Table 3:
Domain | Univariate | Multivariable * | ||
---|---|---|---|---|
Odds Ratio (95% CI) | P-value | Odds Ratio (95% CI) | P-value | |
Awareness | ||||
Q14. Heard about colorectal cancer | 0.65 (0.10–4.08) | 0.649 | 0.97 (0.11–8.30) | 0.980 |
Q15. Heard about colorectal polyp | 23.14 (7.62–70.31) | <0.001 | 22.47 (6.42–78.62) | <0.001 |
Q16. Heard about the FOBT or stool blood test | 180.17 (38.20–849.83) | <0.001 | 245.37 (34.55–1742.75) | <0.001 |
Q19. Heard about colonoscopy | 0.47 (0.11–1.99) | 0.303 | 0.64 (0.12–3.32) | 0.593 |
Behavior | ||||
Q17. Ever been recommended for colon cancer (FOBT) test | 245.44 (47.11–1278.77) | <0.001 | 1473.56 (68.58–31660.02) | <0.001 |
Q18. Ever had a FOBT | 1.00 (0.06–16.44) | 1.000 | 0.23 (0.01–9.56) | 0.437 |
Q20. Ever been recommended for colonoscopy | 68.31 (14.51–321.58) | <0.001 | 91.90 (15.06–560.96) | <0.001 |
Q21. Ever had a colonoscopy | 2.52 (0.72–8.82) | 0.147 | 2.08 (0.51–8.50) | 0.309 |
Q30. Are you willing to undergo colon cancer testing within 6 months? | 84.33 (22.28–319.14) | <0.001 | 87.17 (19.01–399.63) | <0.001 |
Knowledge | ||||
Q22. I do not believe that there is only one screening test for colon cancer.** | 158.39 (32.45–773.18) | <0.001 | 126.63 (23.61–679.07) | <0.001 |
Q23. There is a stool blood test using a “home” test. | 173.73 (21.49–1404.51) | <0.001 | 157.17 (18.02–1370.41) | <0.001 |
Q24. I believe that people are supposed to start getting tested for colon cancer at age of 50. | 1.00 (0.44–2.26) | 1.000 | 0.75 (0.29–1.95) | 0.561 |
Q25. Once people start having stool blood test, they should have them every 3 years **. | 2.57 (1.07–6.15) | 0.034 | 2.64 (0.97–7.21) | 0.057 |
Q26. In general, once people start having colonoscopy exams at age of 50, they should have them every 5 years**. | 0.73 (0.33–1.59) | 0.424 | 0.87 (0.35–2.14) | 0.761 |
Abbreviations: FOBT, fecal occult blood test.
Odds ratio is for the intervention group relative to the control group. 100 observations were used in the multivariable models.
The propensity score (PS) of being in the intervention group was estimated using a multivariable logistic regression model including Q1 through Q11, Q13, and Q28-Q29. Then, the multivariable model was adjusted for the PS.
Models were fitted with a response ‘No’ vs. ‘Yes’ as ‘No’ is the correct answer.