Debunking the July Effect in Cardiac Surgery: A National Analysis of More Than 470,000 Procedures.

BACKGROUND

Recent studies in noncardiac surgery have described worse outcomes in the first month of training. However, the "July effect" in the context of cardiac surgery outcomes is not well understood. We examined whether patient outcomes after cardiac surgery were affected by procedure month or academic year quartile.

METHODS

Using the National Inpatient Sample, we isolated all coronary artery bypass grafting (CABG), surgical aortic valve replacement (AVR), mitral valve repair or replacement (MV), and isolated thoracic aortic aneurysm (TAA) replacement procedures between 2012 and 2014. For each procedure, overall trends in in-hospital mortality and hospital complications were compared by academic year quartiles (ie, between the first academic year quartile vs the fourth quartile) and by procedure month. Outcomes between teaching and nonteaching hospitals were also compared.

RESULTS

Overall, 301,105 CABG, 111,260 AVR, 54,985 MV, and 2,655 TAA procedures met inclusion criteria. In-hospital mortality for each procedure did not vary by procedure month or academic year quartile, even after risk adjustment (all P > .05). Teaching status did not influence risk-adjusted mortality for CABG and isolated TAA replacement (both P > .05). However, teaching hospitals had significantly lower adjusted mortality than nonteaching hospitals for AVR and MV surgery (both P < .01).

CONCLUSIONS

The July effect is not evident for cardiac surgery despite preexisting notions. Teaching hospitals performed at least equivalent, if not better, for major cardiac surgery procedures. These findings highlight the pivotal role of hospital support systems to ensure the safe transition of resident classes without compromising on patient outcomes.