be recorded using a monitoring log in order to be able to monitor intervention intensity and frequency. Engagement with the app content and behavioural feedback (ie, number of page visits per day and duration of time spent on the app) will be monitored via the app. Adherence will be assessed every 4 months, and strategies to improve adherence will be discussed with the CAG as necessary. Additionally, the research team will meet with HHs weekly to determine any concerns or issues with compliance or retention, and strategies will be discussed accordingly. Any adverse events will be recorded and reported as necessary according to standard operating procedures. Given expected attrition of 30%, recruiting 210 mother–infant pairs should yield a final sample size of 150 mother–infant pairs, which will be divided into 2 groups (n=75 each). A sample size of 150 mother–infant pairs divided into 2 groups (n=75 each; control and trial group) will allow for 99% power to detect efficacy of the intervention based on maternal self-efficacy data assessed in the piloting phase of the behavioural feedback intervention (data not yet published). Attrition will be minimised by maintaining regular contact with both groups, by recruiting participants who plan on remaining in Soweto, and by using a flexible and versatile delivery approach including in-person visits, app-based content, telephonic sessions, mobile messaging and site visits. The efficacy of the intervention will be assessed using intention to treat analysis. Secondary outcome data will be compared between the intervention and standard of care arms using Student’s unpaired t-tests. Focus will be placed on the primary efficacy outcome (maternal self-efficacy) due to limited power for secondary outcomes, which will therefore be treated with caution and presented using point estimates and precision (95% CI). Sensitivity analyses will include comparing baseline characteristics between intervention and standard of care arms, examining the effect of intervention dose on outcome data and stratifying analyses by sex. The feasibility and acceptability of the intervention at each phase will be assessed using qualitative and quantitative methods. FGDs will be transcribed and coded line by line, and data will be grouped according to common themes. Themes will be sorted into subthemes and continually reviewed to identify an overarching narrative to identify key findings from each FGD. REDCap will be used for data management (https://www.project-redcap.org). Secure Sockets Layer (SSL) provides the needed security for the REDCap similar to what a Virtual Private Network (VPN) will provide. The Wits University server hosting REDCap is resilient and hardened against hacking and other forms of attacks. All data traffic is encrypted, and an installed network firewall provides secure network access. The REDCap system provides inherent application-level security including access controls and comprehensive audit trails, and all incoming requests get intentionally filtered, sanitised and escaped to prevent against attacks. Data files and metadata will be uploaded and stored on REDCap by the project data manager, on a password-protected Wits server for the duration of the project to enable team access. REDCap is used for data and metadata storage, as well as for data collection. Data will be inputted online with secure web authentication, data logging and SSL encryption. Accelerometery files and video files, as well as FGD notes and transcripts, will be stored on a secure Wits server, as well as on an external hard drive, and backed up as metadata to REDCap. Regular backup copies will be made by the project data manager on a password-protected file space. When the project is completed, the full database will be stored on both Developmental Pathways for Health Research Unit (DPHRU) and University servers. The items and metadata on the institutional repository are backed up on a daily basis and these backups are kept on tape. All data collection devices will be encrypted. Data from this study will be stored for a minimum of 10 years from the completion of the study in order to allow for any verification of data. Long-term data storage will be in.csv files using a standard (simple) American Standard Code for Information Interchange (ASCII) coding scheme in line with UK Data Service recommendations (https://www.ukdataservice.ac.uk/manage-data/format/recommended-formats). Ethical approval has been granted for the establishment of the CAG (M210846), and for the trial (M220217) from the Human Ethics Research Committee of the University of the Witwatersrand, Johannesburg, South Africa. Additional approval has been granted by the Research Committee of Johannesburg Health District (GP_202202_021). Participants will be provided with an information sheet and required to provide written consent (online supplemental document 1) prior to being enrolled. Study results will be shared via publication in peer-reviewed journals, conference presentation and media engagement. bmjopen-2022-064976supp001.pdf

• prioreschi, alessandra
• pearson, rebecca m.
• richter, linda
• bennin, fiona
• theunissen, helene
• cantrell, sarah j.
• maduna, dumsile
• lawlor, debbie a.
• norris, shane a.