• Patrick Gérardin, Christophe Rogier, Stéphane Leteurtre, Philippe Jouvencel, Amadou Sidy Ka, and Patrick Imbert.
• Department of Pediatrics, Hôpital Principal, Dakar, Senegal, Neonatal and Pediatric Intensive Care Unit, Hôpital de Terre-Sainte, Saint-Pierre, Reunion Island, France.
• Pediatr Crit Care Me. 2006 Jan 1; 7 (1): 45-7.

ObjectiveLittle is known about the use of generic severity scores in severe childhood infectious diseases. The purpose of this prospective study was to evaluate the performance of the Pediatric Risk of Mortality (PRISM) scoring system in predicting the outcome of falciparum malaria in African children.Design, Setting, PatientsAll children admitted to a 120-bed pediatric ward in a tertiary care hospital in Dakar, Senegal, with a primary diagnosis of acute malaria were assigned a PRISM score after 24 hrs or at time of death.InterventionsNone.ResultsPRISM discrimination, evaluated by areas under receiver operating characteristic curves (AUC), was good both for all acute malaria cases (n = 311; lethality, 9%; AUC, 0.89; 95% confidence interval [CI], 0.85-0.92) and for severe malaria cases (n = 233; lethality, 12%; AUC, 0.86; 95% CI, 0.81-0.90). However, the number of children who died was greater than the number of deaths predicted by PRISM (standardized mortality ratio, 2.16; 95% CI, 1.46-2.87).ConclusionThis discrepancy observed in five classes of expected mortality (Hosmer-Lemeshow chi-square test, p < .001) may have been due to chance (sample size too small for a valid test), to a lower standard of care in Dakar than in the American hospitals where PRISM was designed, or to a failure of PRISM to classify risk in severe malaria.

16 keywords...

hide…