Temporal Changes in Mortality After Transcatheter and Surgical Aortic Valve Replacement: Retrospective Analysis of US Medicare Patients (2012-2019).

Background The treatment of aortic stenosis is evolving rapidly. Pace of change in the care of patients undergoing transcatheter aortic valve replacement (TAVR) and surgical aortic valve replacement (SAVR) differs. We sought to determine differences in temporal changes in 30-day mortality, 30-day readmission, and length of stay after TAVR and SAVR. Methods and Results We conducted a retrospective cohort study of patients treated in the United States between 2012 and 2019 using data from the Medicare Data Set Analytic File 100% Fee for Service database. We included consecutive patients enrolled in Medicare Parts A and B and aged ≥65 years who had SAVR or transfemoral TAVR. We defined 3 study cohorts, including all SAVR, isolated SAVR (without concomitant procedures), and elective isolated SAVR and TAVR. The primary end point was 30-day mortality; secondary end points were 30-day readmission and length of stay. Statistical models controlled for patient demographics, frailty measured by the Hospital Frailty Risk Score, and comorbidities measured by the Elixhauser Comorbidity Index (ECI). Cox proportional hazard models were developed with TAVR versus SAVR as the main covariates with a 2-way interaction term with index year. We repeated these analyses restricted to full aortic valve replacement hospitals offering both SAVR and TAVR. The main study cohort included 245 269 patients with SAVR and 188 580 patients with TAVR, with mean±SD ages 74.3±6.0 years and 80.7±6.9 years, respectively, and 36.5% and 46.2% female patients, respectively. Patients with TAVR had higher ECI scores (6.4±3.6 versus 4.4±3) and were more frail (55.4% versus 33.5%). Total aortic valve replacement volumes increased 61% during the 7-year span; TAVR volumes surpassed SAVR in 2017. The magnitude of mortality benefit associated with TAVR increased until 2016 in the main cohort (2012: hazard ratio [HR], 0.76 [95% CI, 0.67-0.86]; 2016: HR, 0.39 [95% CI, 0.36-0.43]); although TAVR continued to have lower mortality rates from 2017 to 2019, the magnitude of benefit over SAVR was attenuated. A similar pattern was seen with readmission, with a lower risk of readmission from 2012 to 2016 for patients with TAVR (2012: HR, 0.68 [95% CI, 0.63-0.73]; 2016: HR, 0.43 [95% CI, 0.41-0.45]) followed by a lesser difference from 2017 to 2019. Year over year, TAVR was associated with increasingly shorter lengths of stay compared with SAVR (2012: HR, 1.91 [95% CI, 1.84-1.98]; 2019: HR, 5.34 [95% CI, 5.22-5.45]). These results were consistent in full aortic valve replacement hospitals. Conclusions The rate of improvement in TAVR outpaced SAVR until 2016, with the recent presence of U-shaped phenomena suggesting a narrowing gap between outcomes. Future longitudinal research is needed to determine the long-term implications of lowering risk profiles across treatment options to guide case selection and clinical care.

aortic valve replacement

Elixhauser Comorbidity Index

Hospital Frailty Risk Score

surgical aortic valve replacement

transcatheter aortic valve replacement

Clinical Perspective

What Is New?

The rate of improvement in mortality, readmission, and length of stay and the magnitude of benefit after transcatheter aortic valve replacement has outpaced surgical aortic valve replacement across patient groups.

Recent advances in procedural approaches, multimodality imaging, and the development of streamlined clinical pathways to support early safe discharge home after transcatheter aortic valve replacement have contributed to this temporal trend; the more established practices associated with surgical aortic valve replacement have not been similarly scrutinized.

Both treatments offer excellent options for patients with aortic stenosis, depending on individual risk profiles and patients' goals of care.

What Are the Clinical Implications?

There is a pressing need to ensure equitable access to high‐quality transcatheter aortic valve replacement and surgical aortic valve replacement.

