Introduction: Primary central nervous system(CNS) tumors arise from brain and spinal cord and have an incidence rate of 28.6 per 100,000 people. Primary CNS tumors are included in the two most commonly utilized cancer registries/databases, National Cancer Database(NCDB) and Surveillance, Epidemiology, and End Results(SEER). However, there are some fundamental differences between the two which have to do with the sampling frame as well as the patient, facility, tumor and treatment characteristics. We analyzed primary CNS tumor cases in United States by using NCDB and SEER; we aim to obtain more clear idea about how closely they represent the population in United States in terms of all primary CNS tumors.
Methods: Cases were identified using the International Classification of Disease for Oncology topographical codes designating all primary brain tumors and primary spinal cord tumors. All variables between two databases were compared using standardized difference.
Results: A total of 588,534 primary brain tumor patients were identified. 416,162 patients in NCDB had a mean age(SD) of 56.7(19.4) and 172,372 patients in SEER had a mean age(SD) of 53.6(19.7). A total of 34,827 primary spinal cord tumor patients were identified. 26,602 patients in NCDB had a mean age(SD) of 52.7(19.1) and 8,225 patients in SEER had a mean age(SD) of 50.3(20.1). There were statistically significant differences with regards to age, Hispanic origin, and income between the two databases for both tumors.
Conclusions: It is seen that, two databases differ significantly from most aspects. NCDB, as a facility-based cancer registry, is better in terms of reflecting facility characteristics and detail characteristics related to treatments(inpatient stay after surgery, readmission, 30 and 90-day mortality) and SEER, as a population-based registry, is better in terms of reflecting demographics. However NCDB has data of more patients and has more coverage across the United States, which is also a strong predictor for representativeness.
Patient Care: Big data research is trending among researchers in neurosurgery field with emerging national databases and registries in US. Although they are sources of data for large number of patients, utilization of these databases should be carried out carefully. Presented study aims to guide researchers to choose the more suitable database according to their research question in Neuro-oncology field. This will allow us to see the results for patients with CNS tumors better.
Learning Objectives: Participants will be expected to:
1)Recognize the type of data contained within NCDB and SEER
2)Carefully select which database is appropriate for their future research questions
3)Distinguish NCDB and SEER’s sampling methods and their respective limitations