ADVERTISEMENT
Utilization Trends and Short-Term Outcomes for Transcatheter and Surgical Aortic Valve Replacement Surgery in New York
Abstract
Background. Population-based utilization trends and outcomes of transcatheter aortic valve replacement (TAVR) and surgical aortic valve replacement (SAVR) remain unknown. Objectives. To examine the utilization trends and outcomes of TAVR and SAVR in New York using all-inclusive aggregated statewide cardiac registries. Methods. We described the utilization trends, compared baseline characteristics, and evaluated short-term outcomes of TAVR vs SAVR during 2011-2018 in New York. We applied Cox proportional hazards models to analyze changes in 30-day postoperative mortality for TAVR and SAVR. Results. Of a total 37,566 aortic valve replacement (AVR) patients, 50.8% underwent TAVR and 49.2% received SAVR. TAVR’s annual volume increased from 715 in 2012 to 4849 in 2018 (578.18% increase) whereas SAVR’s annual volume decreased from 2619 in 2012 to 1855 in 2018 (29.17% decrease). TAVR patients were older, more likely to be female and white, and less likely to be Hispanic. Younger patients (<65 years) and Medicare managed-care patients received TAVR (vs SAVR) a lower percentage of the time relative to older patients (≥65 years) and Medicare fee-for-service patients, respectively. In 2018, the unadjusted 30-day mortality rate was 2.37% for TAVR whereas the rate was 0.97% for SAVR. There was significant annual improvement in 30-day mortality for TAVR (annual adjusted hazard ratio, 0.84, 95% confidence interval, 0.80-0.88) but not for SAVR (annual adjusted hazard ratio, 0.96; 95% confidence interval, 0.91-1.01). Conclusions.TAVR and AVR experienced massive growth whereas SAVR decreased in New York. Younger and Medicare managed-care patients had unique utilization trends. TAVR was associated with continuous improvement in 30-day postoperative mortality.
Keywords: short-term outcomes, surgical aortic valve replacement, transcatheter aortic valve replacement, utilization trends
Aortic stenosis (AS) is one of the most common valvular diseases in the United States (US). Approximately 2.5 million people over age 75 years (12% of that population) suffer from this disease.1 As the prevalence of AS increases with age and the US population ages rapidly, the number of elderly patients with AS is projected to more than double by 2050.2 Transcatheter aortic valve replacement (TAVR) is widely perceived as a transformative innovation in treating AS patients. The effectiveness and safety of TAVR have been reported by randomized clinical trials and large observational studies.3,4 TAVR was approved in sequence by the US Food and Drug Administration (FDA) for prohibitive risk (2011), high-risk (2014), intermediate-risk (2016), and low-risk (2019) AS patients.5 Before the introduction of TAVR, open-heart surgical aortic valve replacement (SAVR) was the gold standard for aortic valve replacement (AVR) due to the fact that it was the only effective long-term therapy for AS.
The advent and evolution of TAVR have dramatically changed the overall landscape of AVR during the last decade. Notably, younger and low-risk patients are more likely to receive TAVR in recent years.6-8 According to a recent study, there has been a remarkable increase in both TAVR and AVR but a steady decline in SAVR from 2012 to 2019 among Medicare fee-for-service patients with AS.9 Unlike most previous research, which evaluated TAVR and SAVR in isolation, that study assessed the totality of AVR care by examining the real-world provision of both TAVR and SAVR in the Medicare fee-for-service population. Its findings highlight the importance of fully describing the totality of AVR care and the need to address challenging public health issues related to actual provision of AVR care. However, the study was limited to only Medicare fee-for-service beneficiaries. Thus, it remains unknown about TAVR and AVR utilization and outcomes in increasingly important patient subgroups such as younger patients (<65 years) and Medicare managed-care patients. Moreover, the Medicare claims data lacked important clinical granularity. Therefore, much is unknown about the population-based utilization trends and outcomes of TAVR and SAVR in real-world settings.
