Randomized trials in cardiogenic shock: what's the problem?

The ingredients for a promising randomized study might be expected to include a common condition, a high event rate and existing evidence to suggest at least measurable benefit from the tested intervention. These criteria would seem to apply to cardiogenic shock complicating acute myocardial infarction; even in the thrombolytic era the incidence is almost 10% of hospital admissions with myocardial infarction and mortality hovers persistently about the 90% mark. Pooled results from historical reviews examining the effects of early revascularization in these patients have given cause for optimism, with mortality reduced following CABG or PTCA to 33% and 42%, respectively. Why then do we read in this issue of a prospective, randomized study that was terminated prematurely in the absence of a result? The study aims and trial design were soundly based. Patients in shock for at least 30 min, and within 48 h of infarct symptoms, were allocated to either immediate investigation with a view to emergency revascularization, or medical therapy. With historical data suggesting an 80% mortality in the conservative group and an anticipated mortality of 50% with urgent revascularization, the usually accepted statistical criteria dictated 57 patients to be randomized into each group. Inclusion and exclusion criteria were by no means stringent and reflected practical issues often encountered in managing such critically ill patients. Enrolment of patients in shock despite inotropic support may, however, have selected a group less likely to benefit from intervention and a 30-day mortality of 69% in the actively managed group would support this view. The definition of shock has varied in previous reviews which may have resulted in disparate outcomes and thereby apparently encouraging data to support active management. The result of an intervention in a patient with hypotension alone is likely to be better than when associated with signs of poor peripheral perfusion and markedly deranged