Length of Stay and Discharge Disposition After Transcatheter Versus Surgical Aortic Valve Replacement in the United States

Abstract

Background: As transcatheter aortic valve replacement (TAVR) extends its reach to lower surgical risk patients, the differences between resource utilization for TAVR and surgical AVR (SAVR) will become increasingly important. Methods and Results: AVR procedures between January 2012 and September 2015 at hospitals performing TAVR were identified using the National Inpatient Sample databases. Adults aged ≥50 years with aortic stenosis who underwent isolated TAVR or SAVR were eligible for inclusion. Standardized morbidity ratio weights were calculated using patient demographics, comorbidities, and hospital characteristics. Weighted linear and generalized logistic regression models were used to estimate the effect of undergoing TAVR, compared with undergoing SAVR, on length of stay (LOS) and discharge disposition. In TAVR-performing hospitals, 7266 (40%) patients underwent TAVR (6107 endovascular approach and 1159 transapical approach), while 10 833 (60%) underwent isolated SAVR. Patients undergoing TAVR were older, more likely to be female, and had more comorbidities. From 2012 to 2015, average LOS declined for both TAVR (6.3 days to 4.6 days; P <0.0001) and SAVR (7.5 days to 6.8 days; P <0.0001), with greater reduction in the TAVR group ( P <0.0001). An increase in home/home health discharge was noted with TAVR (67.7%–77.4%; P <0.0001) but not with SAVR (76.8%–79.5%; P =0.25). After standardizing, patients undergoing TAVR had significantly shorter LOS (change in estimate, −2.93, 95% CI, −3.26 to −2.60) and lower incidence of transfer to skilled nursing facility (odds ratio, 0.45; 95% CI, 0.40–0.51) but no difference in in-hospital mortality (odds ratio, 0.85; 95% CI, 0.61–1.20) compared with if they had undergone SAVR. As compared with SAVR, patients who had TAVR performed via an endovascular approach had shorter LOS and lower rates of skilled nursing facility transfer, whereas in the transapical cohort, LOS, and skilled nursing facility transfer were similar to SAVR. Conclusions: As compared with if they undergo SAVR, patients undergoing TAVR (by a nontransapical approach) had a shorter LOS and higher likelihood of home discharge, as opposed to skilled nursing facility. From 2012 to 2015, there was a greater trend towards a reduction of LOS and more home discharges among TAVR, as opposed to SAVR. These data have important implications in the era of constrained resources with a growing emphasis on reducing health care costs.