Association of Physician Certification in Interventional Cardiology With In-Hospital Outcomes of Percutaneous Coronary Intervention

Paul N. Fiorilli, MD; Karl E. Minges, MPH; Jeph Herrin, PhD; John C. Messenger, MD; Henry H. Ting, MD; Brahmajee K. Nallamothu, MD; Rebecca S. Lipner, PhD; Brian J. Hess, PhD; Eric S. Holmboe, MD; Joseph J. Brennan, MD; Jeptha P. Curtis, MD

Circulation. 2015;132:1816-1824. DOI: 10.1161/CIRCULATIONAHA.115.017523.


Dr Teirstein Comments

The authors include highly compensated employees of the ABIM. This is a comparison of patient outcomes following interventional cardiology procedures stratified according to certification status of performing physician. Outcomes in patients receiving PCI by board certified physicians (i.e., participated in a training program and passed the initial boards) versus non-certified physicians (ie practice pathway, never certified) were compared as well as a group of physicians who were once certified but let their certification lapse (i.e. were certified but did not do MOC). Outcomes were better in the board certified physician group (i.e. physicians who did a fellowship and passed the initial certification exam) but no different in the group who let their certification lapse (non MOC) after ten years. Thus, this paper supports initial certification, but not recertification, i.e. MOC.

Dr Cohen Comments

This is a fairly straightforward observational study that used data from the ACC-NCDR database to examine the association between board certification in interventional cardiology (either initial certification or possibly recertification) and in-hospital PCI outcomes. The analysis was adjusted for patient characteristics, physician experience and volume, and hospital characteristics. The main results of the study were that board certification in interventional cardiology was associated with a small but statistically significant reduction in in- hospital mortality and need for emergency bypass surgery but no differences in other outcomes including bleeding, vascular complications, or a composite adverse outcome indicator. There were also no differences in procedural appropriateness. In a secondary analysis in which the group of non-board certified practitioners were separated according to whether they were never certified or were initially certified but lapsed (because they did not complete maintenance of certification), the excess risk associated with lack of board certification appeared to be confined to the group who were never board certified and was not seen in those practitioners who were originally board certified but allowed their certification to lapse.

Overall, this study does suggest that there may be some value to initial board certification with respect to patient outcomes, but the effect is quite weak and may not be clinically important given the low rates of the complications that were affected. This is probably the main limitation of the study in that PCI outcomes are so favorable in the current era that detecting small to modest differences across physician groups may be very challenging. The fact that the benefit of initial board certification appeared to be similar whether or not one completed recertification and MOC suggests that the real benefit is in the initial certification process—not necessarily MOC. However, it should be noted that the subgroup of lapsed practitioners only accounted for 5% of the PCIs performed during the study period and, as such, the comparisons of this group vs. the group with both initial certification and maintenance of certification are relatively underpowered (as reflected by the wide confidence intervals for the adjusted odds ratios). This limitation mainly affects the comparison of emergency CABG rates and is less of an issue for in-hospital death where the upper bound of the confidence limit for the adjusted odds ratio only extends to 1.06.

Dr Ajay Kirtane Comments

This manuscript was a retrospective analysis of data from a large national registry of PCI procedures. The analysis assessed in-hospital outcomes of procedures performed by interventional cardiologists within the registry. The crude outcomes (across endpoints of in- hospital mortality, bleeding, vascular complications, emergency CABG, and the composite endpoint) were indistinguishable among analysis groups. In adjusted analyses, the odds of in- hospital death and emergent CABG were higher among non-certified physicians. Interestingly, when the analyses were further stratified, it appeared that the absence of initial certification was associated with the increased odds of in-hospital mortality or emergent CABG; the concept of further recertification and/or MOC was not directly assessed. Notably, the rates of most complications were among the lowest in physicians with lapsed certification, and the group of physicians who were initially certified but trained in Cardiovascular Disease prior to 1999 (e.g. the longest time from initial training to outcomes assessment in this study) was used as the referent group, indirectly suggesting no improvements in outcomes with recertification/MOC.