Treatment decisions require individual risk stratification, the consensus decisions of expert multidisciplinary teams, and shared decision‐making to consider patients' preferences and priorities.

Future longitudinal research is needed to evaluate quality of care, patient‐reported outcomes and experiences, and cost‐effectiveness.

The treatment of aortic valve disease has evolved rapidly in the past decade. The increasing availability of transcatheter aortic valve replacement (TAVR) has augmented surgical aortic valve replacement (SAVR) as treatment options for patients of varying risk profiles. During this time, there have been significant improvements made to TAVR technology, imaging, procedural approaches, and processes of care that have contributed to improved outcomes. , In contrast, the procedural and technological environments of SAVR have remained more constant given that it is a more mature procedure with a well‐established track record and historically excellent outcomes. In this context, treatment decisions and patient access to SAVR and TAVR have been primarily driven by indications, health policy and funding, geographical location, and social determinants. , Although local access remains heterogenous, TAVR is surpassing SAVR as the most common form of isolated aortic valve replacement (AVR) in the United States and internationally. ,

Differences in mortality have been scrutinized in multiple clinical trials comparing the 2 treatment modalities using various devices across evolving indications, including the more recent low surgical risk trials. , Yet, there is little on temporal trends in mortality for each modality in contemporary real‐world practice to capture the evolution of treatment of aortic stenosis. There is a pressing need to explore changes in mortality in contemporary practice to inform clinical care, health policy, and shared decision‐making. To this end, we examined temporal differences in clinical outcomes for patients undergoing TAVR and SAVR in the United States between 2012 and 2019.

Methods

This retrospective cohort used data from the US Medicare Dataset Standard Analytic Files 100% fee‐for‐service database. The Medicare fee‐for‐service payer database includes information on the health care services that are covered for beneficiaries enrolled in Medicare parts A and B. Coding and validation details are outlined in Data S1. All data used to perform this analysis were deidentified and accessed in compliance with the Health Insurance Portability and Accountability Act. As a retrospective analysis of a deidentified database, the research was exempt from institutional review board review under 45 Code of Federal Regulations 46.101(b). The need for individual patient consent was waived. We adhered to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines for reporting on cohort studies. The data can be provided upon reasonable request to the corresponding author.

Patient Cohort

We included patients enrolled in Medicare part A and part B between January 1, 2012, to December 31, 2019, who were aged >65 years and had an endovascular transfemoral TAVR or SAVR during that period. The index hospitalization for AVR indicated the first time point. We required a minimum of 1 year of continuous enrollment in Medicare before this time point to establish a baseline period to capture covariates. In our main analysis, all patients with SAVR were part of the cohort, including those who had multiple procedures, such as coronary artery bypass graft, surgical maze procedure, and mitral and tricuspid valve procedures. We repeated the analyses restricted to patients with SAVR with no concomitant procedures (“isolated SAVR”). We also stratified our analyses by elective versus urgent status. For the purpose of this study, an urgent procedure was defined as having the SAVR/TAVR as part of an in‐hospital admission as documented by the hospital record.

Study End Points

The primary end point was 30‐day mortality with the index date being the procedural date for TAVR or SAVR. Secondary end points included 30‐day all‐cause readmission from the date of discharge and length of stay from admission to discharge measured in days.

Covariates

Covariates were obtained from the Medicare records and included patient demographics (age, sex, race, and region) and comorbid conditions profile. We captured comorbid profiles using the Elixhauser Comorbidity Index (ECI; Data S1). To further strengthen the prediction models, we measured frailty as determined the Hospital Frailty Risk Score (HFRS), an International Classification of Diseases, Tenth Revision (ICD‐10) claims‐based frailty score previously validated in patients with TAVR (Data S1). ,

Statistical Analysis

Patient demographics and ECI were summarized for SAVR and TAVR cohorts. All statistical models controlled for patient demographics and ECI. Each set of models was done for (1) the full cohort and then (2) repeated restricted to isolated SAVR cases and (3) further stratified by elective versus urgent status.

