tag:blogger.com,1999:blog-8219518322422347333.post8613849305536150954..comments2022-02-25T02:34:40.626-08:00Comments on Unboxing Evidence Based Medicine: Why Unbox Evidence-Based Medicine?Luis Cláudio Correiahttp://www.blogger.com/profile/02909537501158073052noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-8219518322422347333.post-3591462921454885982018-08-30T20:36:19.575-07:002018-08-30T20:36:19.575-07:00Part II
How can we explain such large effect size...Part II<br /><br />How can we explain such large effect size with rather small differences between groups in terms of preventive medications?<br /><br />As you suggested, the tables suggest a rather small percentage difference in the rates of medications among groupos. The problem here is that by averaging the number of patients on certain therapies, we neglect the fact that a substantial number of patients were taken off preventive measures as result of normal CTA results. Table S3 shows that 14% of patients in the CT guided arm received new preventive treatment compared to only 4% in control. At the same time, 4% of the CT arm patients were taken off preventive therapy vs. essentially zero in control. Therefore, follow up averages underestimate the difference in new preventive treatment folks received in the CT guided arm. Furthermore, the data does not reveal if the intensity of prevention/medication dosing changed as result of preventive efforts. It is conceivable that providers increased statin doses in patients already on statins whose CT revealed atherosclerotic disease. <br /><br />It is also important to recognize that the results are not due to one intervention, such as statin therapy, but of a whole array of "preventive measures" --some even not accounted for. In addition to statins—which have been proven to reduce MI and CV death in similar populations--patients were also started on antiplatelet therapy—which is also associated with reductions in MI and CV death. New interventions are typically tested in addition to standard of care and the added benefit is often small. Here, we have the curious situation that some patients did not get any directed prevention (in patients with normal stress test results) and others got a whole array of preventive measures (patients with coronary atherosclerotic disease by CT). It appears the cumulative effect is quite impressive.<br /><br />3) Are the results inflated by ascertainment bias?<br /><br />I agree what there is potential bias since no event adjudication occurred. However, I find it unlikely that there would be a substantial bias based on treatment assignment in these randomized groups. The fact that PROMISE and the European registry showed similar results with shorter follow up and lower risk populations suggests to me that while SCOT-HEART results may be on the upper range of the effect size, study results remain within plausibility. <br /><br /><br />4) What is the external validity of SCOT-HEART?<br /><br />As outlined above, I do not believe that protocolizing preventive measures in response to evidence of coronary atherosclerotic disease reduces external validity of the trial. On the contrary, I believe this measure is not only reasonable but very appropriate and applicable. In other words, the same prescription should be given in clinical practice if similar results are to be expected. <br /><br /><br />5) Conclusions<br /><br />I believe the fact that two RCT and one large registry (encompassing together >100,000 patients in different countries and healthcare systems) revealed a reduction of MI with a CT guided approach vs. stress testing is of major significance and should result in immediate change to our practice (as already done in the UK where they use CTA as first line test in patients with stable CP). We use approximately 10,000,000 stress tests in the US each year. Even if the effect size in clinical practice is smaller than in SCOT-HEART, e.g., a 30% risk reduction for MI when combining SCOT-HEART and PROMISE, we may prevent many thousands of myocardial infarctions each year by identifying symptomatic patients who have coronary atherosclerotic disease—not detectable by stress testing—who benefit from prevention. As such, I believe the SCOT-HEART trial ranks among the most important clinical trials in cardiology of the past years. <br /><br />Kind regards,<br />Armin<br /><br /><br />My disclosure: I am directing a Cardiac CT Lab. My main clinical and investigational focus, however, is CHD and I have been interested in CT solely because I believe it is currently the best tool to identify patients with CHD and guide their management. <br /><br /><br />Anonymoushttps://www.blogger.com/profile/03544038327962819327noreply@blogger.comtag:blogger.com,1999:blog-8219518322422347333.post-8857424459427039352018-08-30T20:31:47.394-07:002018-08-30T20:31:47.394-07:00This comment has been removed by a blog administrator.Anonymoushttps://www.blogger.com/profile/03544038327962819327noreply@blogger.com