The Kusamala Program for primary caregivers of children 6-59 months of age hospitalized with severe acute malnutrition in Malawi: Study protocol for a cluster-randomized controlled trial

Background: Severe acute malnutrition (SAM) is associated with high mortality rates and impairments in growth and development in children that do survive. There are complex nutritional, health, and behavioural risk factors involving severely malnourished children and their primary caregivers, requiring integrated intervention approaches. Methods: A cluster-randomized controlled trial at the Queen Elizabeth Central Hospital in Blantyre, Malawi will be conducted to evaluate the effectiveness of a 4-day hospital-based intervention programme directed at primary caregivers. This programme, titled the Kusamala Program, aims to improve developmental and nutritional outcomes in children with SAM. Up to six primary caregivers and their children will be enrolled to groups each week, which will be randomly allocated to intervention or comparison arms. The intervention package consists of interactive counselling on three modules: 1) nutrition and feeding; 2) water, sanitation, and hygiene (WASH); and 3) psychosocial stimulation. Data collection will be performed at enrolment, at discharge from hospital, and at 6 months following discharge. The primary outcome is child development assessed with the Malawi Developmental Assessment Tool (MDAT), a validated measure of gross and fine motor, language, and social development. Discussion: This intervention programme is unique because it utilizes primary caregivers’ time spent in-hospital while children receive treatment for SAM. The programme has the potential to be effective in addressing multiple aspects of child, nutrition and development. Trial registration: ClinicalTrials.gov, NCT03072433. Registered on 7 March 2017.

The Standard Protocol Items Recommendations for International Trials 2013 Statement has been followed for the reporting of this protocol (see Additional file 1) and for reporting the timeline of the trial (Fig. 1) [21]. Timeline of enrolment, intervention and comparison, and assessments The primary objective of the trial is to evaluate the effects of a 4-day hospital-based intervention for primary caregivers of children with SAM on child development and nutritional status after 6 months. The secondary objective is to understand the effects of this programme on knowledge, attitudes, and practices, as well as mental health status of primary caregivers. The hypothesis is that the intervention programme for primary caregivers will improve developmental and nutritional outcomes in children hospitalized with SAM. This study will evaluate a comprehensive structured intervention package for primary caregivers of children aged 6 to 59 months hospitalized with SAM. It will be a pragmatic trial aimed at replicating how this type of intervention could be delivered in a real nutrition rehabilitation unit (NRU) setting and its effectiveness in this setting [22]. The primary outcome is child development, which will be assessed using the Malawi Developmental Assessment Tool (MDAT), a locally adapted and validated tool for evaluating multiple domains of child development (i.e. gross motor, fine motor, language, and social domains) [23]. The design of the study is a cluster-randomized controlled trial (superiority design) that will be conducted at the Moyo NRU of the Queen Elizabeth Central Hospital in Blantyre, Malawi ({“type”:”clinical-trial”,”attrs”:{“text”:”NCT03072433″,”term_id”:”NCT03072433″}}NCT03072433). A cluster-randomized controlled trial design was selected because the intervention is most feasible if delivered to groups of primary caregivers rather than individual caregivers. This design will also prevent temporal overlap between the two study arms: 1) an intervention arm of primary caregivers and children with SAM participating in an intervention programme; and 2) the comparison arm of primary caregivers and children with SAM receiving standard care. All other clinical and nutritional aspects of care will be equivalent; children will be treated per hospital protocol and WHO guidelines [24]. One day of each week will be used to recruit and enrol an estimated average of four primary caregiver-child pairs. Once formed, the group will be randomly assigned to either the intervention or comparison arm. Recruitment and enrolment will not overlap with the delivery of intervention sessions, although it is possible that children from different groups will be in the NRU at the same time depending on the duration of their in-patient treatment. An internal pilot study (exploratory design) will be conducted within the full randomized trial and will reflect the design of the full trial [25, 26]. The internal pilot study will include at least the first 15 primary caregiver-child pairs recruited to each arm, for a minimum of 30 primary caregiver-child pairs [25, 26]. As there are