Incidence of Severe Acute Malnutrition After Treatment: A Prospective Matched Cohort Study in Sokoto, Nigeria 2018-2019

1. COHORT TIMELINE

   This cohort study was implemented in several phases that took place sequentially, and that included recruitment of OTP-cured children, recruitment of community control children, first home visit (to collect baseline characteristics), and subsequent follow-up home visits. At each phase, different questionnaires were administered.

At each health facility, a CMAM day is held once a week, rotating through all facilities per LGA so that no two facilities within an LGA have a CMAM day on the same day per week. Children who are enrolled in the CMAM programme and their caregivers attend CMAM days for check-ups and treatment. This study recruited OTP-cured children at these CMAM days, identifying children who were discharged as cured.

The data collection timeline was as follows:

1. OTP-cured children were recruited at health facilities on a rolling basis between September and November 2018. There, the study team screened all children who had been discharged as cured on that day for eligibility and consent to participate in the study. This meant enrolling children as they were successfully discharged from the CMAM programme in the 25 health facilities that formed part of this study. At each CMAM day and health facility, teams of two interviewers were present throughout the day to ensure that all children discharged from the programme on that day were screened for possible enrolment into the study and recruited if eligible up until the minimum sample size is reached.

2. Following recruitment into the study, each OTP-cured child was tracked to their community and visited at their home within 2 to 3 weeks of their initial recruitment. Field teams used the information collected at recruitment to locate children in their community. Most communities were accessed either on foot or by motorcycle. This constituted the first home visit where a long baseline questionnaire was administered.

3. Immediately after the first home visit of each OTP-cured child, a search for a suitable community control for that child was conducted. For each OTP-cured child, potential community controls were identified using a snowball approach. In essence, this approach meant that interviewers were referred to potential community control children by the mother of the OTP-cured child. Potential community controls were assessed with respect to their eligibility to enter the study and to whether they matched the corresponding OTP-cured child on a set of criteria (mentioned in the sampling section). The first community control to meet both sets of criteria was recruited into the study and the search for a community control for a given OTP-cured child ended at that stage. Once a control child was identified, the same first home (baseline) questionnaire was administered to the household and mother of that child.

4. Afterwards, both cohorts were followed-up fortnightly for a total of 12 home visits (the 12 visits includes the first baseline home visit).

5. Participation in the study for both cohorts ended at the 12th home visit, unless a child developed SAM earlier or dropped out of the study (e.g. family no longer consented to participate in the study or moved out of the community, or child had died), whichever came first. In total, children were followed up for a duration of up to six months after discharge from OTP. Within the six months of follow-up, if children were identified as being SAM by the field team, they proceeded to exit the study and interviewers referred those children to CMAM services.

2. DESCRIPTION OF QUESTIONNAIRES

There were a number of questionnaires that were used for this study.

During recruitment of OTP-cured children at the health facility

A recruitment questionnaire for OTP-cured children was administered to the mother of the child on the day children were discharged and recruited into the study. This questionnaire assessed eligibility of the child and collected some information to help with locating the home of the child.

Additionally, data on children's health status at admission and discharge from the OTP were also collected from registration and treatment tracking cards kept at the facility by staff (OTP cards and Ration cards). This data was scanned on enumerator's tablets and later on entered into a database by data entry staff. Information that was entered from these records included anthropometric measurements and morbidity at admission, duration of treatment and anthropometric measurements at each visit to the OTP. Note that data from the scanned OTP and Ration cards has not been uploaded for public use due to data quality concerns. The data suffers from many missing observations, given that this data was not directly collected by the enumerators but relied on health facility staff filling in the OTP and Ration cards for the treated children.

During this phase, a health facility questionnaire was also administered in each health facility to assess adherence of the health facility to the Nigeria CMAM national guidelines and availability of OTP-related drugs and equipment and the general quality of infrastructure and resources. The survey also collected data on shocks that affected the catchment area of facilities in the year prior to the survey, such as drought, floods, sandstorms, and security-related events. In each health facility, this survey was administered once, on the first day the interview team visited the health facility. The survey used direct observation as well as interviews with the head of the health facility and the CMAM focal person in charge. If either of these individuals were not available on the day, other knowledgeable health facility member was asked to respond to the questions.

