Table 3.
Models | OR (95% CI) | p Value |
---|---|---|
Unadjusted data | 0.70 (0.55–0.89) | < 0.01 |
Third principal model* | 0.70 (0.55–0.90) | < 0.01 |
Sensitivity analyses | ||
1. Proportional odds model (angina as 3-level categorical variable)† | 0.69 (0.55–0.87) | < 0.01 |
2. Third principal model + adjusted for ntiangina medication post-randomization | 0.78 (0.60–1.00) | 0.05 |
3. Worst-case scenario | 0.77 (0.61–0.98) | 0.03 |
Of the 1,212 patients randomized in STICH, 1,089 with angina assessed at least once after randomization were included in the statistical model
The model is the final adjusted binary logistic regression model. The model remained stable after internal validation (OR: 0.69; 95% CI: 0.54 to 0.89), and after crossover patients were taken into consideration (OR: 0.57; 95% CI: 0.45 to 0.74; p < 0.01).
Proportional odds model defines angina relief as an ordinal 3-level categorical variable with values representing worsening angina, stable angina, and angina relief. The odds ratio supplied is predicting the probability of angina relief averaging over worsening angina and stable angina.