Table 2.
Results for Intervention Effect on Colonoscopy Uptake within 15 months
| Odds of getting Colonoscopy | 95% Confidence Interval | P Value | Medically Verified Colonoscopies | ||
|---|---|---|---|---|---|
| Modela | TeleCARE vs. Education | [Lower, Upper] | TeleCARE % (n) | Education Brochure % (n) | |
| Cases with known outcome | 2.50 | [1.61-3.88] | <0.001 | 44.3% (78/176) | 24.2% (50/207) |
| Negative outcome imputationb | 2.02 | [1.32-3.08] | 0.001 | 33.6% (78/232) | 20.1% (50/249) |
| Multiple imputationc | 2.37 | [1.59-3.52] | <0.001 | 42.7% (99/232) | 24.1% (60/249) |
a
Each of the 3 separate models represents a different treatment of missing outcomes and included a random effect for family.
b
Negative outcome imputation treated unknown colonoscopy outcome as no colonoscopy.
c
Average number of colonoscopies based on 5 imputation sets from the SAS procedure MI