In reply.

coronary syndromes (VIP trial) and high-risk PTCA (COURT trial), the latter not quoted by the authors. Both studies showed that there was no difference in terms of major cardiac events (VIP trial) or even a 45% reduction with a non-ionic contrast medium (iodixanol) when compared to an ionic medium (ioxaglate) in the COURT trial. These discrepancies could result from some methodological flaws noted in the study of Scheller et al. Actually, in the Scheller study, the randomization process was not really adequately performed. It mentioned that one angio-lab was working with one contrast medium when the other lab used the alternative agent. Even if both angio-labs have the same equipment, we cannot be sure that one is not older than the other, that the more seriously ill patients were treated in the new, better laboratory, or that the most skilled, experienced operators were working in only one of the two angiolabs. Thus, we cannot trust this ‘pseudo-randomization’. As a result, there was an important heterogeneity in the two arms of the study population, and more seriously ill patients (ACS+CCS Class IV patients) were in the non-ionic arm. In addition, it is stated that diagnostic coronary angiography was performed with non-ionic contrast media. However, in the interventional procedure, the patients received either non-ionic or ioxaglate, depending on the daily availability in the allocated room. Thus, taking into account the current practice of ad hoc percutaneous coronary intervention, this means that half the patients initially received ionic contrast medium followed by non-ionic contrast medium . . . How can we seriously compare the results in those conditions? Furthermore, the ionic contrast medium was compared to six different non-ionic agents: the respective administration of these agents ranged from 5 to 39% respectively. Finally, the authors mentioned that in contrast to the VIP study, iodixanol had the highest rate of subacute stent occlusion (5%), but only 5% (i.e. 90 patients) received this agent. This means that only four of them had subacute occlusion and, again, is it possible to draw any conclusions from these limited data? There are, as listed, several flaws in the concept and design of Scheller et al.’s paper. This study cannot challenge properly conducted, randomized studies, such as the VIP and COURT trials.