# # Key findings
- Study scope: 103 DHS/MICS-based studies from sub-Saharan Africa (41 on diarrhea, 62 on ARI).
- Missing data reporting: Only 3 diarrhea studies (7.3%) and no ARI studies explicitly reported how they handled missing data.
- Analytic methods: Complete-case analysis (CCA) was used in 80.5% of diarrhea studies and 100% of ARI studies. Multiple imputation (MI) was used in only two diarrhea studies (4.9%) and no ARI studies.
- Empirical missingness: In the Nigeria DHS 2018 validation sample (N = 33,924 under-five children), the missingness rate for both diarrhea and ARI outcomes was 9.8% across the full analytical sample.
- Impact on effect estimates: Using clean cooking fuel as an example, CCA yielded an ARI odds ratio of 0.82 and a diarrhea odds ratio of 0.66; multiple imputation yielded ORs of 0.81 and 0.64, respectively, with biases of -0.9% and -3.5%.
# # Limitations
The review notes several constraints: the empirical missingness analysis is based solely on the Nigeria DHS 2018 validation dataset, limiting generalizability to other sub-Saharan African countries and survey years; included studies may not be exhaustive despite a multi-database search; assessment of missing data handling relies on what authors reported rather than independent data verification; and no citation data are available to assess the uptake or influence of the findings.