Table 3.
Adjusted predictors for IMRT diffusiona |
|||
---|---|---|---|
Characteristics | OR | 95% CI | P |
Race | |||
CA | 1.00 | Reference | — |
AA (urban only) | 0.20 | [0.09, 0.42] | <.0001 |
Age at diagnosis, (years) | |||
66-69 | 1.00 | Reference | — |
70-74 | 0.99 | [0.87, 1.13] | .921 |
≥75 | 0.93 | [0.82, 1.05] | .0.252 |
Year of radiation | |||
2002 | 1.00 | Reference | — |
2003 | 2.93 | [2.53, 3.39] | <.0001 |
2004 | 7.69 | [6.62, 8.94] | <.0001 |
2005 | 22.40 | [19.03, 26.37] | <.0001 |
2006 | 52.00 | [43.31, 62.43] | <.0001 |
Marital status | |||
Married | 1.00 | Reference | — |
Not married/unknown | 1.00 | [0.95, 1.05] | .0987 |
NCI combined comorbidity score | |||
0 | 1.00 | Reference | — |
>0 | 0.82 | [0.74, 0.90] | <.0001 |
Income (census tract) | |||
Quartile 1 (0%-25%) | 1.00 | Reference | — |
Quartile 2 (26%-50%) | 1.04 | [0.90, 1.21] | .575 |
Quartile 3 (51%-75%) | 1.18 | [1.00, 1.38] | .0.0496 |
Quartile 4 (76%-100%) | 1.21 | [0.99, 1.47] | .061 |
Education (census tract) | |||
Quartile 1 (0-%25%) | 1.00 | Reference | — |
Quartile 2 (26%-50%) | 1.30 | [1.12, 1.50] | .0006 |
Quartile 3 (51%-75%) | 1.69 | [1.43, 2.00] | <.0001 |
Quartile 4 (76%-100%) | 2.28 | [1.87, 2.77] | <.0001 |
Population density | |||
Urban | 1.00 | Reference | — |
Rural | 0.17 | [0.08, 0.37] | <.0001 |
Tumor grade | |||
I | 1.00 | Reference | — |
II | 0.95 | [0.87, 1.05] | .374 |
III/IV | 0.81 | [0.63, 1.04] | .099 |
Clinical stage | |||
T1/T2 | 1.00 | Reference | — |
T3/T4 | 0.91 | [0.87-0.96] | .0003 |
Geographic region | |||
South | 1.00 | Reference | — |
Northeast | 2.38 | [2.05, 2.76] | <.0001 |
Central | 0.90 | [0.77, 1.04] | .159 |
West | 1.77 | [1.53, 2.04] | <.0001 |
Interaction terms | |||
Race (AA) * Urban | 4.39 | 2.17-8.89 | <.0001 |
Note. NCI = National Cancer Institute; SEER = Surveillance, Epidemiology, and End Results; 95% CI = confidence interval; OR = odds radio; AA = African American; CA = Caucasian; IMRT = intensity-modulated radiation therapy. SEER sites were grouped into four geographic regions for analysis.
Logistic regression adjusted for race, age (3 categories), year of radiation, marital status, Charlson comorbidity index, income, education, population density, tumor grade, clinical stage, geographic region, and race (AA) * urban.