Critical illness epidemiology and mortality risk in pediatric oncology

Objective Pediatric oncology patients admitted to the pediatric intensive care unit (PICU) are at high risk of mortality. This study aims to describe the epidemiology of and the risk factors for mortality in these patients. Study design This is a retrospective cohort study including all consecutive PICU oncology admissions from 2011 to 2017. Demographic and clinical risk factors between survivors and nonsurvivors were compared. Both univariate and multivariate Cox proportional hazard regression models were used to quantify the association between 60-day mortality and admission categories, accounting for other covariates (Pediatric Risk Of Mortality [PRISM] III score and previous bacteremia). Main outcome measures The primary outcome was 60-day mortality. Results The median (interquartile range) age and PRISM III scores of pediatric oncology patients admitted to the PICU were 7 (3, 12) years and 3 (0, 5), respectively. The most common underlying oncological diagnoses were brain tumors (73/200 [36.5%]) and acute lymphoblastic leukemia (36/200 [18.0%]). Emergency admissions accounted for approximately half of all admissions (108/200 [54.0%]), including cardiovascular (24/108 [22.2%]), neurology (24/108 [22.2%]), respiratory (22/108 [20.4%]), and "other" indications (38/108 [35.2%]). The overall 60-day mortality was 35 of 200 (17.5%). Independent risk factors for mortality were emergency respiratory and neurology categories of admission (adjusted hazard ratio[aHR]: 5.62, 95% confidence interval [95% CI]: 1.57, 20.19; P = .008 and aHR: 6.96, 95% CI: 2.04, 23.75; P = .002, respectively) and previous bacteremia (aHR: 3.37, 95% CI: 1.57, 7.20; P = .002). Conclusion Emergency respiratory and neurology admissions and previous bacteremia were independent risk factors for 60-day mortality for pediatric oncological patients admitted to the PICU.