To fill the knowledge gap, high-quality cardiac registries need to be linked to the population-based administrative claims in order to address crucial implementation and health policy questions specific to AVR care.5 In this study, we aimed to examine real-world population-based utilization trends and short-term outcomes of TAVR and SAVR after the first FDA approval of TAVR use in New York with the all-payer aggregated statewide cardiac registries.
Methods
Data source and study sample. We used all-inclusive aggregated data from the New York State (NYS) Cardiac Surgery Reporting System (CSRS), the Statewide Planning and Research Cooperative System (SPARCS), and the Statewide Vital Statistics dataset from 2011 through 2018.
The NYS CSRS registry was developed by the NYS Department of Health in 1989 with the goal of evaluating and improving the quality of cardiac surgery in New York.10 It is a comprehensive and high-quality registry that has been routinely used to publish statewide public reports on quality and outcomes for cardiac procedures. It includes important information on patient demographics, preprocedural risk factors, hemodynamic status, procedure details, hospital and operator identifiers, postprocedure complications, and discharge status for all cardiac surgeries performed in non-federal hospitals with Certificate of Need (CON) approval to perform the procedures in NYS. The CSRS’s evident strengths include its completeness, quality, reliability, and timeliness of clinically relevant data that are ensured through rigorous systematic efforts. In addition, the data elements and variable definitions of the CSRS registry are determined, reviewed, and continuously updated by the NYS Cardiac Advisory Committee (CAC), which consists of cardiologists, cardiac surgeons, and experts in cardiovascular quality and outcomes, access, and ethics.
The NYS SPARCS database was created in 1979 to serve as a statewide comprehensive all-payer claims data reporting system. The SPARCS dataset contains all-inclusive patient discharges from non-federal acute care hospitals in NYS. It contains information on patient demographics, diagnoses, procedures, admission date, discharge date, and discharge disposition. In this study, SPARCS data were used to confirm TAVR and SAVR hospitalizations and length of stay (LOS) for each TAVR/SAVR hospitalization during 2011-2018.
The NYS Vital Statistics dataset was matched to the CSRS registry using unique patient identifiers to obtain 30-day postoperative deaths that occurred after the index TAVR/SAVR procedure. In this study, we limited patients to NYS residents to minimize the chance of postdischarge mortality occurring outside of NYS.
Patient baseline characteristics. Patient baseline characteristics include demographics (age, gender, race, ethnicity, primary payer), priority status (elective, urgent, emergency, and salvage), cardiovascular risk factors (number of coronary vessels diseased, left main lesion, acute myocardial infarction [AMI], left ventricular ejection fraction, cerebrovascular disease, peripheral vascular disease, malignant ventricular arrhythmia, heart failure [HF], previous percutaneous coronary intervention [PCI], previous coronary artery bypass graft [CABG], previous valve surgery), and other comorbidities (diabetes, hepatic failure, renal dysfunction, etc). These patient-level baseline characteristics have been used in NYS routine public reporting for quality and outcomes in cardiac surgery and prior studies published by our research team and collaborators.10,11Supplemental Table S1, Part 1 and Supplemental Table S1, Part 2 provides a complete list of variables with definitions.
Outcome. The primary outcome measure was 30-day postoperative mortality after the index TAVR/SAVR surgery. This was defined as all-cause death occurring within the 30-day frame from the index TAVR/SAVR procedure. Secondary outcome measures included LOS and hospital discharge disposition. The LOS information was obtained from the CSRS registry and verified by the SPARCS database. The hospital discharge disposition was categorized into: (1) home; (2) acute care facility; (3) skilled nursing facility; (4) physical medicine and rehabilitation; and (5) other facilities, such as hospice.
