389 - Predictors of Delayed Diagnosis of Pediatric CNS Tumors in the Emergency Department

Poster Number: 389
Publication Number: 389.314

Ann L. Young, Boston Children's Hospital, Boston, MA, United States; Michael C. Monuteaux, Boston Children's Hospital, Boston, MA, United States; kenneth michelson, Boston Children's Hospital, Boston, MA, United States

Background: The diagnosis of central nervous system (CNS) tumors in children is frequently delayed, leading to adverse outcomes.

Objective: Delays often occur in emergency departments (EDs), so examination of factors associated with delayed ED diagnosis could identify approaches to reduce delays.

Design/Methods: This case-control study used the Healthcare Cost and Utilization Project State ED and Inpatient Databases from 2014-2017 for five states (NY, MD, FL, WI, IA). We included children aged 6 months-17 years if their first diagnosis of CNS tumor was made in the ED, excluding patients with prior malignancies and tumor predispositions. Cases had a delayed diagnosis, defined as having one or more ED visits in the 140 days preceding tumor diagnosis (the mean pre-diagnostic symptomatic interval for pediatric CNS tumors in the United States). Controls had no ED visits in the preceding 140 days. We constructed a logistic regression model to examine the association of patient- and ED-level predictors with delayed diagnosis, defined by case-control status.

Results: 2707 children (2053 controls, 76%; 654 cases, 24%) were included. The case rate did not significantly change over time. Among cases, 68% had one preceding ED visit and 32% had 2+ visits. The top five preceding visit primary diagnoses in descending frequency were: “Epilepsy; convulsions,” “Headache; migraine,” “Other nervous system disorders,” “Nausea; vomiting,” and “Other gastrointestinal disorders.” Patient-level predictors of delayed diagnosis included age ≤5 years (adjusted odds ratio [aOR] 1.59 compared with >5 years; 95% confidence interval [CI] 1.16-2.12), public insurance (aOR 1.39 compared with private; 95% CI 1.10-1.77), chronic complex condition (aOR 9.43; 95% CI 6.38-13.92), rural home address (aOR 0.53; 95% CI 0.33-0.83), seeking care within state (aOR 3.98 compared with crossing a state line; 95% CI 2.26-7.00), and weekend presentation (aOR 1.50; 95% CI 1.14-1.96). ED-level predictors of delayed diagnosis included non-teaching hospitals (aOR 2.90; 95% CI 1.90-4.42) and rural location (aOR 4.20; 95% CI 1.25-14.04).Conclusion(s): A large proportion of children with CNS tumors experienced delayed ED diagnosis. One-third of cases required three or more visits before their diagnosis. Preceding visit diagnoses overlapped with CNS tumor symptoms. The strongest patient-level predictors were chronic complex conditions and staying within state to seek care. The strongest ED-level predictors were rural and non-teaching hospitals. Further research is needed to evaluate ED readiness to diagnose pediatric patients with rare but life-threatening conditions.
CVAnn Young_CV.pdf