Skip to main content
. Author manuscript; available in PMC: 2016 Sep 1.
Published in final edited form as: Pediatr Blood Cancer. 2015 Mar 21;62(9):1495–1500. doi: 10.1002/pbc.25506

TABLE I.

Characteristics of Administrative Datasets and Registries Used for the Study of Pediatric Cancer

Nationwide Inpatient Sample
(NIS)
Kids Inpatient Database (KID) Pediatric Health Information
System (PHIS)
Surveillance, Epidemiology, and
End Results Database (SEER)
National Cancer Data Base
(NCDB)
Description Publically-available nationally-representative all-payer inpatient care database Publically-available nationally-representative all-payer inpatient care database for children Privately administered database from >40 children’s hospitals Publically-available representative population-based registry of patients with cancer Private cancer registry of patients at hospitals in Commission on Cancer
Data Source State public and private data organizations State public and private data organizations Hospital encounters, including admissions, ambulatory medical and/or short-stay, ED visits Public and private data organizations, cancer and hospital registries Hospital cancer registries
Sample Data Patient demographics Patient demographics Patient demographics Patient demographics Patient demographics
Primary and secondary diagnoses, procedures Primary and secondary diagnoses, procedures Primary and secondary diagnoses, procedures Disease incidence Tumor characteristics
Hospital characteristics Discharge status Date-stamped billing data Tumor characteristics Hospital characteristics
Expected payment source Hospital characteristics Payer information First-course treatment Payer information
Total charges Expected payment source Survival and vital status First-course treatment
Discharge status Total charges Survival
Length of stay Length of stay
Severity and comorbidity measures Severity and comorbidity measures
Limitations Lacks clinical data, cannot track patients across time or settings Lacks clinical data, limited tracking across time or settings Coding errors, lack of clinical detail, limited classification of complications, limited tracking across time or settings Some populations overrepresented, frequency of tumor characteristics and survival limited to the racial/ethnic groups for which population denominators available Not population-based, excludes hospitals not in the Commission on Cancer