Background

Vaccination coverage levels fall short of the Global Vaccine and Action Plan (GVAP) 90% target in low- and middle- income countries (LMICs). Having identified traditional and religious leaders (TRLs) as potential public health change agents, this study aimed at assessing the effect of training them to support routine immunisation for the purpose of improving uptake of childhood vaccines in Cross River State, Nigeria.

Methods

A cluster-randomised controlled study was conducted between 2016 and 2019. Of the 18 Local Government Areas (LGA) in Cross River State, eight (four urban and four rural LGAs) were randomized into the intervention and control study arms. A multi-component intervention involving the training of traditional and religious leaders was implemented in the four intervention LGAs. Baseline, midline and endline surveys collected information on children aged 0-23 months. The effect of the intervention on outcomes including the proportion fully up-to-date with vaccination, timely vaccination for pentavalent and measles vaccines, and pentavalent 1-3 dropout rates were estimated using logistic regression models using random effects to account for the clustered data.

Results

A total of 2598 children at baseline, 2570 at midline, and 2550 at endline were included. The intervention was effective in increasing the proportion with at least one vaccine (OR 12.13 95% CI 6.03-24.41p<0.001). However, there was no evidence of an impact on the proportion of children up-to-date with vaccination (p=0.69). It was effective in improving timeliness of Pentavalent 3 (OR 1.55; 95% CI: 1.14, 2.12; p= 0.005) and Measles (OR 2.81; 96% CI: 1.93-4.1; p<0.001) vaccination. The odds of completing Pentavalent vaccination increased (OR=1.66 95% CI: 1.08,2.55).

Conclusion

Informal training to enhance the traditional and religious leaders’ knowledge of vaccination and their leadership role can empower them to be good influencers for childhood vaccination. They constitute untapped resources in the community to boost routine immunisation.

Selection of children for the surveys

Independent household surveys were carried out at baseline, mid-line (after 9 months of intervention) and end-line (after 18 months of intervention).

 In each selected village, 25 households with children aged 0-23 months were selected using the WHO spin-the-pen method because there was no list of all households. A team member dedicated to sampling of households went to the centre of the village and spun a bottle to choose a random direction. The “sampler” then walked in the direction indicated until the edge of the village was reached, sketching a map of all the households passed, and numbering them as they went. One of these houses was selected at random as the starting point, or “house 1” of the village. At this house, a bottle was spun to choose a random direction, and the sampler walked in that direction until they came to another household, which was the second house of the village, and so on. If there was a junction in the path, the bottle was spun again to select from the choices available. This procedure was repeated until 25 households with children were counted.

Questionnaire

A semi-structured interviewer-administered questionnaire was used for the survey on immunisation coverage. The questionnaire was adapted from the WHO vaccination coverage tool. Verbal consent was obtained from the child’s caregiver before applying the questionnaire. Information was collected on household characteristics, immunization status of the child, caregiver’s knowledge of vaccination, prevalence of selected childhood diseases, mother’s health facility utilization, and delivery. Data on the child’s immunization history was extracted from the vaccination card. When this was not available parental recall was used. Open Data Kit (ODK) was used for data collection Coding of the paper tool into the mobile device included the creation of built-in data validation logic, constraints and loops.

Statistical analysis

The effect of the intervention was estimated using logistic regression, including random effects for village, Ward and LGA was used to account for the non-independence of the clustered observations. We included the baseline survey in the model to allow each village to account for baseline differences. The effect of the intervention was estimated as the additional increase in vaccine uptake between the baseline and either the midline or endline surveys in the intervention arm compared to the control arm. Covariates such as the age of caregiver, residing in a hard-to-reach community, distance to the nearest health facility, hard-to-reach and rural/urban setting were assessed for imbalance at baseline. The analysis was carried out in R.

# TRL_study
data and code
<br>
This repository includes anonymised data for the three surveys (baseline, midterm and final)<br>
<br>
Code for data preparations and analysis are provided as R scripts <br>
master.r -  describes and calls the other R scripts in order