Statistical analysis. We evaluated baseline characteristics for patients who received isolated SAVR and isolated TAVR during the study period. We defined isolated SAVR/TAVR as patients received either SAVR or TAVR surgery without any other heart operations during the hospitalization. Means and standard deviations or counts and percentages were reported to present patient-level demographics, priority status, cardiovascular risk factors, and other comorbidities. For the primary outcome measure of postoperative 30-day mortality, counts and percentages were used. For the LOS outcome measure, medians and interquartile ranges (IQRs) were reported. For the hospital discharge disposition outcome measure, counts and percentages were used. We compared patient-level baseline characteristics between isolated SAVR and isolated TAVR groups for 3 different time frames: the full study period of 2011-2018, initial years of TAVR adoption 2011-2012, and the last year of the study period 2018. We applied Chi-square tests for categorical data and t tests for continuous data in comparing baseline characteristics between the 2 groups. We also compared utilization trends and short-term outcomes between younger (<65 years) and older patients (≥65 years) as well as between Medicare fee-for-service and Medicare managed-care patients. In addition, we computed risk-adjusted 30-day postoperative mortality rates for TAVR, SAVR, and AVR in 2012 and 2018 by using a multivariable equation that was derived from the risk-adjustment model accounting for a comprehensive set of patient-level risk factors and using all AVR cases during the whole study period. To examine the annual change in 30-day postoperative mortality rates for TAVR, we used a Cox proportional hazard models with hospital random intercepts and adjusted for patient demographics, priority status, cardiovascular risk factors, and comorbidities. The dependent variable was time until the event, 30-day death after the isolated TAVR/SAVR surgery. The model also included an ordinal time variable, ranging from 0-7, representing the years 2011 to 2018 in chronological order. Using the same Cox proportional hazards model specification, we repeated the analysis for SAVR and AVR, respectively. We used SAS, version 9.4 software (SAS Institute) for all descriptive and regression analyses. We conducted 2-sided tests with a .05 significance level for hypothesis testing.
Results
Trends in TAVR/SAVR/AVR utilization. During the study period of 2011-2018, there was a total of 37,566 isolated AVR hospitalizations in NYS. Among them, there were 19,099 isolated TAVR hospitalizations (50.84%) and 18,467 isolated SAVR hospitalizations (49.16%). As 2011 was the first year of TAVR adoption and the case volume of TAVR was very small, we decided to report utilization trends of TAVR and SAVR starting from 2012. TAVR’s annual volume increased from 715 in 2012 to 4849 in 2018 (578.18% increase) whereas SAVR’s annual volume decreased from 2619 in 2012 to 1855 in 2018 (29.17% decrease) (Figure 1 and Supplemental Table S2, Part 1, Part 2, Part 3, and Part 4; Supplemental Table S3, Part 1, Part 2, Part 3, and Part 4; and Supplemental Table S4, Part 1, Part 2, Part 3, and Part 4 for TAVR, SAVR, and AVR utilization trends and patient baseline characteristics).
For younger (<65 years) patients, TAVR’s annual volume increased remarkably from 33 in 2012 to 186 in 2018 (463.63% increase) and SAVR’s annual volume increased slightly from 711 in 2012 to 806 in 2018 (13.36% increase). For older (≥65 years) patients, TAVR’s annual volume increased at a faster pace from 682 in 2012 to 4663 in 2018 (583.72% increase) and SAVR’s annual volume decreased markedly from 1908 in 2012 to 1049 in 2018 (45.02% decrease) (Figure 2). From 2012 to 2018, the percentage of getting TAVR in younger patients increased from 4.44% (33 TAVRs out of 744 AVRs) to 18.75% (186 TAVRs out of 992 AVRs) whereas the percentage of getting TAVR in older patients increased from 26.33% (682 TAVRs out of 2590 AVRs) to 81.64% (4663 TAVRs out of 5712 AVRs).
