We read with great interest, Kim et al’s study regarding the association of hospital volume with in-hospital outcomes after transcatheter aortic valve implantation (TAVI) in the United States. We would like to bring to the investigators’ attention the results from our study using the Nationwide Inpatient Sample from 2012 published earlier this year. We also studied the impact of hospital volume divided into quartiles on in-hospital outcomes after TAVI. Our results demonstrated the highest hospital volume quartile to be significantly predictive of lower in-hospital mortality (odds ratio 0.38, 95% CI 0.27 to 0.54, p <0.001) and procedural complications after multivariate analysis. In addition, we also demonstrated that higher hospital volume was predictive of shorter length of stay and lower hospitalization costs. These results were further reiterated by a spline analysis, which demonstrated decreasing predicted probability of mortality with increasing hospital volume. A similar inverse relationship between increasing hospital volume and reduced hospitalization costs was noted on spline analysis.

The investigators in their present study also showed that hospitals with a high transapical TAVI volume had lower mortality after surgical aortic valve replacement (SAVR). This reiterates the results of another study, where we demonstrated lower in-hospital mortality and complications after SAVR in TAVI capable centers. Indeed, TAVI centers have developed special multidisciplinary heart valve teams for optimizing patient selection, planning procedural details and potential complication management. These extensive heart valve programs can be posited to contribute toward improving overall procedural outcomes even in SAVR.

We would also like to point out that the current International Classification of Diseases code 35.05 for TAVI used by the investigators is for endovascular TAVI. This code also includes the subclavian and aortic approach and not just transfemoral approach. The invesigators have primarily used weighted values to represent their final results. Contrarily, we mentioned both the actual and weighted values in our study. Although weighted values from Nationwide Inpatient Sample are believed to provide national estimates, they might not represent the true procedural numbers as have been noted by several of our previous reviewers. Moreover, it is not entirely clear from the study if the investigators excluded all patients with primary diagnosis of stroke, myocardial infarction and so forth to prevent false classification of a preexisting condition. Nonetheless, the results from the Kim et al study further reiterate the fact that a higher hospital volume is an important quality metric in predicting superior procedural outcomes in various complex structural interventions including TAVI.