High surgeon, center volume for LVAD implantation may lower in-hospital mortality risk

New findings from two studies suggest that increased volume of left ventricular assist device implantation procedures confers a reduced risk for in-hospital mortality.

Results from one study found that patients implanted in centers with high annual LVAD volume have lower rates of in-hospital mortality and a shorter length of stay. Data from another study found that high surgeon volume was a predictor of reduced in-hospital mortality, but high center volume was not. Both studies were published in the Journal of Cardiac Failure.

Center volume

Neeraj Shah, MD, MPH, and colleagues analyzed the Nationwide Inpatient Sample and identified 1,749 patients aged 18 years or older (mean age, 55.4 years; 23% women) who had an LVAD implant from 2008 to 2011. They computed annual LVAD implantation volume for each hospital and determined predictors of in-hospital mortality and length of stay. They did not analyze surgeon volume.

For hospitals in the lowest tertile of volume (one to 22 procedures per year), the in-hospital mortality rate was 20.9% vs. 13.7% for hospitals in the highest tertile of volume (35 or more procedures per year; adjusted OR = 0.41; 95% CI, 0.26-0.64), according to the researchers.

Median length of stay was 34 days in the lowest tertile vs. 28 days in the highest tertile (adjusted OR = 0.41; 95% CI, 0.23-0.73), Shah, from the department of cardiology, Lehigh Valley Health Network in Allentown, Pennsylvania, and colleagues found.

When they performed a cubic spline analysis, they found that a center volume of more than 20 LVAD implantations per year conferred mortality rates of less than 10%. That optimal volume “is considerably larger than the current criterion of 10 LVADs over 3 years,” they wrote.

Surgeon volume

Katherine F. Davis, RN, BSN, MSHSM, and colleagues performed a retrospective cross-sectional analysis of 7,714 patients implanted with an LVAD between 2007 and June 2012 by 581 surgeons in 88 hospitals from the University HealthSystem Consortium of academic medical centers.

They determined annual surgeon and hospital volumes and evaluated them by quintiles as well as continuous variables.

Davis, from the division of comparative data and informatics at the University HealthSystem Consortium, Chicago, and colleagues found that hospital volume did not affect in-hospital mortality after controlling for patient and hospital characteristics.

However, they found that patients who had procedures performed by a surgeon in the highest quintile of volume were less likely to experience in-hospital mortality (adjusted OR = 1.69; 95% CI, 1.44-1.97), and a model with that as a significant predictor had an area under the receiver operating characteristic curve of 0.79. They wrote that optimal surgeon volume is more than 12 cases per year, much higher than called for in current guidelines.

“These results suggest that the surgeon’s technical skills, which are refined over time through higher procedure volume, plays a more important role in reducing in-hospital mortality that the more general refinement of care processes for patients receiving an LVAD,” they wrote.

Centralize procedures

In a related editorial, John M. Fallon, MD, and David A. Axelrod, MD, MBA, wrote that the study by Shah and colleagues “was limited by the inherent imprecision of administrative databases,” but “it is likely that these conclusions are robust,” whereas the results of the study by Davis and colleagues “may reflect that the [University HealthSystem Consortium] database is already a selected group of academic centers. Consequently, the conclusions may not be generalizable to smaller community hospitals or nonacademic centers.”

Fallon and Axelrod, both from the department of surgery, Dartmouth-Hitchcock Medical Center in Lebanon, New Hampshire, wrote that hospital volume may be the more persuasive metric because “few would argue for VAD implantation by a high-volume surgeon in a low-volume hospital,” given the support that is required.

“There appears to be a strong argument to centralize high-complexity procedures in high-volume centers, although it is not clear what the long-term impact of future concentration of medical care would be,” they wrote. – by Erik Swain

References:

Davis KF, et al. J Card Fail. 2016;doi:10.1016/j.cardfail.2015.10.012.

Fallon JM, Axelrod DA. J Card Fail. 2016;doi:10.1016/j.cardfail.2016.01.003.

Shah N, et al. J Card Fail. 2016;doi:10.1016/j.cardfail.2015.10.016.

Disclosure: Both sets of researchers and Fallon report no relevant financial disclosures. Axelrod reports being an owner of XynManagement.