few previous trials evaluating long-term development in children with SAM, this internal pilot study will inform the design of the larger randomized control trial by estimating the variance of MDAT scores. This will allow for a more accurate sample size calculation. The internal pilot study will also establish the adherence and engagement rates of participants to the intervention [25]. Children admitted to hospital with complicated SAM and their primary caregivers will be recruited at the NRU after being screened for eligibility within 3 days of admission to the NRU (–t 1). It is estimated based on hospital records from the NRU that there will be sufficient SAM admissions to recruit primary caregivers and children to each group. Prior to seeking informed consent, information about benefits and risks of this study will be provided verbally and on an information sheet in Chichewa to the primary caregivers of children admitted to hospital. Written informed consent will be requested of the primary caregivers on behalf of themselves and their children (–t 1). For those unable to read or write, the information will be read to them in Chichewa. In these cases, informed consent by signature or thumbprint will be observed by an impartial witness. This study will not exclude children with cerebral palsy and other types of neurodisability as they represent a significant portion of admitted children with SAM and could benefit from the intervention. The Kusamala Program is a 4-day counselling programme, led by NRU nurses, for primary caregivers of children with SAM in an NRU setting. Kusamala means “to take care” in Chichewa. The programme incorporates multiple modules aimed at addressing the diverse causes and consequences of SAM. The modules include: Each of the 4 days (t 1 –t 4) of the Kusamala Program involves 45 min of counselling performed in the back bay of the NRU, which is separated from the rest of the ward by a wall. The first three sessions cover each of the three modules (i.e. one module per session), whereas the final day is a summary of all modules. Materials for the first two modules of the Kusamala Program were based on adaptable nutrition and WASH communication and support materials from the United Nations Children’s Fund (UNICEF) [27, 28]. Nutritional messages and materials for children above 2 years of age have been added since these UNICEF materials focus on infants and young children. Furthermore, there will be an additional emphasis on prevention and early identification of malnutrition. For the psychosocial stimulation module on the third day of the intervention, the WHO Care for Child Development Package will be used to promote sensitive and responsive caregiving for age- or ability-appropriate play and communication activities [29]. Psychosocial stimulation is an integral component of the Kusamala Program. For this reason, primary caregivers and their children are also involved in 45 min of supervised and interactive play on each of the 4 days after the counselling sessions. A basket of play items is available in the NRU for children and primary caregivers to use during the play sessions. There are traditional Malawian toys, including small drums, rattles, and shakers, in addition to “Western” toys, such as colourful stacking blocks. Primary caregivers can also make their own toys such as shakers by adding beans or bottle caps to empty plastic jars available in the basket. Nurses supervise these sessions and actively encourage primary caregivers to engage in play and communication activities appropriate for the age and stage of development of their children. Four behaviour change techniques including information, materials, media, and performance have been integrated into the Kusamala Program. Results from 24 studies to improve health and nutritional outcomes showed that effect sizes on outcomes were higher when three or four techniques were applied [30]. Firstly, nurses use a flipbook with coloured images to provide information to participants [30]. Primary caregivers will be given: two take-home images of nutrition and WASH messages, respectively (i.e. media); a toy that is appropriate for their child’s stage of development (i.e. materials); and a certificate of completion of the Kusamala Program [30]. During the intervention sessions, nurses also involve primary caregivers in practicing relevant activities including feeding, handwashing, and play (i.e. performance) [30]. Participants in the comparison group will receive the standard care at the NRU including clinical stabilization and nutritional rehabilitation. Besides this, nurses will instruct primary caregivers to take their children to the area of the NRU where play items will be available for use during their hospital stay. However, nurses will not be involved in the same interactive psychosocial stimulation sessions in which participants in the intervention groups partake. Participants will not be counselled about the