Randomised controlled pilot feasibility trial of an early intervention programme for young infants with neurodevelopmental impairment in Uganda: A study protocol

Introduction Early intervention programmes (EIPs) for infants with neurodevelopmental impairment have been poorly studied especially in low-income settings. We aim to evaluate the feasibility and acceptability of a group participatory EIP, the â € ABAaNA EIP’, for young children with neurodevelopmental impairment in Uganda. Methods and analysis We will conduct a pilot feasibility, single-blinded, randomised controlled trial comparing the EIP with standard care across two study sites (one urban, one rural) in central Uganda. Eligible infants (n=126, age 6-11 completed months) with neurodevelopmental impairment (defined as a developmental quotient <70 on Griffiths Scales of Mental Development, and, or Hammersmith Infant Neurological Examination score <60) will be recruited and randomised to the intervention or standard care arm. Intervention arm families will receive the 10-modular, peer-facilitated, participatory, community-based programme over 6 months. Recruited families will be followed up at 6 and 12 months after recruitment, and assessors will be blinded to the trial allocation. The primary hypothesis is that the ABAaNA EIP is feasible and acceptable when compared with standard care. Primary outcomes of interest are feasibility (number recruited and randomised at baseline) and acceptability (protocol violation of arm allocation and number of sessions attended) and family and child quality of life. Guided by the study aim, the qualitative data analysis will use a data-led thematic framework approach. The findings will inform scalability and sustainability of the programme. Ethics and dissemination The trial protocol has been approved by the relevant Ugandan and UK ethics committees. Recruited families will give written informed consent and we will follow international codes for ethics and good clinical practice. Dissemination will be through peer-reviewed publications, conference presentations and public engagement. Trial registration number ISRCTN44380971; protocol version 3.0, 19th February 2018. We will conduct a pilot feasibility, single-blind, RCT with two parallel groups: one receiving the EIP and the other SC. The study is based at two Ugandan sites: one urban (Mulago Hospital, Kampala) and one rural (Kiwoko Hospital, Nakaseke). Neither site has existing family support services for children with NDI. Mulago National Referral Hospital is the largest in Kampala, Uganda’s capital city, taking high-risk pregnancies from across surrounding areas. Children’s services include acute admissions, an inpatient malnutrition unit and outpatients, with a weekly paediatric neurology clinic providing investigation and management of neurological conditions including seizures, and a clinic-based physiotherapy and occupational therapy service for children with cerebral palsy and other NDIs. Kiwoko Hospital in Nakaseke District, central Uganda, serves a catchment area of 800 000 people and provides comprehensive medical services, including neonatal inpatient care for >1200 infants per year. The trial implementation partner, Adara Development, has worked in partnership with Kiwoko Hospital since 1998, and the government to improve neonatal health in Nakaseke district. Together they provide HIV services, maternal health services and community-based healthcare to 44 villages surrounding Kiwoko Hospital. Participants will be young children with NDI and their caregivers. A Standard Protocol Items: Recommendations for Interventional Trials diagram showing the planned flow of participants is presented in figure 1. Flow of participants. *In-depth interviews (IDI) with caregivers on impact of disability, confidence level of the parents, level of participation in family and community life and experience of stigma/discrimination. EIP, early intervention programme; GMDS, Griffiths Mental Developmental Scales; HINE, Hammersmith Infant Neurological Examination; HOME, Home Observation for the Measurement of the Environment; MDAT, Malawi Developmental Assessment Tool; MIRI, Maternal Infant Responsiveness Inventory; NDI, neurodevelopmental impairment; PedQL, Pediatric Quality of Life tool; PEDI, Pediatric Evaluation Disability Inventory; PSI, Parent Stress Index; SRQ, Self-Referral Questionnaire. Infants at high-risk of NDI will be identified from (1) neonatal admission registers and neonatal follow-up services, (2) local paediatric outpatient services and (3) attendance for early child health services following community sensitisation. Sensitisation will include public health announcements on local radio raising awareness of the research and appropriate child development more generally. Caregivers of high-risk infants (survivors of neonatal encephalopathy, prematurity, neonatal septicaemias/meningitis and severe jaundice) will be contacted by phone and invited to attend an appointment when the child is 6–11 completed months old. After informed written consent, they will be screened for NDI by trained study staff using the Malawi Developmental Assessment Tool (MDAT).12 If two or more items in any MDAT domain are not achieved, the child will be referred for comprehensive neurodevelopmental assessment. If the child fails one item in two or more domains, they will be invited back for an assessment in 1 month. If the child’s MDAT scores are age appropriate across all domains, advice will be given on play and stimulation, communication, nutrition and immunisations and the child discharged. Caregivers of infants screening positive on MDAT will be invited to an appointment for written informed consent, and if provided, comprehensive neurodevelopmental assessment by study staff using the Griffiths Mental Developmental Scales (GMDS)13 and the Hammersmith Infant Neurological Examination (HINE).14 An overall developmental quotient will be derived, from the GMDS subscales assessing locomotor, personal–social, hearing/language, eye-hand coordination and performance skills.13 Neuromotor impairment will be further assessed according to the HINE, a standardised paediatric neurological examination and classified by type. We have used both these tests extensively in previous studies in Uganda and found them easy to administer in this setting and