Table 2.
Logistic Regression (LR) Models of 6 Month (T3) Colorectal Cancer Outcomes
| Outcomes & Randomized Groups | Best-Estimate Data (Medical Record and Self-Report) | Self-Report Data (Stage) | ||||
|---|---|---|---|---|---|---|
| Model 1 (Binary LR Model) T3 Screen (yes/no) | Model 2 (Generalized LR Model; reference category = T3 Precontemplation) | |||||
| T3 Contemplation | T3 Action | |||||
| Adjusted OR | p-value | Adjusted OR | p-value | Adjusted OR | p-value | |
| Any Colorectal Cancer screening test | ||||||
| Web only | 1.01 (0.63, 1.62) | .9626 | 1.97 (1.14, 3.40) | .0152 | 1.81 (0.99, 3.30) | .0537 |
| Phone only | 4.00 (2.60, 6.16) | <.0001 | 2.37 (1.28, 4.39) | .0058 | 7.94 (4.33, 14.56) | <.0001 |
| Web + Phone | 2.69 (1.73, 4.18) | <.0001 | 2.52 (1.33, 4.77) | .0045 | 6.68 (3.56, 12.54) | <.0001 |
| Risk-appropriate colorectal cancer test | ||||||
| Web only | 0.94 (0.59, 1.51) | .7982 | 1.92 (1.11, 3.32) | .0195 | 1.62 (0.89, 2.96) | .1163 |
| Phone only | 4.00 (2.60, 6.17) | <.0001 | 1.99 (1.08, 3.67) | .0277 | 7.07 (3.89, 12.85) | <.0001 |
| Web + Phone | 2.59 (1.67, 4.03) | <.0001 | 2.37 (1.26, 4.48) | .0076 | 6.08 (3.26, 11.34) | <.0001 |
| Stool test | ||||||
| Web only | 1.20 (0.64, 2.24) | .5772 | 2.07 (1.17, 3.67) | .0126 | 1.78 (0.86, 3.68) | .1208 |
| Phone only | 6.80 (3.98, 11.60) | <.0001 | 1.93 (1.04, 3.57) | .0364 | 9.81 (5.21, 18.48) | <.0001 |
| Web + Phone | 5.37 (3.11, 9.29) | <.0001 | 2.86 (1.50, 5.45) | .0014 | 12.14 (6.26, 23.57) | <.0001 |
| Colonoscopy | ||||||
| Web only | 0.70 (0.37, 1.33) | .2794 | 1.02 (0.57, 1.80) | .9516 | 0.96 (0.46, 1.97) | .9005 |
| Phone only | 1.39 (0.77, 2.52) | .2717 | 0.64 (0.35, 1.14) | .1285 | 1.38 (0.70, 2.70) | .3534 |
| Web + Phone | 0.88 (0.48, 1.61) | .6750 | 0.83 (0.46, 1.49) | .5275 | 0.84 (0.41, 1.72) | .6274 |
Models adjusted for baseline characteristics including age, race (African American vs Other), education, income, marital status, BMI, whether depression limits patient’s activities (yes/no), family history of 1 or more blood relatives with colon cancer (yes/no), perceived risk, doctor recommendation (yes/no), number of past-year primary care visits excluding eye care and dentistry (>=3), number of self- reported health problems, baseline adherence to mammography screening (yes/no), baseline stage of change for colorectal cancer screening, knowledge, susceptibility, benefits, fear, fatalism, self-efficacy, and barriers. Self-efficacy and barriers specific for colonoscopy or stool test were used in colonoscopy and stool test models, respectively. Sample sizes for Models 1 and 2, respectively, were: any colorectal cancer (843, 683), risk appropriate colorectal cancer (842, 681), stool test (836, 642), and colonoscopy (835, 643).