30‐Day Mortality

We developed Cox proportional hazard models, which incorporated a 2‐way interaction term with index year and TAVR versus SAVR. The dependent variable was time to death. This model allowed us to determine if the relative difference in mortality between TAVR and SAVR changed over time.

30‐Day All‐Cause Hospital Readmission

We developed cause‐specific Cox proportional hazard models to account for the competing risk of death with the exclusion of patients who died during their index admission. Similar to the mortality models, the main covariate was TAVR versus SAVR, with an interaction term for year of procedure to evaluate a temporal effect.

Length of Stay

Length of stay was modeled using a time to event (ie, discharge) model with in‐hospital death treated as a competing risk event. The cumulative probability of discharge to home was estimated using the cumulative incidence function and compared by Gray's test. Fine and Gray subdistribution hazard modeling was used to evaluate the effect of TAVR versus SAVR on the adjusted probability of discharge while adjusting for age, sex, and ECI scores. A 2‐way interaction term of cohort with index year was included to test whether the probability of discharge for TAVR versus SAVR has changed over time. In interpreting these outputs, a hazard ratio (HR) >1 indicates a shorter time to discharge home.

Effect of AVR Program on Mortality

We developed Cox proportional hazard models with the center included as a random effect to explore the relative mortality impact depending on whether a program performed SAVR only in comparison with a center that offered both TAVR and SAVR. Given that the number of centers performing both SAVR and TAVR procedures increased over time, we conducted this analysis stratified on the index year of the procedure. This was designed to discern if any mortality difference between SAVR and TAVR was restricted to differences in hospital availability of both procedures versus differential temporal improvements in SAVR and TAVR.

Tabulation of summary statistics was performed using the Instant Health Data platform from Boston Health Economics. Models were run using Statistical Analysis Software 9.4, and plots were illustrated in STATA 16. A 2‐sided P value of <0.05 was considered statistically significant.

Results

A total of 433 849 patients who underwent AVR were recorded in the Medicare fee‐for‐service payer database from 2012 to 2019, including 245 269 (56.5%) patients with SAVR and 188 580 (43.4%) patients with TAVR. After applying exclusions for patients who were <65 years at the time of index procedure, did not demonstrate continuous 1‐year enrollment in Medicare, or lacked documentation of AVR (SAVR, 34 023; TAVR, 8683), the main study cohort included 211 246 patients with SAVR and 179 897 patients with TAVR. The subanalysis cohorts included isolated SAVR (n=95 016) and TAVR and elective isolated SAVR (n=76 079) and elective TAVR (n=147 099; Figure 1).

Figure 1

Study cohort (2012–2019; data source: US Medicare Dataset Standard Analytic Files fee‐for‐service database).

AVR indicates aortic valve replacement; SAVR, surgical aortic valve replacement; and TAVR, transcatheter aortic valve replacement.

Baseline Characteristics

In the main cohort, the mean±SD age of the main cohort was 74.3±6.0 years for SAVR and 80.7±6.9 years for TAVR, with women comprising 36.5% and 46.2%, respectively, and White race for 92.5% and 93.3%, respectively (Table). Patients with SAVR had a mean±SD ECI of 4.4±3.0 and a mean±SD HFRS of 4.8±6.1 (33.5% categorized as frail with HFRS ≥5), whereas patients with TAVR reported a mean±SD ECI score of 6.4±3.6 (55.4% categorized as frail) and a mean±SD HFRS of 8.4±8.6. Similar baseline characteristics were reported for the cohorts of isolated SAVR and TAVR and elective isolated SAVR and elective TAVR (Table S1). Among the patients in the full cohort, an elective admission was recorded for 74.8% of SAVR and 81.8% of TAVR.

