Table 2.

Characteristics of data sources used to measure patterns of colorectal cancer care in the United States*

Percentage of studies	Data source	Data type	Population coverage	Patient data		Health services data
				Date of diagnosis	Stage at diagnosis	Surgery	Radiotherapy	Chemotherapy	Post-diagnosis surveillance	End of life
40%	SEER–Medicare	Registry linked to Medicare claims	≥65 y diagnosed and treated in SEER regions	√	√	√	√	√	√	√
10%	SEER	Registry	17 US regions	√	√	√	√
10%	State registry linked†	Registry linked to supplemental data source	State	√	√	√	√	√	√	√
8%	State registry alone†	Registry	State	√	√	√	√	√	√	√
7%	HCUP–NIS	Hospital discharge data	Nation			√
7%	Special study (eg, CanCORS)† ‡	Varies	Varies	√	√	√	√	√	√	√
5%	Single institution†	Hospital medical records	Institution	√	√	√	√	√	√	√
5%	Medical claims alone	Medical claims	Medical insurance beneficiaries			√	√	√	√	√
3%	NCI Patterns of Care	Registry linked to physician surveys	17 US regions	√	√	√	√	√
3%	NCDB		Nation	√	√	√	√	√
3%	NCCN CRC outcomes database	Chart review	Patients treated at 8 NCI comprehensive cancer centers	√	√	√	√	√

* CanCORS = Cancer Care Outcome Research and Surveillance Consortium; CRC = colorectal cancer; HCUP–NIS = Healthcare Cost and Utilization Project–Nationwide Inpatient Sample; NCCN = National Comprehensive Cancer Network; NCDB = National Cancer Data Base; NCI = National Cancer Institute; SEER = Surveillance, Epidemiology, and End Results.

† Availability of health services data may vary by state, study, or institution.

‡ Availability of patient data may vary by study.