De-identified analytic dataset and reproducibility code for household material resources and between-school variation in asymptomatic Plasmodium falciparum parasitaemia among Nigerian primary-school children

Abstract

This record contains the manuscript-specific, disclosure-minimised analytic dataset and complete Python reproducibility materials supporting the clustered cross-sectional study “Household material resources and between-school variation in asymptomatic Plasmodium falciparum parasitaemia among primary-school children in southwestern Nigeria: a clustered cross-sectional study”. The study included 317 afebrile, symptom-screen-negative children aged 5–9 years attending 20 public primary schools in southwestern Nigeria between 10 March and 30 June 2024. Two co-located schools shared one pre-specified cluster identifier, yielding 19 independent school-compound clusters. The outcome was microscopy-detectable Plasmodium falciparum parasitaemia: 106 children were microscopy-positive and 211 were microscopy-negative. One microscopy-positive child lacked data for all 13 household-resource items, leaving 316 children in the primary resource-index analysis. The package contains analysis-ready CSV and Stata datasets, a data dictionary, study and dataset metadata, a standalone Python analysis script, pinned software requirements, locked expected results, citation metadata, checksums and a reproducibility validation report. The script reconstructs the 13-item principal-component household-resource index and reproduces the reported descriptive analyses, Poisson generalised estimating equation models, within-between decomposition, school fixed-effects analysis, cluster heterogeneity statistics, sensitivity analyses, leave-one-cluster-out analyses, tables and figures. The public CSV and Stata files are analytically equivalent. Independent package validation reproduced all 82 locked numerical results within an absolute tolerance of 1 × 10−10. To reduce disclosure risk, source participant and laboratory identifiers, names, school names, exact addresses, geographic coordinates, specimen-collection dates, free-text responses, biomarkers and variables not required for this manuscript were excluded. Rows were randomly reordered before fresh analytic identifiers were assigned, and no linkage crosswalk is included. Ethical approval was granted by the Oyo State Ministry of Health Research Ethics Committee, Ibadan, Nigeria. Written informed consent was obtained from each child’s parent or guardian, and children aged seven years or older provided assent. Users should cite this Zenodo record and the associated manuscript when reusing the dataset or analysis code.