Author + information
- Rohan Khera,
- Ambarish Pandey,
- Brahmajee Nallamothu,
- Sandeep Das,
- Mark H. Drazner,
- Michael Jessen,
- Ajay Kirtane,
- Deepak Bhatt,
- James A. de Lemos and
- Dharam Kumbhani
Background: Hospital volume is used as a surrogate for quality among cardiac surgical patients but data are mainly for revascularization procedures. We assessed volume among low- and high-performing hospitals for isolated aortic valve replacement (AVR) and mitral valve repair/replacement (MVR), identified based on risk-standardized in-hospital mortality rates (RSMR).
Methods: We assessed annual volume using a hospital identifier-linked Nationwide Inpatient Sample (2007-2011). RSMR were calculated using hierarchical models with empirical Bayes techniques.
Results: We identified 651 hospitals; median annual volume 45 (IQR 26-79) for AVR, 22 (IQR 12-40) for MVR. Median RSMR were 4.6% (IQR 3.9-5.4) for AVR and 5.6% (IQR 5.1- 6.6) for MVR. There was a weak inverse volume-RSMR association for AVR (ρ -0.16) and MVR (ρ -0.15) (p <0.001 both). There was wide variation in RSMR among highest and lowest hospital volume quartiles for both AVR/MVR (Figure). Proportion of hospitals with above-median RSMR in the highest volume quartile: 42% AVR, 44% MVR; similarly, proportion with below-median RSMR in the lowest volume quartile: 44% AVR, 49% MVR. If volume rather than RSMR was used to define quality, 21% AVR and 24% MVR hospitals would be misclassified.
Discussion: A pure volume-based metric has the potential to misclassify a substantial proportion of low- and high-performing US hospitals for AVR and MVR. It may be more valuable to focus on risk-adjusted outcomes than on volume thresholds alone to determine quality.
Poster Hall, Hall C
Saturday, March 18, 2017, 3:45 p.m.-4:30 p.m.
Session Title: Interventional Cardiology: TAVR 3
Abstract Category: 17. Interventional Cardiology: Aortic Valve Disease
Presentation Number: 1243-174
- 2017 American College of Cardiology Foundation