importance of play; primary caregivers will not be advised which toys are appropriate and the play sessions will not be supervised. Having primary caregivers and children go to the back bay also acts as a method of blinding study personnel and participants so they are unaware of the allocation. At the time of discharge from the NRU, nurses counsel primary caregivers on nutrition and WASH for an estimated 15 min. Nurses have been trained on nutrition and WASH messages to discuss with participants at discharge, yet this counselling does not follow a strict protocol. Five Moyo NRU nurses were trained to conduct the Kusamala Program. Prior to the training they were given a test based on the Food and Agriculture Organization Guidelines for assessing nutrition-related knowledge, attitudes and practices to determine their baseline understanding [31]. Questions from the following modules were used: feeding young children (6–23 months), malnutrition, personal hygiene, water and sanitation, and food-based dietary guidelines [31]. Mean scores out of a possible 25 points were 21.8 ± 0.8, indicating that nurses already had familiarity of nutrition and WASH components as expected considering their work in the NRU. Nurses then received training for a duration of 6 hours and, over the next 2 weeks following training, the nurses practiced at least two sessions with primary caregivers and children in the ward. Nurses were then given the same test; mean scores were 24.2 ± 1.3. Differences between the pre-test and post-test were significantly different, showing an improvement in mean scores of 2.4 points (p = 0.02). Refresher training of NRU nurses will be conducted bi-annually. In addition, nurses received training in care for child development. The primary outcome is child development 6 months after discharge (t 6). Child development will be assessed with the MDAT. The MDAT has been evaluated in 1426 healthy Malawian children as well as children with neurodisability (n = 80) or with severe wasting (n = 120) [23]. In this evaluation, sensitivity was 97%, and 18% of healthy children failed, translating to specificity of 82% [23]. These results show that the MDAT is an appropriate tool for assessing child development in Malawi. The MDAT will be used at discharge (t 5) and 6 months later (t 6) to examine development in four domains: gross motor, fine motor, language, and social development [23]. Children will be given a pass or fail depending on their ability to complete increasingly advanced items on the MDAT [23]. The sum of passed items can be calculated and translated to age-specific z scores based on healthy Malawian children as reference standards [23]. Data will be collected at three time points: at enrollment (–t 1), discharge from hospital (t 5), and 6 months after discharge (t 6). All outcomes and measures are described in Table 1. Primary, secondary, and implementation outcomes and measures Important child outcomes are the nutritional status assessed by anthropometry (i.e. MUAC, WLZ or WHZ, oedema, and height-for-age z scores) and dietary intake based on a locally adapted 24-h dietary recall which will be performed only at follow-up (t 6) [32, 33]. Readmission to hospital and mortality at any time point are other secondary outcomes of interest. Child co-variates that will be considered include sex, age, and HIV status. There are important co-variates in relation to households and primary caregivers measured at enrolment (–t 1) that could influence the ability of primary caregivers to care for their children during hospital treatment and in the home, and both are known to influence child outcomes [12, 18, 34]. First, the mental health status of primary caregivers will be measured with the Self-Reporting Questionnaire, which has been translated and validated for use in Malawi [35, 36]. Maternal depressive symptoms have been shown to be significantly associated with adverse outcomes in infants such as cognitive and socioemotional problems [12, 18, 34]. Furthermore, nutritional status of primary caregivers according to two anthropometric outcomes, body mass index and MUAC, will be assessed. Other measured co-variates that could influence child outcomes include HIV status, age, education level, and marital status of the primary caregiver as well as location (i.e. urban or rural) and monthly income of the household, for example. Primary caregiver outcomes of interest at 6 months (t 6) will also include mental health status, nutritional status, as well as knowledge, attitudes, and practices relating to malnutrition, nutrition and feeding, and WASH based on the WHO Indicators for assessing infant and young child feeding practices [33]. Lastly, the environment in the home will also be assessed at follow-up 6 months after discharge (t 6) using an adapted and translated version of the Home Observation of the Measurement of the Environment (HOME) Inventory [37]. The HOME Inventory includes observations and interview