Abstract / Summary
Objectives Risk stratification tools for childhood pneumonia may improve identification of children at highest risk of death in low-resource settings. However, their added value over the WHO Integrated Management of Childhood Illness (IMCI) criteria and danger signs remains uncertain. We compared the performance of published paediatric pneumonia risk scores with WHO-IMCI criteria and WHO danger signs using data from a multi-country African clinical trial. Design Secondary analysis of a multi-country randomised controlled trial Setting Thirteen hospitals in Mozambique, South Africa, Uganda, Zambia, and Zimbabwe. Participants Children aged 2-59 months without HIV hospitalised with WHO-defined severe community-acquired pneumonia in the PediCAP trial. Primary and secondary outcome measures The primary outcome was in-hospital mortality. Secondary outcomes were mortality during hospitalisation or within 7 days of discharge, and the exploratory outcome of readmission or death by day 28. Discrimination was assessed using area under the receiver operating characteristic curve (ROC-AUC). Results Among 1052 children, 18 (1.7%) died in hospital and 22 (2.1%) died in hospital or within 7 days of discharge. Among 1010 children with day-28 follow-up, 63 (6.2%) died or were readmitted by day 28. ROC-AUC point estimates for in-hospital mortality ranged from approximately 0.73 to 0.87. WHO danger signs had lower point estimate than the other scores, although confidence intervals overlapped substantially and no score demonstrated clear superiority. Similar findings were observed for mortality during hospitalisation or within 7 days of discharge. In contrast, all approaches showed limited discrimination for the exploratory outcome of 28-day readmission or death (ROC-AUC approximately 0.54-0.57). Conclusions In this multi-country cohort of children hospitalised with severe pneumonia, no score demonstrated clear superiority for mortality prediction, although modest differences cannot be excluded. Published risk scores and WHO-IMCI criteria showed broadly similar performance. These findings suggest that strengthening implementation of WHO clinical criteria may be at least as important as introducing additional risk stratification tools in low-resource settings.