For Medicare fee-for-service patients, TAVR’s annual volume increased substantially from 422 in 2012 to 3101 in 2018 (634.83% increase) and SAVR’s annual volume decreased greatly from 1144 in 2012 to 605 in 2018 (47.12% decrease). For Medicare managed-care patients, TAVR’s annual volume increased even more remarkably from 142 in 2012 to 1262 in 2018 (788.73% increase) and SAVR’s annual volume decreased from 486 in 2012 to 375 in 2018 (22.84% decrease) (Figure 3). From 2012 to 2018, the percentage of getting TAVR in Medicare fee-for-service AVR patients increased from 26.95% (422 TAVRs out of 1566 AVRs) to 83.68% (3101 TAVRs out of 3706 AVRs) whereas the percentage of getting TAVR in Medicare managed-care AVR patients increased from 22.61% (142 TAVRs out of 628 AVRs) to 77.09% (1262 TAVRs out of 1637 AVRs).
Patient characteristics. The mean age increased over time from 73.09 years in 2012 to 76.62 years in 2018 for all AVR hospitalizations. However, the mean age decreased from 82.90 years to 81.20 years for TAVR hospitalizations and the mean age also decreased from 70.41 years to 64.63 years for SAVR hospitalizations. Thus, our data demonstrated the "Will Rogers" phenomenon whereby the direction of changes in the subgroups (ie, TAVR and SAVR subgroups) contradicts the overall AVR group’s trend with regard to the patients’ mean age change during the study period.12 This finding highlights the importance of examining the totality of AVR care using real-world all-inclusive datasets.
The percentage of female patients remained similar in all AVR hospitalizations (47.54% in 2012 and 45.51% in 2018). However, for isolated TAVR hospitalizations, the percentage of female patients decreased from 54.13% in 2012 to 49.14% in 2018. For isolated SAVR hospitalizations, the percentage of female patients had an even larger decrease from 45.74% to 36.01%. Among patients who received isolated TAVR surgery, the percentage of White patients decreased from 95.24% in 2012 to 90.93% in 2018, whereas the percentage slightly increased for Blacks (3.78% to 4.50%), Asian Americans (0.98% to 1.84%), and Hispanic ethnicity patients (3.64% to 5.28%). Yet, when we looked at the numbers of TAVR cases, we found that the number of TAVR cases kept increasing every year across subgroups of gender, race, ethnicity, and payer.
A significant decrease (defined as >10% decrease) in prevalence of comorbidity over time was previous CABG with patent graft before hospitalization for TAVR (28.81% in 2012 to 16.46% in 2018). There was no significant increase (defined as >10% increase) in prevalence of any comorbidity over time for either TAVR or SAVR hospitalizations. In the all AVR group, there was a significant increase in the prevalence of HF over time (45.02% to 62.43%).
When we compared patient baseline characteristics between TAVR and SAVR groups for the entire study period, we found that TAVR patients were older, more likely to be female, White, and have Medicare insurance, less likely to be Hispanic and have coronary artery disease, and more likely to have AMI, cerebrovascular disease, peripheral artery disease, HF, diabetes, previous PCI, previous CABG patent grafts, and previous valve surgery (see Table 1, Part 1; Table 1, Part 2; and Supplemental Table S5, Part 1, Part 2, Part 3, and Part 4 for full details). When we repeated the comparisons using data from the early years of 2011-2012 and data from the last year of 2018, we found that these significant differences in baseline characteristics between TAVR and SAVR groups persisted.