at this age.15 The assessments will be conducted in the local language using the standard manual material to ensure internal consistency in the assessments technique. Inclusion and exclusion criteria are outlined in box 1. Infant and caregiver demographic information will be recorded at baseline, including date of birth, age, sex, birth order, parity, antepartum, intrapartum and postpartum history, family and medical history, developmental history, mother’s education and occupation, family details including family size, and ages, household incomes, household SES and residence. All outcome measures will also be measured at baseline enabling preintervention and postintervention comparisons. Infants and their caregivers will be randomised in a 1:1 ratio to either the EIP or SC arm. Randomisation will be stratified by recruitment centre. Randomisation lists indicating a randomisation number and trial arm allocation will be prepared by the trial statistician using a random number generator in Stata (V.15) prior to the commencement of the study, and stored on a secure, password-protected computer at the Medical Research Council/Uganda Virus Research Institute (MRC/UVRI) and London School of Hygiene & Tropical Medicine (LSHTM) Uganda Research Unit by a statistician otherwise not involved in the study. When a participant is eligible for recruitment and consent obtained, study staff will contact the MRC/UVRI statistician who will inform the study staff of the study number and trial arm to which the participant is to be allocated. The personnel in charge of the randomisation will not be involved in other study procedures, including assessment of outcomes. The EIP is a community-based, peer-led group programme with caregivers at a community level, using a participatory approach driven by adult learning theory.16 The programme manual is freely available to download (https://www.ubuntu-hub.org). Development of the programme is described in box 2. Core themes and content of the ABAaNA early intervention programme. Participating families are encouraged to share experiences through discussion and reflection, prioritise problems and identify solutions together. Facilitators of the group sessions are ‘expert parents’, themselves parents of children with NDI, who have undergone 5 days of core training followed by regular supervision, face-to-face mentoring meetings and telephone discussions with existing in-country master facilitators (trained therapists in Uganda). Each EIP group involves 6–10 families; groups are selected depending on locality for ease of attendance. The training is divided into 10 modules covering understanding disability, positioning and carrying, feeding, mobilising, communication, play, everyday activities and experiences in the local community (figure 2, table 1). Individual module sessions are delivered every 1–2 weeks and last 2–3 hours including time for facilitated discussion; the entire programme is designed to be delivered over 6 months including at least one home visit conducted by the expert parent facilitator. Description of the programme modules EIP facilitators will receive a 5-day training programme delivered by two master facilitators, which includes facilitation skills, knowledge transfer on the core contents of the EIP manual and translation of knowledge to practice through simulated sessions with families and children with NDI. All trial intervention groups will be co-facilitated by a master facilitator providing supportive supervision to new facilitators. After each modular group meeting, a short-facilitated feedback session will be conducted, and the content of the module delivered will be recorded. Attendance of individual caregivers and children at the group sessions will be recorded. Facilitators will emphasise to caregivers the importance of attending all sessions, with phone calls prior to each session to promote adherence. If missed, a catch-up session may be offered before the next module. SC refers to care that is currently available in established local services. In both sites, this includes referral to physiotherapy, seizure management and nutritional support. Information on access to local medical, therapy and family services will be collected. Families in the SC arm will be offered delayed entry into the EIP after completing the 18-month assessment. Contamination of the SC arm by exposure of SC families to intervention will be monitored and reported. Participants in both arms will be assessed by study staff masked to trial allocation at two time points; at age 12–17 months (which corresponds to completion of the EIP in the intervention arm, 6 months after recruitment) and again at age 18–23 months (12 months after recruitment). (figure 1). Caregivers will be phoned a week before the follow-up assessments to arrange a time for interview. Assessments will be primarily conducted in the study-site clinics. Where caregivers cannot be contacted by phone or are unable to attend the clinic, a community visit will be arranged, and assessments completed at home. Outcome assessments will be conducted by Mulago assessors for children recruited at Kiwoko, and vice versa to ensure assessors are blind to allocation arm. Two assessors will independently assess a small proportion of the children and inter-rater reliability will be calculated. The primary outcomes of the study will be: In-depth interviews (IDIs) will be conducted with five randomly selected caregivers from each arm at each site. Focus group discussions (FGDs) will be conducted with caregivers, at baseline, 6 months post recruitment and again 6 months later in both the intervention and SC arms. Among intervention arm families, qualitative techniques will be used to capture information on the feasibility, acceptability and impact of the EIP intervention using qualitative tools including FGDs, IDIs and observation. We will describe the experiences of children and caregivers relating to the intervention received including the impact of the disability, parental confidence level, inclusion in community life and experience of stigma and discrimination. We will examine changes in these domains over the follow-up period and explore attributions of change. In addition, we will perform social mapping of parent networks and group discussions with staff on their perspectives and experiences