Table 1

Baseline Characteristics by Study Cohort

	Full cohort		Isolated SAVR	Elective TAVR and elective isolated SAVR
	SAVR	TAVR		SAVR	TAVR
Total patients	211 246	179 897	95 016	76 079	147 099
Age, mean±SD, y	74.3±6.0	80.7±6.9	73.9±6.0	73.8±5.9	80.6±6.9
Sex, female patient	77 078 (36.5)	83 194 (46.2)	39 874 (42.0)	32 069 (42.2)	67 931 (46.2)
White race	195 409 (92.5)	167 801 (93.3)	87 272 (91.8)	70 485 (92.6)	137 923 (93.8)
Elective procedure	158 053 (74.8)	147 099 (81.8)	76 079 (80.1)	N/A	N/A
Congestive heart failure	58 120 (27.5)	99 102 (55.1)	24 185 (25.5)	18 909 (24.9)	80 585 (54.8)
Cardiac arrhythmia	71 265 (33.7)	90 917 (50.5)	27 103 (28.5)	21 682 (28.5)	74 455 (50.6)
Hypertension*	39 660 (18.8)	79 805 (44.4)	16 179 (17.0)	12 597 (16.6)	64 834 (44.1)
Chronic pulmonary disease	55 017 (26.0)	63 088 (35.1)	24 695 (26.0)	20 000 (26.3)	51 313 (34.9)
Diabetes*	19 198 (9.1)	35 258 (19.6)	7259 (7.6)	5487 (7.2)	28 352 (19.3)
Peripheral vascular disorders	57 083 (27.0)	70 679 (39.3)	26 591 (28.0)	22 636 (29.8)	59 784 (40.6)
Renal failure	31 296 (14.8)	56 842 (31.6)	12 655 (13.3)	9564 (12.6)	45 197 (30.7)
Obesity	30 343 (14.4)	32 607 (18.1)	13 870 (14.6)	11 438 (15.0)	27 104 (18.4)
Liver disease	8144 (3.9)	11 718 (6.5)	3889 (4.1)	3133 (4.1)	9865 (6.7)
Deficiency anemia	14 942 (7.1)	24 261 (13.5)	6366 (6.7)	4629 (6.1)	19 018 (12.9)
Depression	16 814 (8.0)	22 099 (12.3)	7871 (8.3)	6074 (8.0)	17 740 (12.1)
Elixhauser Comorbidity Index	4.4±3.0	6.4±3.6	4.3±2.9	4.4±2.7	6.4±3.5
Hospital Frailty Risk Score	4.8±6.1	8.4±8.6	4.7±6.0	4.4±5.5	8.2±8.3
Frail, Hospital Frailty Risk Score ≥5	70 720 (33.5)	99 740 (55.4)	30 859 (32.5)	23 854 (31.4)	80 657 (54.8)

N/A, not applicable.

Data are provided as number (percentage) or mean±SD. The table highlights the most pertinent comorbidities. Study analyses were conducted with the full complement of the 31 variables included in the Elixhauser Comorbidity Index (Data S1). SAVR indicates surgical aortic valve replacement; and TAVR, transcatheter aortic valve replacement.

Classified as “complicated.”

Temporal Changes in Procedure Volumes

Total AVR volumes increased 61% from 36 861 in 2012 to 59 357 in 2019. TAVR volumes grew annually, equaled, and then surpassed SAVR in 2016 and 2017, respectively (Figure 2A). The distribution ratio of TAVR versus SAVR changed from 12% in 2012 to 72% in 2019 (Figure 2B).

Figure 2

Temporal changes in procedure volumes and availability of TAVR in the United States (2012–2019).

A, Total aortic valve replacement volumes. B, Ratio of SAVR vs TAVR. C, Growth of TAVR centers and median annual volume of procedures. AVR indicates aortic valve replacement; SAVR, surgical aortic valve replacement; and TAVR, transcatheter aortic valve replacement.