Trends in outcomes. Between 2012 and 2018, the unadjusted 30-day postoperative mortality rate declined for isolated TAVR from 6.43% to 2.37%, declined for isolated SAVR from 1.57% to 0.97%, and declined for all AVR from 2.61% to 1.98% (Figure 4). The median LOS for isolated TAVR decreased from 8 days (IQR, 5-14) in 2012 to 2 days (IQR, 1-5) in 2018, decreased for isolated SAVR from median LOS of 8 days (IQR, 6-12) in 2012 to 6 days (IQR, 5-9) in 2018, and decreased for isolated AVR from median LOS of 8 days (IQR, 6-12) in 2012 to 3 days (IQR, 2-7). As for the hospital discharge disposition from 2012 to 2018, there was a steady increase in percentage of being discharged home directly for isolated TAVR (66.12% to 88.47%), isolated SAVR (71.17% to 88.80%), and all AVR (70.12% to 88.56%) and a stable decrease in percentages of being discharged to skilled nursing facilities and physical medicine facilities for TAVR (14.79% to 8.14% and 17.75% to 2.70%, respectively), for SAVR (13.09% to 5.98% and 11.81% to 4.08%, respectively), and for all AVR (13.45% to 7.54% and 13.05% to 3.08%, respectively) (Table 2).
The risk-adjusted 30-day postoperative mortality rate for TAVR decreased from 4.39% (95% confidence interval [CI], 2.96-6.27) in 2012 to 2.17% (95% CI, 1.78-2.63) in 2018. During the same period, the risk-adjusted 30-day postoperative mortality rate for SAVR decreased from 1.09% (95% CI, 0.76-1.52) to 0.84% (95% CI, 0.46-1.41) and the risk-adjusted 30-day postoperative mortality rate for all AVR decreased from 2.30% (95% CI, 1.77-2.93) to 1.61% (95% CI, 1.33-1.93) (Table 3). In addition, the annual adjusted changes in 30-day postoperative mortality rates derived from the Cox proportional hazards models were expressed as annual adjusted hazard ratios (AAHRs). We found that there was significant improvement in AAHR for 30-day postoperative mortality rate for isolated TAVR (AAHR, 0.84; 95% CI, 0.80-0.88) and for all AVR (AAHR, 0.88; 95% CI, 0.86-0.91) but not for isolated SAVR (AAHR, 0.96; 95% CI, 0.91-1.01).
Discussion
To the best of our knowledge, this is the first large study to present the overall landscape of real-world AVR using population-based aggregated cardiac registry datasets. Notably, we examined increasingly important patient subgroups such as younger patients and Medicare managed-care patients who underwent TAVR/SAVR. We found that the volumes of TAVR and AVR experienced rapid growth while SAVR volume decreased steadily in NYS. And, TAVR and SAVR patients had significantly different baseline characteristics. In 2018, the risk-adjusted 30-day postoperative mortality rate was 2.17% for TAVR whereas the rate was 0.84% for SAVR. There was significant improvement in annual 30-day postoperative mortality over time for TAVR and AVR, but not for SAVR.
Our study has several unique strengths. First, our study fills the knowledge gap by providing real-world evidence of TAVR utilization for previously understudied patient subgroups such as younger (<65 years) patients and Medicare managed-care patients. Our findings suggest that both TAVR and SAVR annual case volumes in younger patients increased over time. We also observed that the percentage of getting TAVR increased over time for both Medicare fee-for-service patients (26.95% to 83.68%) and Medicare managed-care patients (22.61% to 77.09%). These results warrant further investigation to identify the causes for detected differences in these subgroups. As enrollment in Medicare managed-care plans continues to increase rapidly in the US and an increasing number of younger patients will likely receive TAVR/SAVR, we believe that it is crucial to continuously monitor and evaluate the baseline characteristics, access, and outcomes in AVR care across subgroups. Also, our findings suggest that Mori et al might overestimate the TAVR prevalence rate in the Medicare population, as that study did not include both younger patients and Medicare managed-care patients, who were found to have lower TAVR rates in our study.9 Second, we examined the real-world total AVR care with the aggregated cardiac registry datasets. Therefore, our study has advantages over previous publications that have evaluated patient profiles and outcomes separately for TAVR and SAVR due to the fact that SAVR data were very limited13 or included only older Medicare fee-for-service beneficiaries but excluded Medicare managed-care plan patients and the increasing numbers of younger patients with other types of insurance coverage (eg, private insurance, Medicaid, etc),9,14 or use only healthcare claims datasets, which usually lack important clinical details and suffer from data quality issues.15 It is noteworthy that linked registry datasets with claims database such as ours were recently recommended for real-world AVR/TAVR/SAVR practice research.5 Third, the high quality of the CSRS registry allowed us to compare a comprehensive list of patient-level factors between TAVR and SAVR groups. Our findings highlight the urgent need to address public health issues in AVR care, such as access and disparity issues related to TAVR for racial/ethnic minorities and challenges in providing personalized AVR care. Additionally, we identified the Will Rogers phenomenon in patients’ mean age and prevalence rate of HF in AVR patients. Our finding of TAVR/SAVR/AVR patients’ mean age change was similar to the result reported by Mori et al in their recent paper that studied TAVR utilization trends with Medicare fee-for-service claims dataset.9 However, our findings of the Will Rogers phenomenon in the prevalence rate of HF had not been presented by any previous literature.
