Table 2.
Association between screening colonoscopy and risk of incident late-stage CRC according to data source, SEARCH Study 2006–2008, n = 1,039
| Data source according to screening definition used | Odds Ratio and 95% CI | % Difference in beta coefficients | P-value of difference* |
|---|---|---|---|
|
Screening defined as ‘probable’ or ‘definite’ |
|
|
|
| Progress note |
0.31 (0.14-0.70) |
−9.5 |
0.64 |
| Referral note |
0.46 (0.22-0.98) |
28.0 |
0.41 |
| Procedure report |
0.50 (0.25-1.02) |
31.2 |
0.26 |
| All sources combined |
0.30 (0.14-0.65) |
−21.6 |
0.52 |
| Adjudicated indication |
0.36 (0.19-0.68) |
Ref |
Ref |
|
Same definition as above plus ‘high-risk’ screening exposures |
|
|
|
| Progress note |
0.32 (0.17-0.64) |
2.4 |
0.98 |
| Referral note |
0.45 (0.23-0.85) |
34.9 |
0.12 |
| Procedure report |
0.43 (0.23-0.79) |
27.4 |
0.23 |
| All sources combined |
0.31 (0.16-0.58) |
−6.2 |
0.69 |
| Adjudicated indication | 0.33 (0.17-0.62) | Ref | Ref |
Conditional logistic regression modelling was performed with separate indicator variables for colonoscopy and sigmoidoscopy and adjusted for receipt of ‘definite’ screening barium enema and FOBT, census block-group poverty levels (as a continuous variable), number of preventive health care visits, family history of colorectal cancer, and comorbidity index at baseline. Missing values of poverty level were imputed using predictive mean matching.
*Two-sided Wald Chi-square (χ2) P-values of the difference between effect sizes.