In the unadjusted model for the main cohort, we found that the 30‐day mortality rates decreased from 4.8% to 4.6% for SAVR and from 6.3% to 2.0% for TAVR during the study period (Figure S1). When adjusting for age, sex, ECI, and HFRS, we found a statistically significant temporal effect on the relative efficacy of TAVR versus SAVR (Figure 3). In all of the years, TAVR was associated with a lower mortality rate than SAVR. This relationship was complex and showed a U‐shaped curve. Year over year, the magnitude of benefit associated with TAVR increased until 2016, with a HR of 0.76 (95% CI, 0.67–0.86) in 2012 to 0.39 (95% CI, 0.36–0.43) in 2016. Although TAVR continued to have lower mortality rates from 2017 to 2019, the magnitude of benefit over SAVR was attenuated. Analyses of the TAVR versus isolated SAVR and the elective TAVR versus elective isolated SAVR cohorts showed similar results.

Figure 3

End point analysis—Cox proportional hazard models of time to 30‐day mortality and 30‐day readmission and Fine and Gray subdistribution models on length of stay (2‐way interaction term with index year).

The diamond markers represent the HRs, and the bars indicate the 95% CIs. HR indicates hazard ratio; SAVR, surgical aortic valve replacement; and TAVR, transcatheter aortic valve replacement.

30‐Day All‐Cause Readmission

In the main cohort, year over year, the incidence of unadjusted 30‐day all‐cause readmission declined for both SAVR (from 18.2% to 13.6%) and TAVR (20.6% to 10.9%; Figure S2). When adjusted for baseline differences,TAVR was associated with a lower risk of all‐cause readmission with the magnitude of benefit of TAVR over SAVR varying over time from a HR of 0.68 (95% CI, 0.63–0.73) in 2012 to a low of 0.43 (95% CI, 0.41–0.45) in 2016 and increasing to 0.75 (95% CI, 0.72–0.80; Figure 3). A similar trend was seen across the additional isolated SAVR and elective TAVR versus elective isolated SAVR study cohorts.

In‐Hospital Length of Stay

In all cohorts, the pattern of temporal change in length of stay was consistent: there were small decreases in SAVR (13.9% decrease in mean length of stay in the main cohort: 12.2 days in 2012 versus 10.5 in 2019) and larger decreases for TAVR (61.2% decrease in the main cohort: 9.1 days in 2012 versus 3.5 in 2019; Figure S3). When adjusted for covariates and assessed for a temporal effect, a consistent benefit in length of stay was seen favoring TAVR over SAVR, with an increase in the magnitude of this benefit year over year. Specifically, the adjusted HR increased from 1.91 (95% CI, 1.84, 1.98) in 2012 to 5.34 (95% CI, 5.22, 5.45) in 2019 in the main cohort (Figure 3).

Effect of AVR Program on 30‐Day Mortality

Over time, the proportion of hospitals that offered both SAVR and TAVR grew from 22.9% (2012, n=258) to 66.7% (2019, n=665), with the median number of TAVR procedures performed increasing from 11 (interquartile range, 4–22) to 46 (interquartile range, 24–77; Figure 2C). When we restricted our 30‐day mortality modeling to include only hospitals that performed both TAVR and SAVR, we found a consistent benefit associated with TAVR over SAVR that followed the same U‐shaped curve as our primary results (Figure 3).

Discussion

In this retrospective analysis of temporal trends in outcomes after TAVR and SAVR in the United States, we found that over time, there was a complex relationship in terms of degree of improvement in 30‐day mortality and readmission and in‐hospital length of stay favoring TAVR. To our knowledge, this is the first study to examine changes over time in a large series of real‐world patients treated with AVR across multiple centers and during the course of ongoing changes in indications, technology, and clinical care. This study provides novel evidence about the degree and pace of improvement between the 2 treatment options over time as indications, case selection, technology, and clinical pathways continue to evolve. The findings strengthen the evidence that TAVR offers a therapy that continues to rapidly gain an advantage in comparison to the rate of change of SAVR and translates in growing magnitude of benefit across patient groups.

