Dear Dr. Topol,

It was a little surprising that this trial did not receive the kind of coverage that it deserved after publication in the Lancet. As a physician who routine prescribed perioperative beta-blockers at the in a formalized perioperative clinic, it is clear that the POISE design of "one size fits all" is a serious issue. The large simple clinical trial design was certainly not ideal for this question. As you are well aware, no one uses 400 mg of metoprolol a day in the perioperative setting, especially in beta-blocker naive elderly patients with significant carotid disease and a prior h/o stroke and significant PAD. The titration and holding criteria for beta-blockers were especially liberal and not appropriate for high-risk patients. Including emergency surgery patients was also problematic because these patients are especially prone to post-operative complications, especially infections. Beta-blockers at very high doses can obviously mask the signs of sepsis and blunt responsiveness to pressors--there is no mention of the types of pressors used post-operatively in the study. In addition, it is hard to keep a trial of high-dose beta-blockers truly blinded because of their obvious hemodynamic effects. The accompanying editorial by Poldermans and Fleischer addresses some of these issues but there are many other issues that remain. Data from Iran and Colombia was discarded--not surprising, because perioperative beta-blockade is a very resource intensive strategy and perhaps current best suited to environments where intense monitoring and response to complications is quick and regimented. This also may have been an issue in countries where peri-operative beta-blockade is a novel phenomenon. As you are also aware, Poldermans's group from the Netherlands has published extensively on this subject in the past and never demonstrated a signal for harm with these agents. They have used these agents in accordance with the RCRI risk index and used dobutamine echo with careful attention to the numbers of dyskinetic segments and careful preoperative dose titration over weeks, with very encouraging data. It is true that their numbers are much smaller but big numbers can be deceptive at times. We can go on and on about the metodological flaws with POISE. The bottomline is that the guidelines will be changed because of this study, which is not applicable to the population it was meant to study. People will be denied the benefits of a good treatment strategy because of the fear of litigation--all post-operative hypotension and bradycardia will be mislabeled as beta-blocker related, even though there are many other causes for these complications. Your thoughts?

 Thanks