Introduction: Despite meticulous efforts in providing affordable insurance coverage, disparities in access to healthcare resources and outcomes seem evident. We investigate racial disparity in short-term outcomes and costs in pediatric patients undergoing craniotomy for brain tumor resection.

Methods: Study design: An observational, cohort study using the HCUP KID databases to identify all patients harboring malignant, benign (meningioma and vestibular schwannoma) and pituitary tumors was formulated. Outcome endpoints: Inpatient mortality, discharge disposition, LOS, cost, and post-operative complications (cardiac; neurological; DVT; pulmonary embolism; treated hydrocephalus, wound infections and complications). Exposure variable: Primary exposure variable of interest was to evaluate racial differences in outcome endpoints (African Americans, Hispanics, Asians, others with reference to Caucasians). Statistical methods: Each tumor sub-type analyzed separately by constructing multivariable models to evaluate the likelihood of individual endpoints across all ethnic groups with reference to Caucasian. Models are adjusted for patient demographics (age, sex, race, insurance, quartiles of median household income based upon residential zipcode); Hospital characteristics (annual case-volume; bed-size, academic status, region); and medical comorbidities as stratified by the scoring method proposed by Rhee et al for risk-adjustments for pediatric patients undergoing major surgery. All models are fit with generalizing estimating equations using sandwich-covariance matrix estimator to restrict clustering of similar outcomes within hospitals. For missing values, model-based multiple-imputation approach was utilized. Sensitivity analysis via bootstrap validation based upon nested clusters within hospitals using 1000 replacement samples affirmed the validity of the derived estimates.

Results: Significant racial disparities for various outcomes for individual tumor types in pediatric patients was noted. Quantification of differences from regression analysis are depicted in corresponding forest plots for individual outcomes.

Conclusions: Our results quantify estimates across diverse racial population with four subtypes of brain tumors in pediatric age groups. The data provides supporting evidence for policy makers to formulate tailored measures to mitigate these differences.

Patient Care: The findings has direct implications for policy makers to formulate measures to attenuate the racial disparity to healthcare access. For physicians, it provides a baseline overview for expectations in outcomes stratified, albeit should be tailored based upon individual outcomes.

Learning Objectives: 1.) To provide a brief overview of population-based cohort data on the magnitude of racial disparity in post-craniotomy outcomes and costs in pediatric patients.

References: Rhee D et al. A novel multispecialty surgical risk score for children.Pediatrics. 2013 Mar;131(3):e829-36.