In our analysis of 26 106 patients admitted to six hospitals in five sub-Saharan African countries, we found transfusion to be associated with decreased odds of death in site-adjusted and severity-adjusted analysis (odds ratio [OR] 0·50, 95% CI 0·42–0·60). This association could be subject to immortal time bias if many patients died before having an opportunity to receive transfusion—a concern that James Watson and colleagues have raised and that we explicitly stated as a possible limitation in our Article.
Infection with Plasmodium falciparum leads to severe malaria and death in approximately 400 000 children each year in sub-Saharan Africa. Blood transfusion might benefit some patients with malaria but could potentially harm others. The aim of this study was to estimate the association between transfusion and death among children admitted to hospital with P falciparum malaria.