Colon cancer care and survival: income and insurance are more predictive in the USA, community primary care physician supply more so in Canada

Table 3 Logistic regression of main effects and interaction of community primary care physician density and country on 10-year survival of non-metastasized colon cancer: California and Ontario, 1996—2010

Predictor variables	Odds ratio	95 % Confidence interval
Model 1: community primary care physician density and country
Country	0.94	0.82, 1.09
Community primary care physician (PCP) density	1.60	1.30, 1.96
Community PCP density by country	1.95	1.50, 2.52
Model 2: colon cancer care variables entered
Had surgical resection	3.12	1.64, 5.93
>15 regional lymph nodes examined	1.23	1.07, 1.41
Received chemotherapy	1.28	1.00, 1.63
Waited > 90 days for treatments^a	0.69^b	0.45, 1.07
Country	0.98	0.88, 1.11
Community primary care physician (PCP) density	1.30	1.06, 1.60
Community PCP density by country	1.11	0.82, 1.50

Notes. Total samples were 6,043 for model 1 and 5,908 for model 2. All effects were adjusted for age, tumor stage and grade, and place (small or large urban or rural). Gender did not enter either model so this pattern is likely the same for women and men. Statistically significant odds ratios are bolded
^aTotal wait time = wait after diagnosis for surgery + wait after surgery for chemotherapy
^bApproached significance, 90 % confidence interval (CI) = 0.48, 1.00

ISSN: 1475-9276