The accuracy of chemotherapy ascertainment among colorectal cancer patients in the surveillance, epidemiology, and end results registry program

Table 3 Multivariable logistic regression models of chemotherapy under-ascertainment

Characteristic	Odds Ratio	95% Confidence Interval	P
Age, years			0.08
< 50	Ref	Ref
50–64	2.14	0.96–4.79
65–74	2.12	0.84–5.37
75+	3.12	1.27–7.70
Race			0.28
White	Ref	Ref
Black	1.51	0.91–2.51
Hispanic	1.38	0.16–11.56
Socioeconomic Status	1.01	0.78–1.31	0.93
Insurance			0.54
Other	Ref	Ref
None	0.47	0.13–1.67
Medicaid	1.50	0.52–4.35
Medicare	1.07	0.59–1.96
Geographic Site			0.14
Detroit	Ref	Ref
Georgia	0.71	0.46–1.12
Hospital Bed Size			0.75
< 100	Ref	Ref
300–499	0.86	0.49–1.51
≥500	1.06	0.64–1.76

We present here the covariates of most interest; the multivariable regression model was adjusted for all covariates (age, sex, marital status, comorbid conditions, race, composite socioeconomic status, insurance, primary disease site, geographic site, hospital bed size, and American College of Surgeons cancer program status). Covariates not shown in the table were not significantly associated with under-ascertainment of chemotherapy receipt

ISSN: 1471-2407