Table 1 Summary of methods and results after adjustment for potential confounders
Study design | Outcome | Statistical methods | Estimate of effect of statin exposure * | P |
|---|---|---|---|---|
Matched cohort | Use of OS in the observation period | Conditional logistic regression | Odds ratio 1.064 (0.879,1.288) | 0.523 |
| Â | Number of days on OS therapy in the observation period | Zero inflated negative binomial model | a) Odds ratio 1.025 (0.900,1.169) | 0.704 |
| Â | Â | Â | b) Rate ratio 0.879 (0.758,1.021) | 0.091 |
Unmatched cohort ^ | Time to initiation of OS therapy | Cox model | Hazard ratio 0.956 (0.901,1.015) | 0.14 |
| Â | Time to cessation of OS therapy | Cox model | Hazard ratio 0.954 (0.866,1.051) | 0.34 |