OBJECTIVETo investigate the early predictive indices of critical condition in infants and young children with severe pneumonia, and to provide reference for diagnosis and treatment of the disease.

METHODSClinical data were collected on 411 patients (aged 1-36 months) with severe pneumonia who were admitted from January 2009 to December 2011, and multivariate logistic regression analysis was performed using 23 potential indices. These cases were divided into a critical group of 139 cases who died in hospital or needed rescue or mechanical ventilation during the course of disease and an ordinary group of 411 cases.

RESULTSEight indices with statistical significance were selected to predict the critical condition after multivariate logistic regression analysis, including hypocalcemia with the highest odds ratio (OR) (11.488), followed by sinus tachycardia (7.506), congenital heart disease (5.977), brain disorder symptoms (5.182), premature birth (4.978), blood potassium abnormality (2.910), metabolic acidosis (2.489) and malnutrition (2.048).

CONCLUSIONSThe predictive indices of critical condition in infants and young children with severe pneumonia are hypocalcemia, sinus tachycardia, congenital heart disease, brain disorder symptoms, premature birth, blood potassium abnormality, metabolic acidosis and malnutrition. The infants and young children with these risk factors need intensive care.