We read with great interest the report by Khera et al describing the use of mechanical support devices in percutaneous coronary interventions (PCI) using the Nationwide Inpatient Sample data. The investigators did not find any evidence of clear benefit with the use of percutaneous ventricular assist devices (PVADs), that is Impella and TandemHeart, compared with intraaortic balloon pump (IABP) in patients undergoing PCI in the propensity-matched population or in the subgroups of patients with cardiogenic shock, acute myocardial infarction without cardiogenic shock, and non–acute myocardial infarction/non cardiogenic shock (elective high-risk PCI). In another recent study from Nationwide Inpatient Sample, the use of PVADs was found to be a significant predictor of reduced mortality (odds ratio 0.55, 95% CI 0.36 to −0.83; p = 0.004) which was particularly evident in the subgroup of patients who underwent elective high-risk PCI. The propensity score–matched analysis also showed a significantly lower mortality rate associated with PVADs compared with IABPs in that study. These 2 studies may provide a conflicting message to the reader however one must note the following. The current PCI guidelines give a class IIb recommendation for the use of mechanical support devices (IABP or PVAD) during elective high-risk PCI. This is largely based on the A Prospective, Randomized Clinical Trial of Hemodynamic Support With Impella 2.5 Versus Intra-Aortic Balloon Pump in Patients Undergoing High-Risk Percutaneous Coronary Intervention 2 trial. The previous studies, including the one by Khera et al, have noted a numerically lower but statistically nonsignificant reduction in mortality when PVADs are compared with IABP in patients undergoing elective high-risk PCI. Some minor differences in the methods of the 2 studies such as inclusion of sample from 2008 by Patel et al and differences in the propensity matching algorithm may have lead to these findings. The take-home point, however, from both the studies, is that the only group where PVADs may have superior outcomes compared with IABPs is that of patients undergoing elective high-risk PCI. We agree with Khera et al regarding necessity of more robust data to establish the effectiveness of PVADs in this group.

Another important subgroup of patient, which deserves mention, is that of patients who received both IABP and PVADs during the same hospitalization. This group, previously excluded from all studies, was found to have an in-hospital mortality of 41% by Patel et al which highlights the concern about escalating support from IABP to PVAD in patients undergoing PCI.