questions to the primary caregiver to establish the amount and quality of interaction and input in the child’s home environment; this may be related to developmental outcomes in children [11, 37]. Participant engagement and adherence to the Kusamala Program will be evaluated based on the proportion of participants that attend the first day of the programme (t 1) and the number of participants that attend all 4 days of the programme (t 1–t 4), respectively. Fidelity of 20% of the intervention sessions will be assessed using a tool designed to capture duration of sessions, delivery of specific intervention content, and counselling skills according to the WHO Care for Child Development Package [29]. Following the intervention, knowledge and attitudes of primary caregivers who participate in the Kusamala Program will be demonstrated with the use of a questionnaire at discharge from hospital and at follow-up. The sample size estimate is based on cross-sectional study data of development of children with SAM at discharge from the NRU using the MDAT [38]. Results from this study showed that, based on MDAT z scores, the variance for gross motor was 1.9 SD, for fine motor was 1.2 SD, and for language was 1.7 SD [38]. Therefore, the composite variance for those domains was 1.6 SD. This was used to calculate the sample size using α = 0.05 and 80% power, an estimated intracluster correlation coefficient of 0.001, and with an expected effect size of 0.5 using Cohen’s d [39, 40]. With this approach, the initial estimate of the sample size for the primary outcome is a minimum of 160 primary caregiver-child pairs per arm (n = 320). As discussed previously, an internal pilot study will be conducted to determine the variance of the primary outcome in the first 15 children in each study group at 6 months post-discharge from hospital which will be used to recalculate the sample size [25]. A biostatistician (CB) completed the computer-generated random allocation sequence for each study week and placed the sequence in sealed envelopes to be opened at the start of each week by a nurse at the NRU. NRU staff (apart from NRU nurses) and those doing screening, recruitment, enrolment, and data collection will be blinded. There are no expected reasons for un-blinding of study team members to occur. Blinding of the analysis will be performed by a biostatistician with the use of a dummy variable to code for each group. A data monitoring committee will not be necessary to analyse results un-blinded during the trial because the intervention is of very low risk. Nonetheless, an interim analysis will be completed with the internal pilot study data. Data collection will be performed by three different enumerators who have been trained on all relevant assessment tools and questionnaires, including the MDAT. Baseline data collection will take place over approximately 80-week-long cycles, depending on enrolment rates, with assessments performed at enrolment (–t 1) and at discharge (t 5). These data collections will be done privately in the NRU, either in the procedures room or in the back bay depending on availability of these locations. Home visits will be conducted for follow-up assessments 6 months after children are discharged from hospital (t 6) over the course of 80 weeks. For participants who prefer not to be visited in the home, they will be asked to come back to the NRU for follow-up assessments. Investigators will maintain medical and research records for this trial in compliance with regulatory and institutional requirements for the protection of confidentiality of subjects. All personnel with access to data and information related to this trial will take all reasonable precautions to maintain the confidentiality of information. Data will be collected using paper forms which will be kept in a locked filing cabinet and will be accessed only by the investigators. Data entry will be performed into a Research Electronic Data Capture database, which includes automated checks for data values with coded numbers [41]. De-identified data will be exported in a password-encrypted file. Bias will be determined by maintaining a screening log of all potentially eligible participants and those who were excluded or recruited to the two arms. This will be presented as a flow diagram in the final report according to the Consolidated Standards of Reporting Trials (CONSORT) guidelines (Fig. 2) [42]. Modified version of the CONSORT 2010 flow diagram Data will be analysed using the statistical software Stata 14, R v3.2.5, and WHO Anthro v3.2.2 for anthropometric z-score calculations [43–45]. Missing data will be accounted for depending on the nature of drop-outs and loss to follow-up. Descriptive statistics will be performed to summarize baseline characteristics of participants in both the intervention and comparison groups. More specifically, means and SD will be calculated for continuous variables such as the age of children, whereas the