Brief background: statins are drugs that lower cholesterol. Trials comparing statins with placebo show that statins lower mortality (in terms of hazard ratio) by almost 20% (hazard ratio 0.8). There are numerous trials conducted with similar findings. Mean follow-up is approx 4 years in these studies.
Now comes this study: http://bmjopen.bmj.com/content/5/9/e007118 - which is based on the statin trials and this study concludes that in terms of survival time, satins prolong life by 5 days.
This cannot be mathematically sound.
Could I go backwards and estimate the hazard ratio that these numbers would yield? Because the hazard ratio should be approimately equal to the studies it was based on. This is a way of prooving that the study results are incorrect.