Table 3. Significant predictors of failure to return the FIT kit (13.3%).
Univariate, Unadjusted | Multivariable, Adjusted | |||||
---|---|---|---|---|---|---|
| ||||||
Predictor | OR | 95% CI | p-value | OR | 95% CI | p-value |
| ||||||
Greater residential distance from cancer center (miles) (vs. less residential distance from cancer center) | 1.03 | 1.01-1.05 | .006 | 1.03 | 1.01-1.05 | .015 |
| ||||||
Lower health literacy (vs. higher) | 1.18 | 1.06-1.33 | .003 | |||
| ||||||
Lower awareness score (vs. higher) | 1.19 | 1.04-1.37 | .014 | |||
| ||||||
Higher religious beliefs score (vs. lower) | 1.08 | 1.02-1.15 | .012 | 1.09 | 1.02-1.16 | .015 |
| ||||||
No prior CRC screening (vs. prior CRC screening) | 2.33 | 1.00-5.44 | .050 | |||
| ||||||
Employment status | ||||||
Unable to work | 1.00 | .003 | 1.00 | .010 | ||
Employed | 0.38 | 0.18-0.80 | 0.49 | 0.23-1.08 | ||
Not employed | 0.14 | 0.05-0.45 | 0.15 | 0.05-0.46 | ||
Retired | 0.36 | 0.13-1.00 | 0.45 | 0.16-1.26 |
Note: OR = odds ratio; CI = confidence interval; CRC = colorectal cancer.
Health literacy, awareness, religious beliefs, and residential distance from cancer center in miles were each continuous variables, whereas prior colorectal cancer screening and employment status were categorical variables.
Significant univariate predictors were further assessed using multivariable logistic regression.
Adjusted models included the following variables as potential confounds: gender, racial heritage, education, employment status, annual physical exam, prior CRC screening, residential distance from cancer center, health literacy, and salience and coherence.
Non-significant univariate analyses results not shown.