of using the EIP. The themes guiding our analysis will be drawn from objectives of the trial and from the data, should additional areas of interest emerge during interviews and discussions. The interviews will be conducted by social scientists who have experience in qualitative research. Data collected in the clinic or at field visits will be entered on standard clinical record forms (CRFs). Clinical data will be recorded under a unique study ID number. Completed CRFs will be checked by and double entered into a trial-specific MS Access database. Data from both IDIs and FGDs will be collected in the form of audiotapes, transcripts and field notes. All data entry and data management will be overseen by a statistician/data manager at the MRC/UVRI Unit. Data will be maintained on the host institution server and backed up following standardised operating procedures. Paper CRFs will be stored in lockable filing cabinets at the sites. Access to these data during the trial will be restricted to essential personnel (the principal investigators, site co-investigators, medical research officers and data clerks). All research team members will receive training in confidentiality. Data will be stored without personal identifiers, except where names must be included to ensure identification of the correct participants for procedures. All data will be stored on password-protected computers, accessible only to research team members. The trial will recruit 126 children and their caregivers, 63 per arm. Allowing for a 20% dropout rate, this sample size will give 90% power to detect a minimal relative difference of 20% on PedQL Family Impact score between the intervention and control arms, at 5% significance level, assuming a mean PedQL score of 65 in the SC arm and SD of 20 in both arms. Assumptions are based on data from the pilot study showing a mean caregiver PedQL score for families before the intervention of 64.9 (SD 19.6) and mean score of 78.9 for families after receiving the intervention (SD 17.5). The first primary outcome, feasibility of participant recruitment and randomisation, will be assessed by the total number recruited and randomised to each arm. Recruitment and randomisation feasibility will be demonstrated if the target sample size of 126 is achieved. Data on participants screened, eligible and randomised will be displayed in a Consolidated Standards of Reporting Trials flow chart. Descriptive statistics (frequencies, means, medians, SD and IQRs) will be used to describe the sample at baseline, by trial arm. The second primary outcome, acceptability, will be assessed quantitatively by (1) calculating the protocol violation rate and (2) summarising the number of programme sessions attended between baseline and programme completion for those in the intervention arm. Protocol violation rate will be calculated as the number of participants for whom one or more protocol violations occur divided by the total number of participants, and will be presented both overall, and by trial arm. For participants in the EIP trial arm, the overall number of modules attended by each participant will be tabulated. Acceptability on the basis of number of programme sessions will be defined as attendance of at least six modules. For the third primary outcome and secondary outcomes, analyses will compare outcomes between intervention and control arms at the end of the programme, when the participants will be aged 12–17 months, and again 6 months later. Analysis will be on an intention-to-treat basis and missing data will not be imputed. Data for each outcome measure will be summarised by trial arm, using proportions for binary outcomes and means or medians for quantitative outcomes, depending on normality of the distribution. Differences in means/proportions between trial arms together with 95% CIs will be calculated. We do not plan any formal statistical tests due to the preliminary nature of the trial; instead CIs will provide a possible range of effect sizes. Regression models (linear regression for continuous outcomes, logistic regression for binary outcomes) will be used to adjust comparisons for baseline measures of the outcomes, which were collected at enrolment into the trial, in order to improve precision of effectiveness estimates. For skewed continuous outcomes, data will be normalised before analysis using suitable transformations or quantile regression will be considered. No subgroup analyses are planned. Qualitative data will be analysed using a thematic framework approach. Themes will be based on the study objectives and those emerging from the data. Social scientists (two people) will agree the coding frame and undertake analysis collaboratively to ensure agreement on the coding approach. Thematic summaries will be developed and shared with the wider team for discussion. The Trial Steering Committee (TSC)22 will oversee progress of the study towards its objectives, review relevant information from other sources (eg, other related trials) and receive reports from the Data and Safety Monitoring Board (DSMB). All adverse events, whether related to the intervention or not, will be noted and reported. A Data Monitoring and Safety Committee has been established independent of the investigators and the TSC but reporting to the TSC and the sponsor. The DSMB includes an expert on global child heath, a senior statistician and a senior academic working in newborn and early child health research in Uganda, independent of the investigators. The DSMB will have access to all data on request. Resulting from the initial meeting of the DSMB on 28 June 2017, no formal stopping rules will be applied. Children with NDI and particularly those with seizure disorders and difficulties with swallowing are at increased mortality risk. All adverse events, whether related to the intervention or not, will be investigated and reported according to the UVRI Research Ethics Committee (REC) in accordance with good clinical practice requirements. All deaths, hospitalisations and other serious adverse effects will be reported to the relevant ethics committee irrespective of whether the death or event is related to disease progression or not. Trial data monitoring will be conducted by an internal independent monitor at initiation, 6 months into data collection, again after 1 year and end of data collection.