Table 4.
Logistic regression analysis of association between GI events and treatment initiation
| Effect | Initiated any treatment vs. no treatment initiated | ||
|---|---|---|---|
| Post-index GI event (ref: no post-index GI event) | Odds ratio | 95 % CI | p value |
| Among patients without pre-index GI problems | 0.293 | (0.25, 0.34) | <0.0001 |
| Among patients with pre-index GI problems | 0.257 | (0.20, 0.31) | <0.0001 |
| Age group | |||
| 65–74 vs. 55–64 | 1.319 | (1.14, 1.47) | 0.0024 |
| 75–84 vs. 55–64 | 1.538 | (1.26, 1.72) | <0.0001 |
| 85+ vs. 55–64 | 1.271 | (1.05, 1.63) | 0.1432 |
| Pre-index medication use | |||
| Gastro-protective agents | 1.755 | (1.58, 1.95) | <0.0001 |
| NSAIDs | 1.337 | (1.23, 1.45) | <0.0001 |
| Glucocorticoids | 1.254 | (1.13, 1.39) | <0.0001 |
| Charlson comorbidity index score | 0.929 | (0.89, 0.97) | 0.0093 |
GI gastrointestinal, NSAIDs non-steroidal anti-inflammatory drugs
CI confidence interval