Epidemiology and mortality predictors for severe childhood community-acquired pneumonia in ICUs: A retrospective observational study.

Abstract

To identify the epidemiology and mortality predictors of severe childhood community-acquired pneumonia (CAP) and evaluate the influence of medications on clinical outcomes in the real world. We conducted a multicenter retrospective observational study among children aged ≤5 years with severe CAP, separately comparing the detailed information between those who experienced in-hospital death and those who survived in three different age groups. A multivariate logistic regression model was used to determine mortality predictors. A total of 945 children were recruited: 341 young children aged 2-59 months, 47 infants aged 29 days to 2 months, and 557 neonates aged less than 28 days. A total of 88 deaths occurred (9.3%). There was low adherence to antimicrobial guidelines in the group aged 2-59 months, and carbapenems widely served as initial empirical regimens. However, analysis of all three age groups showed that the efficacy of antibacterial drugs with initial empirical selection grades higher than those recommended by the guidelines was not better than that of antibacterial drugs with grades recommended by the guidelines. In multivariate analyses, very severe pneumonia (odds ratio (OR): 3.48; 95% confidence interval (CI): 1.36-8.93), lower birth weight (OR: 4.64; 95% CI: 1.78-12.20), severe underweight (OR: 6.06; 95% CI: 2.34-15.63), mechanical ventilation (OR: 2.58; 95% CI: 1.00-6.62; OR: 15.63; 95% CI 3.25-76.92), a higher number of comorbidities (OR: 8.40; 95% CI: 1.89-37.04), comorbidities including anemia (OR: 5.24; 95% CI: 2.33-11.76) and gastrointestinal hemorrhage (OR: 3.79; 95% CI: 1.36-10.53), and the use of sedative-hypnotics (OR: 2.60; 95% CI: 1.14-5.95) were independent risk factors for death; infants treated with probiotics had a lower mortality rate (OR: 0.14; 95% CI: 0.06-0.33). Severe CAP remains a primary cause of death in children under 5 years of age. Clinical characteristics, comorbidities and medications are evidently associated with death. Importantly, we should pay particular attention to the identification of mortality predictors and establish prophylactic measures to reduce mortality.