TAVR has experienced significant evolution since 2012. In the early days, TAVR was reserved for the sickest patients, many of whom were ineligible for surgery. Over time, indications expanded to patients who were healthier and case selection was refined to exclude excessively patients unlikely to derive the survival and quality‐of‐life benefits. Thus, it is not surprising that TAVR outcomes have improved over time. The presence of a U‐shape phenomena may indicate that in the most recent years, as mostly at very low surgical risk are undergoing SAVR, there is a narrowing of the gap between the outcomes after SAVR and TAVR.

Technology continues to improve to minimize known complications, including vascular injury, bleeding, paravalvular leak, stroke, and arrhythmias, although ongoing quality improvement efforts remain under way. , In parallel, there has been a shift to the adoption of a streamlined clinical pathway focused on reducing risks associated with the use of invasive interventions and in‐hospital deconditioning and to reduce length of stay. , , The adoption of a minimalist approach with a strategy of local anesthesia and/or procedural sedation, the avoidance of invasive monitoring lines, and peri‐procedure strategies aimed at increasing the predictability of uncomplicated hemostasis and hemodynamic stability have facilitated the rapid reconditioning of patients with TAVR. The 3M TAVR (Multimodality, Multidisciplinary but Minimalist TAVR) study demonstrated the safety, feasibility, and reproducibility of the Vancouver Clinical Pathway to facilitate a safe next‐day discharge home. Other studies have added to this evidence. , Consequently, many centers have replaced historical clinical practices informed by cardiac surgery protocols in favor of practices better matched to TAVR, patient risks, and contemporary evidence. In contrast, the long‐standing excellent outcomes achieved by SAVR centers may not have created similar pressures to scrutinize all aspects of the patients' journey of care to identify opportunities for quality improvement across programs. SAVR programs may continue to report the ceiling effect of the invasiness of surgery that limits the magnitude of possible reductions in length of stay. Advances to surgical protocols, including the use of mini‐thoracotomies and robotic and off‐pump surgeries, and the implementation of enhanced recovery after surgery pathways continue to yield improved outcomes. , In our study, we found that 30‐day all‐cause readmission rates for patients with SAVR followed a similarly robust trend compared with TAVR and decreased from 18.2% in 2012 to 13.6%, indicating improved transitions of care after surgery. Nevertheless, the relative slower rate of improved outcomes seen in SAVR in this study and the resultant growing gap between therapies may reflect missed opportunities to address complications such as the incidence of delirium, surgical site or other infections, and delayed mobilization and discharge that are known to adversely impact the primarily older aortic stenosis population. ,

We found that this temporal observation was consistent, albeit attenuated, when we restricted our analyses to centers that offer both TAVR and SAVR. This is a critical observation as it reinforces 2 key points. First, that even in centers that have both TAVR and SAVR, there was a greater degree of improvement in TAVR outcomes compared with SAVR over time. Second, if a hospital only offered SAVR, there was an even greater difference in outcomes, as patients who would have received TAVR in other centers were receiving SAVR in that center. This raises further questions about the potential impact of smaller AVR centers that may not offer the full complement of services or may be adversely impacted by low procedural volumes. ,