number of participants and percentages will be calculated for categorical variables such as the presence of oedema. MDAT z scores standardized by age and based on a reference population of Malawian children will be calculated [23]. Z scores of each domain will be analysed separately to understand if the intervention influences certain domains more than others and a composite score of overall child development will be calculated if statistically appropriate. Upon completion of data collection for the internal pilot study, the variance of the primary outcome, child development using the MDAT, will be determined. The variance will be used to recalculate the sample size for the full trial [25, 26]. To determine the effect of the intervention on the primary outcome measures and continuous secondary outcomes, analysis of co-variance (ANCOVA) will be conducted to determine if there are differences between the intervention groups and the comparison groups while allowing for adjustment of co-variates in the analysis [46]. Multiple logistical regression will be performed to determine whether the intervention influenced binary outcomes such as the presence or absence of oedema. Subgroup analyses will be performed with participant age categories (i.e. infants and older children) and types of SAM (i.e. marasmus and kwashiorkor), and neurodisability, since children with neurodisability such as cerebral palsy are included in this study. NRU staff members involved in the trial, including nurses, will be taught about background information of the trial, scientific and ethical aspects of clinical research and trials, data collection tools, and communication to participants. Written consent for all participants will be taken by an enumerator after screening. All of these NRU staff members have completed Good Clinical Practice training and are fluent in Chichewa and English. Participation in this study is completely voluntary for primary caregivers in the intervention and comparison groups. If primary caregivers are enrolled but choose to withdraw from the study at any time, they can do so without being asked to respond and without consequences. Confidentiality and respect of privacy of participants and those that withdraw will be maintained. The Kusamala Program, which will be provided to participants in the intervention groups, is of very low risk and of potential benefit. The intervention package gives primary caregivers an opportunity to learn more about nutrition, health, and development of children. Adoption of caregiving practices requires behaviour change, which is dependent on the situation, motivation, and decisions of the primary caregivers. If primary caregivers choose to adopt recommended care practices, there may be positive outcomes for children involved in the study, other children in the household, and/or primary caregivers themselves. Ethical approval for this study was obtained from the College of Medicine Research and Ethics Committee (P.04/16/1930) in Blantyre, Malawi, and the Research Ethics Board (1000053578) at the Hospital for Sick Children in Toronto, Canada. If any modifications are to be made to the study protocol, amendments will be submitted to both specified ethics boards and will be communicated to investigators and other study team members. It is expected that there will be no adverse events as a result of this low-risk intervention, yet if any adverse events do occur they will be recorded on a Hospital for Sick Children Adverse Event form within 48 h and will be a component of continuing review by the Hospital for Sick Children Research Ethics Board per the Tri-Council Policy Statement. Dissemination of results to the participants is difficult in this study primarily for logistical reasons. Firstly, it will take place in a low-resource setting with many participants living far from the NRU. Upon completion of the study, primary caregivers are unlikely to return to the hospital if their child is doing well. Results of the trial with lessons learned will be communicated to colleagues at the NRU with the goal of transforming health services delivery for children admitted for in-patient treatment of SAM. Lessons learned and results will also be shared with collaborators and other international researchers through activities such as presentations at global health and nutrition conferences. To ensure that the study findings reach a broad audience, manuscripts will be submitted to open access peer-reviewed journals. Authorship will be granted for investigators and other personnel who play a significant role in the design and implementation of the intervention, data collection and management, analysis of results, or writing of the manuscript for this study. All authors will have access to the final trial dataset.

• daniel, allison i.
• van den heuvel, meta
• voskuijl, wieger
• gladstone, melissa
• bwanali, mike
• potani, isabel
• bourdon, celine
• njirammadzi, jenala
• bandsma, robert h.j.