Neurosurgical Outcomes for Pediatric Central Nervous System Tumors in the United States.

ABSTRACT

BACKGROUND: Limited data exist on pediatric central nervous system (CNS) tumors, and the results from the National Cancer Database, the largest multicenter national cancer registry, have not previously been comprehensively reported. OBJECTIVE: To capture pediatric neurosurgical outcomes and investigate possible disparities of care. METHODS: The National Cancer Database was queried for pediatric patients who were diagnosed with CNS tumors from 2004 to 2018. Primary outcomes included 30/90 days postoperative mortality (30M/90M), readmission within 30 days of discharge (30R), and length of inpatient stay (LOS). RESULTS: Twenty four thousand nine hundred thirty cases met the inclusion criteria, of which were 4753 (19.1%) juvenile pilocytic astrocytomas, 3262 (13.1%) medulloblastomas, 2200 (8.8%) neuronal/mixed neuronal-glial tumors, and 2135 (8.6%) ependymal tumors. Patients aged 0 to 4 years had significantly poorer outcomes than patients in older age groups (90M 3.5% vs 0.7%-0.9%; 30R 6.5% vs 3.6%-4.8%; LOS 12.0 days vs 6.0-8.9 days). Tumor size was a strong predictor of poor outcomes with each additional cm in diameter conferring a 26%, 7%, and 23% increased risk of 90M, 30R, and prolonged LOS, respectively. Data over the study period demonstrated year over year improvements of 4%, 3%, and 2%, respectively, for 90M, 30R, and prolonged LOS. Facilities with a high volume of pediatric tumor cases had improved 90M (1.1% vs 1.5%, P = .041) and LOS (7.6 vs 8.6 days, P < .001). Patients with private health insurance had better outcomes than patients with government insurance. CONCLUSION: There is substantial variability in surgical morbidity and mortality of pediatric CNS tumors. Additional investigation is warranted to reduce outcome differences that may be based on socioeconomic factors.