As such, our study has important implications for promoting patient access to both treatment options and advocating for AVR centers to offer high‐quality TAVR and SAVR supported by a coordinated process, a multidisciplinary evaluation pathway and treatment decision, and an embedded adoption of shared decision‐making to tailor recommendations to patients' unique risks, values, and priorities. , Access to both treatment options with surveillance of equity of access and quality of care remains essential. There has been a vigorous health policy debate related to promoting quality of care for TAVR and concerns about volume/outcome relationships. , Recent analyses of the US TVT (Transcatheter Valve Therapy) Registry reported a significant inverse association between annual transfemoral TAVR volumes and mortality, with hospitals in the lowest volume quartile (mean, 27 procedures) reporting higher and more variable 30‐day mortality (3.2%; 95% CI, 2.8–3.7) compared with hospitals in the highest volume quartile (mean, 143 procedures; 30‐day mortality, 2.7%; 95% CI, 2.5–2.9). This trend was previously reported in an earlier report and is further supported by evidence that higher volume single operators have superior outcomes. These findings informed the recent US Centers for Medicare & Medicaid Services National Coverage Determination that requires centers to perform ≥20 TAVR/year in addition to meeting other volume, programmatic, and reporting requirements. Similar scrutiny of SAVR has not been similarly intense or debated. The shift to a comprehensive management of patients with aortic stenosis requires the removal of silos that currently separate TAVR and SAVR clinical processes, health policy and funding models, surveillance of access and outcomes, and quality improvement to achieve a more patient‐centred and procedure‐agnostic approach to disease management.

Additional considerations include the measurement of health status and economic analyses. In low‐risk clinical trials, TAVR was associated with better health status than SAVR at 1 month , ; in a recent analysis of very early changes in quality of life in the 3M TAVR study, most patients reported large improvements by 2 weeks after the procedure, with modest additional benefit from 2 weeks to 1 month and sustained improvement through 1 year follow‐up as seen in clinical trials. In contrast, among patients undergoing SAVR, deriving quality‐of‐life benefit requires a longer postsurgical recovery. , In the context of our study, the inclusion of patient‐reported outcomes in treatment recommendation and shared decision‐making enables the clinician to convey the best evidence about risks and benefits, whereas the patients can inform their providers about their goals, personal preferences, and values. Similarly, our study findings may further inform the consideration of the economic impact of treatment, as TAVR is projected to be economically dominant by providing greater quality‐adjusted life expectancy and lower long‐term costs than SAVR, driven in part by the lower health resource requirements and shorter length of stay associated with TAVR. Momentum for the adoption of minimalist procedural approaches have cost‐lowering implications for TAVR programs. Ongoing research is needed to integrate a comprehensive evaluation of contemporary practice inclusive of clinician‐reported and patient‐reported outcomes and health service use.

Limitations

This was a retrospective observational study of site‐reported administrative data. Medicare fee for service does not include the full cohort of the Medicare patients with the rate of Medicare Advantage enrollment increasing annually since 2012. Adjustment for patient and procedural factors accounted for reported multiple factors, including frailty; nevertheless, we may not have captured the full complement of determinants of outcomes. The study does not fully account for the diversity of other potentially complex issues such as valve technology, repeat procedures, in‐hospital complications, evolving case selection for both SAVR and TAVR, operator and program experience, and variations in reporting. Analyses were limited to early outcomes.

Conclusions

In this large‐scale, real‐world report of temporal changes in outcomes in patients treated with AVR, we demonstrated an accelerated improvement of TAVR over SAVR. As indications for TAVR continue to expand, our data highlight opportunities to pursue technological improvements and the implementation of clinical best practices to continue to drive quality improvement. A similar focus on SAVR may offer patients who require a more invasive approach to achieve an optimal outcome.

Sources of Funding

Data analyses were funded by Edwards LifeSciences.

Disclosures

Dr Lauck has worked as a consultant for Edwards Lifesciences and Medtronic. Dr Baron has worked as a consultant for Boston Scientific Corporation, Abiomed, Abbott, Edwards Lifesciences, and Mitra Labs and has received research support from Abiomed. Dr Wood has worked as a consultant for Edwards Lifesciences and has received research support from Edwards Lifesciences and Abbott. Dr Webb has worked as a consultant for Edwards Lifesciences, Abbott, and Boston Scientific. Dr Moore and S. Clancy are employees of Edwards Lifesciences. The remaining authors have no disclosures to report.

Data S1

Table S1

Figures S1–S3

Click here for additional data file.