The approach does not disqualify prosthesis–patient mismatch

We have read with interest the article by Knez et al. [1] and we thank these authors for their kind comments with regards to our previous work. However, we cannot agree with their conclusion as reflected in the title of the paper.

We and others [2] have always defined prosthesis-patient mismatch on the basis of the indexed effective orifice area (EOA) whereas Knez et al. [1] have divided their patients between expected mismatch (EXMIS) and no mismatch (NOMIS) on the basis of the indexed geometric orifice area (GOA) which is a measurement derived from the internal diameter of the prosthesis measured in vitro by the manufacturer. Indeed, we have recently shown that the indexed GOA is poorly related to postoperative gradients and that the indexed EOA is a much better predictor of postoperative hemodynamics [3]. Moreover, closer scrutiny of the data of Knez et al. shows that there is no significant difference for postoperative indexed EOA's or gradients between the two groups. It thus follows that with regards to the usual definition of mismatch these two groups offer no discrimination between patients with and without mismatch and that no conclusion with regards to mismatch can therefore be derived from these data.

The overall good results with regards to the regression of left ventricular hypertrophy in both groups can possibly be explained by the relatively good hemodynamic performance of mechanical valves in general and a possibly lesser incidence of mismatch in this population. Nonetheless, to suggest that the patient's BSA should not be taken into consideration when implanting this type of prosthesis is both premature and without scientific justification. Moreover, the prospective strategy we propose to document and prevent mismatch is very easy to perform and takes only a few minutes [2,3]. As an analogy, we would submit that omitting to do a blood lipid profile in a patient with coronary disease may not have immediate consequences but, in light of current knowledge, no one would question its great relevance in establishing therapeutic goals for the patient. We believe the same applies to prosthesis-patient mismatch.