During recruitment of community control children

A recruitment questionnaire for community control children was administered to the mothers of the children to assess eligibility and matching criteria and decide if they can be recruited.

During the first home visit

At the first home visit, a long baseline questionnaire was administered to the mother of the recruited child and the household head to collect baseline information across several domains related to the child, mother, and household. Children's MUAC was measured using the WHO/UNICEF-recommended MUAC tape and measurement protocol, whereas height and length were measured with a precision of 0.1 cm, using boards manufactured by SECA: standing boards for children who were able to stand and lying-down boards for children unable to. The domains included in the baseline questionnaire include:

• Child level: height/length, mid-upper arm circumference; demographics; breastfeeding history; co-morbidities in the 2 weeks prior to the survey; immunization status; dietary diversity (24 hours prior to survey).
• Mother level: demographics; economic activity and education status; knowledge on child feeding and health-seeking behaviour; reproductive history and care; perceived OTP experience; networks in community.
• Household level: household demographics and composition; economic activity and education of household head; household assets and wealth; water, sanitation, and hygiene infrastructure; household food security and dietary diversity; deaths in the household in the year prior to the survey.

During the follow-up home visits

At each follow-up home visit, a short follow-up questionnaire was administered to the mother of the child to collect child-level co-morbidity data in the 2 weeks preceding the visit (a subset of the questions asked in the baseline questionnaire) and to measure the child's MUAC.

In the final follow-up visit (i.e. the visit when the child exited the study either because they developed SAM or if they reached the final 12th visit), additional questions were asked of the household including on mother's employment status, changes in the breastfeeding and pregnancy status of mothers, deaths in the household, household food security, child feeding, household and child dietary diversity, and mother's feeding knowledge and practices. These questions were a subset of those asked in the baseline questionnaire and they were added in order to understand if household, mother, or child conditions assessed at the first home visit might have changed at the point of exit. Note that these additional questions were not asked of children who dropped out of the study (because the interviewers would not have known that the previous visit was going to be the final exit visit).

All questionnaires were administered using the Computer-Assisted Personal Interviewing software CSPro (Version 7.1.3), and OTP and Ration cards were scanned and data entered digitally using the SurveyCTO software. Questionnaires were translated into Hausa and administered to all respondents in Hausa.

3. NOTE ON THE DEFINITION OF SAM

   SAM was determined using the WHO and national MUAC criteria of MUAC <115 mm. Given that this study's objective was to identify definite relapses and cases of SAM that would require treatment, we classify a child as having SAM if his/her MUAC =112 mm at any home visit or if his/her MUAC is between 112 and 115mm for two consecutive visits. The reason we do this is to account for the possibility of measurement error, i.e. it is difficult to identify whether children around the 115mm MUAC cut-off temporarily dip into SAM or whether they are a certain SAM case that requires treatment.

4. NOTES ON THE DATA COLLECTION THAT MIGHT BE RELEVANT FOR DATA ANALYSIS

   Survival analysis techniques can be used to analyse the data.

An important point to emphasise is that for OTP-cured children there was a lag of up to three weeks between their recruitment at the health facility and the first visit at home (where we collected baseline data). Some children had already relapsed into SAM at the first home visit before additional data on these children could be collected. For those children who relapsed between recruitment and the first home visit, it would therefore not be possible to assess whether certain time-varying characteristics collected at the first home visit - e.g. child-level health indicators - materialised as a consequence of relapse or prior to relapse. This is a limitation and could present implications for data analysis, depending on the type of analysis the users of the data wish to conduct. Specifically, it is important for the analysis not to suffer from endogeneity if for instance users are interested in assessing the effect of certain factors on relapse rates. There are options to deal with this limitation, for instance, i) limiting the analysis to the factors/covariates that could reasonably be assumed to be time-invariant between recruitment and the first home visit, or ii) defining the time origin for OTP-cured children as the first home visit (as opposed to their recruitment from the health facility) and restricting the analysis to the subsample of children that had not relapsed into SAM at the first home visit (though this option would entail a significant reduction in sample size).