Do hospitals with low mortality rates in coronary artery bypass also perform well in valve replacement?

While hospital performance in coronary artery bypass graft (CABG) surgery is reported widely, patients may find it difficult to learn about their hospital's performance in heart valve replacement. We sought to determine if a hospital's performance in CABG is correlated to its performance in heart valve replacement.

We studied operative mortality after CABG, aortic valve replacement (AVR), and mitral valve replacement (MVR) using the 1994 to 1999 national Medicare database. After excluding any hospital that did not perform at least 50 CABGs and 20 valve replacements per year we examined the correlation between hospital mortality in CABG and hospital mortality in AVR and MVR using least-squares simple linear regression models. Operative mortality was adjusted for patient characteristics using logistic regression models.

A total of 684 hospitals performed 817,606 isolated CABGs, 142,488 AVRs (54% with concomitant CABG), and 61,252 MVRs (45% with concomitant CABG). Hospital mortality rates with AVR ranged from 6.0% to 13.0% between hospitals in the lowest and highest, respectively, 10th percentile of CABG performance. Similarly hospital mortality rates with MVR ranged from 10.1% to 20.5% in the lowest and highest respectively, 10th percentile of CABG performance. Adjusted mortality rates for both AVR and MVR were closely correlated with isolated CABG mortality rates (correlation coefficients 0.592 and 0.538, respectively; p = 0.001 for both correlations). In stratified analyses these correlations persisted regardless of whether valve replacement was performed with or without concomitant CABG or whether valve replacement was performed in a high- or low-volume hospital.

Hospital mortality rates with CABG are closely correlated with mortality rates with valve replacement. These findings suggest that shared processes and systems of care are important determinants of performance in cardiac surgery.