Objectives
The trial will evaluate the potential benefits of US on early childhood development and child care by augmenting a routine antenatal service currently provided mainly in tertiary public health facilities in South Africa, with plans to be rolled out to district hospitals. The goal of the intervention is to increase parental investment not only in child survival, but also for the wellbeing and healthy growth and development of their child. We aim to establish whether, compared to routine care, messages about fetal and child development delivered during one (<25 weeks) or two (< 25 weeks and again <36 weeks) fetal ultrasounds improve ECD and growth, exclusive breastfeeding, mother/father-infant attachment, maternal and paternal well-being, father involvement and pre- and post-natal clinic attendance.
Specific objectives are to evaluate whether child development messages delivered during one or two ultrasounds:
- Improve early child development
- Improve infant breastfeeding practices and healthy child growth
- Improve parent-infant attachment and interaction during pregnancy and six months after birth
- Increase paternal engagement in a young child’s development and care
Design
Eligible mothers will be randomised to one of three arms. Arm 1, the control group, will receive standard practice of care during pregnancy; arm 2 will receive the child development intervention at the first research ultrasound visit, and arm 3 will receive the intervention at two research ultrasound visits. All mothers, infants and partners (where available) will be followed-up 6 weeks and 6 months postnatally for assessments. Participants will continue to receive routine ante-, peri- and postnatal care as provided by the tertiary care facility from which all participants are being recruited from.
Setting
The study is situated in Soweto, South Africa. Soweto is adjacent to Johannesburg and is the most populous Black urban residential area in South Africa with over 1.3 million people counted in the most recent census (43). Participants will be recruited from women attending antenatal care at Chris Hani-Baragwanath Academic Hospital (CHBH), a tertiary teaching hospital that attends to over 22,000 births annually. Women are referred to CHBH from community health centres with risk factors including HIV and hypertension. Recruited women will attend their study ultrasound visits at the SAMRC Development Pathways for Health Research Unit (DPHRU), which is located on the grounds of CHBH and in walking distance of the CHBH antenatal services.
Procedures
Recruitment
Approximately 500 women present to CHBH each month for their first pregnancy ultrasound, of which, 50 percent are <25 weeks gestation. This offers a potential pool of 250 women a month to be included in the study. Participants will be recruited from women attending the Foetal Medicine Unit (FMU) at CHBH, where routine fetal ultrasounds are conducted. Eligible women will be resident within Soweto, 18 years and older, with a singleton pregnancy and presenting <25 weeks gestation. Major fetal abnormalities detected during the first routine ultrasound conducted by an experienced radiologist in the FMU, or severe maternal morbidities, will exclude a pregnant woman from the trial. These women will be transferred to specialised clinical services at CHBH where she will receive appropriate treatment.
Sample size calculations were based on detecting a 0.5 standard deviation (SD) difference between arms on both the Bayley III Language and Cognitive Scales, and z-scores for weight-for-age (WAZ), length-for-age (LAZ) and weight-for-length (WLZ) measured at 6 months of age at 90% power, which is 85 per arm. Allowing for 15% potential attrition during pregnancy and during follow-up to six months of infant age, sample size in each of the three arms will be 100. This makes a total sample size of 300 women for recruitment.
Intervention
Women will be randomised to one of three arms on recruitment at CHBH <25 weeks gestation after having a fetal ultrasound as part of routine antenatal hospital services. The women will be blind to their arm allocation and remain blinded until study completion. Using computer-generated simple randomisation, each consecutive woman screened will be randomly allocated to a study arm. Research assistants responsible for screening will receive the study arm numbers for allocation from the study coordinator on a daily basis. The allocation sequence will be concealed from the research assistants enrolling the participants. Women recruited for arm 1 will act as the control group and will receive one further standard US at DPHRU <25 weeks gestation when they will receive standard fetal growth measurements, but no child development messages.
In addition to their routine antenatal US at CHBH, women recruited to arm 2 will receive an additional US <25 weeks gestation at DPHRU, during which they will receive messages to promote early childhood development. During the US session, the sonographer will talk with the mother and partner, if present, about the fetus’ sensory development and capacity for learning, fetal responsivity to the environment, maternal wellbeing and the dependence of the fetus on maternal and paternal health and wellbeing before and after birth.
In addition to their routine antenatal US at CHBH, women recruited to arm 3 will receive two additional USs at DPHRU, one <25 weeks gestation (early US) and another < 36 weeks gestation (late US). They will receive the same child development messages as mothers in arm 2 at the first US. During the second US, in addition to childhood development messages, the sonographer will encourage the woman and her partner, if present, to personalise the baby, strengthen intentions to exclusively breastfeed, talk and sing to the child before and after birth, and counsel them on how to respond to their young infants emotional and learning needs.
Partner involvement and interest in the baby is encouraged by inviting partners to attend the ultrasound sessions of women in all three arms. For mothers recruited to arms 2 and 3, digital ultrasound images and a photograph of the woman (with her partner where present) having the US will be sent to parents via WhatsApp and printed images given to the parents. A small baby book with the child development messages, as appropriate to the arm, is given to the mothers in arms 2 and 3, and they are encouraged to paste the printed pictures into the book and to share the US images and photographs with family and friends.
