Table 3. Logistic Regression Analysis Modeling the Odds of Evaluation by a Surgeon^*.

Factor (reference group)	Group	Odds Ratio	95% CI
Race/Ethnicity (white)	Black	0.58	0.43 – 0.78
Age (continuous)	Per increasing year of age	0.92	0.90 – 0.93
Charlson score (score = 0)	Score = 1	0.75	0.62 – 0.92
	Score = 2	0.85	0.63 – 1.14
	Score = 3	0.54	0.38 – 0.76
Site of tumor (head)	Body/tail	0.89	0.67 – 1.18
	Not specified	0.53	0.43 – 0.66
Abdominal imaging (no)	Imaging	3.37	2.38 – 4.76
Resectable (not resectable)	Resectable	1.26	1.01 – 1.56
Oncology evaluation (no)	Seen by oncology	1.37	1.14 – 1.63
GI evaluation (no)	Seen by GI	0.71	0.57 – 0.87
Primary care physician (no)	Have a PCP	1.37	1.15 – 1.63
SEER Region**	Individual region not shown		Type 3 P-value <0.0001

^*Year of diagnosis, gender, income, education, marital status, and population are controlled for in the above model but did not influence surgical evaluation so OR are not shown.

^**Individual SEER region (Connecticut, Detroit, Greater California, Hawaii, Iowa, Kentucky, Los Angeles, Louisiana, Metropolitan Atlanta, New Jersey, New Mexico, Rural Georgia, San Francisco-Oakland, San Jose-Monterey, Seattle-Puget Sound, and Utah) odds ratios not shown. OR reported above control for SEER region and SEER region influences surgical evaluation, evidenced by the significant type 3 P-value.

Complete data available in 3,614 of 3,777 patients.