Two qualified and registered sonographers will be trained to deliver the intervention and a Philips HD9 Ultrasound System will be used to conduct all ultrasounds. Due to the nature of the intervention, the sonographers will not be blind to the mother arm allocation.
Data collection
Study data will be collected and managed using REDCap (Research Electronic Data Capture) hosted by the University of the Witwatersrand. Data will be collected on the enrolled women, child and partners (where appropriate) at three stages; 1) baseline at first research ultrasound at DPHRU <25 weeks; 2) 6-week post-natal follow-up, and 3) 6-month post-natal follow-up. Data will be collected on tablets at stages one and two by two trained data collectors under the supervision of the study coordinator. At stage three, data will be collected by occupational therapists or physiotherapists trained in administering child development measures, under the supervision of the study coordinator. All data collectors will be trained on the principles of research ethics, good clinical practice, referral procedures and management of difficult situations in the field.
Table 1 shows the schedule of enrolment, interventions and data collection.
[Table 1 placed here]
The data collected at each stage is as follows:
Baseline at first research ultrasound at DPHRU. Data at the first research visit (<25weeks) will include participant contact details, demographic characteristics and reproductive history, socio-economic status, social and instrumental support, and breastfeeding intentions. Maternal and paternal fetal attachment is assessed using the Maternal and Paternal Attachment Scales (44, 45), and the mother and her partner, if available, will complete a social support questionnaire and the Edinburgh Postnatal Depression Scale (46). The woman and her partner, if available, will also be asked to complete an ultrasound experience questionnaire after the US.
Six week postnatal follow-up. Two trained outcome assessors, blind to the mother’s arm allocation, will complete the follow-up questionnaires with the mothers, infants and partners, where available. Questionnaires will assess perinatal and postnatal information, breastfeeding practices and partner and social support. Postnatal depression will be assessed for the mother and partner using the Edinburgh Postnatal Depression Scale. Infant growth will be assessed using standard procedures for infant weight and length (the SECA 416 infantometer for length and the SECA 367 infant scale for weight). Measures will be converted to z-scores using the 2006 WHO Growth Standards for LAZ, WAZ and WLZ. Birth weight and length as well as key information on maternal and infant birth outcomes will be collected from the infant’s Road to Health Booklet.
Six month postnatal follow-up. Occupational therapists or physiotherapists trained in child development will complete the six-month follow-up assessments with the mothers, infants and partners, where available. The primary outcome for the six-month follow-up is child development as measured by the Bayley Scales of Infant and Toddler Development, Third edition (47).
All the study outcome measures are shown in Table 2.
[Table 2 placed here]
Formative and process data
Study procedures were informed by formative qualitative research conducted with pregnant women attending antenatal care at CHBH prior to the trial inception. We have written standard operating procedures (SOPs) to support the research team to deliver the study procedures and intervention according to the protocol. Participant adherence will be assessed by counts (for example, visit attendance) and intervention delivery will be monitored by observing contacts between the sonographer (who will conduct the ultrasound and deliver the ECD messages) and participants (intervention and control) to record the nature of interactions. We will monitor intervention dose and experience by recording the duration of the ultrasound visit (with and without delivery of the intervention) and by gathering participant and partner views on the ultrasound experience. Ongoing consultation with a key stakeholder network will inform intervention procedures and assist with contextualising the intervention and addressing any challenges that may arise during trial implementation.
Due to the well-documented challenges with engaging men in research and health care in South Africa, (54) we will adopt pro-active recruitment strategies, which will include phone calls to partners after participant enrolment and follow-up calls closer to the visit to encourage attendance. We have also included Saturday sessions for those partners who are unable to attend during the week due to work commitments. Research staff will make regular contact with participants in-between study visits to maintain accuracy of records and to promote retention of participants in the study. We will follow up participants who default and those who withdraw to gain an understanding of why they dropped out during the course of the trial.
Data analysis
All available data at baseline will be analysed descriptively for each intervention group and in total. Categorical data will be summarised by numbers and percentages. Continuous data will be summarised by mean, SD and range if data are normal and median, IQR and range if data are skewed. Minimum and maximum values will also be presented for continuous data. Differences regarding baseline variables will be checked for clinical relevance between the three treatment groups.
A CONSORT flow diagram will be used to summarise the number of participants who were: assessed for eligibility at screening, were eligible at screening, eligible and randomised, lost to or discontinued the intervention, included in the primary outcome analysis. If possible respondents and non-respondents will be compared.
The analysis of the primary outcome will be performed on the intention to treat population with no imputation of missing data. In the primary analysis, child development scores will be compared to the control group. If outcome data are normally distributed, we will fit linear regression models. If not normally distributed, outcome data will be categorised and logistic regression methods will be used to assess differences between intervention and control groups. Tests will be two-sided with a significance level of 0.05 and 95% Confidence Intervals will also be reported. In case of relevant differences in baseline characteristics additional adjustment(s) for these factors will be performed within the models to address potential confounding.
A stratified analysis will be done for three pair-wise comparisons, i.e. Arm 1 vs. control, Arm 2 vs. control, Arm 1 vs. Arm 2 (as described in 6.2) according to the following subgroups: Partner vs no partner accompaniment and parity.
STATA, Version 15 (StataCorp, College Station, Texas, USA) will be used for all analyses. Missing data and data from those who withdraw from the study will not be interpreted.