In terms of marked decline in TAVR 30-day postoperative mortality and LOS over time as well as the steady rise in TAVR discharge back home, we think that these findings can be largely explained by the evolution of TAVR technique and devices along with the accumulated experience of operators and cardiac centers. Our analysis also suggests the continuous improvement in risk-adjusted 30-day postoperative mortality rate for TAVR surgery. Further research is needed to assess long-term outcomes in TAVR/SAVR patients.
Study limitations. This study has several limitations. First, the study period did not cover the year of 2019 and afterward. This was due to the unavailability of Vital Statistics data, which require extra time in data preparation during the COVID pandemic. As a result, we were unable to present the latest population-based AVR/TAVR/SAVR utilization trends after the FDA’s 2019 approval of TAVR use for low-risk AS patients in this study. Second, our study focused on utilization trends and short-term outcomes in NYS residents who received isolated TAVR/SAVR in NYS. Caution is needed in interpreting our findings as patients from other states or regions might have different regulatory policies and/or practice environments for TAVR diffusion. Third, we did not directly evaluate certain specific subgroups such as patients with bicuspid aortic valves who are likely to receive SAVR at an earlier age. Given that the population-based bicuspid aortic valve prevalence rate is approximately 1%, we do not think that it will significantly affect our major findings.16
Conclusion
TAVR and AVR experienced massive growth, whereas SAVR decreased in NYS. Significant differences in demographics and baseline characteristics between TAVR and SAVR patients existed. Younger patients and Medicare managed-care patients received TAVR (vs SAVR) a lower percentage of the time. TAVR was associated with continuous improvement in 30-day postoperative mortality.
Acknowledgments. The authors thank Kimberly S. Cozzens, the NYS cardiac service program team, and the cardiac surgery departments of the participating hospitals for their efforts to ensure the timeliness, completeness, and accuracy of the registry data.
Affiliations and Disclosures
From the Department of Health Policy, Management, and Behavior, School of Public Health, University at Albany – State University of New York, Rensselaer, New York.
Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Qian reports stock in Coinbase and options in AiDANT Intelligent Technology (Canada) and AUM Biosciences (Singapore). The remaining authors report no conflicts of interest regarding the content herein.
Manuscript accepted July 26, 2022.
Address for correspondence: Dr Feng Qian, Department of Health Policy, Management, and Behavior, School of Public Health, University at Albany – State University of New York, Room 169, GEC Building, 1 University Pl, Rensselaer, NY 12144. Email: fqian@albany.edu. Twitter: @UAlbanySPH
References
1. Osnabrugge RL, Mylotte D, Head SJ, et al. Aortic stenosis in the elderly: disease prevalence and number of candidates for transcatheter aortic valve replacement: a meta-analysis and modeling study. J Am Coll Cardiol. 2013;62(11):1002-1012. Epub 2013 May 30. doi:10.1016/j.jacc.2013.05.015
2. Virani SS, Alonso A, Aparicio HJ, et al. Heart disease and stroke statistics-2021 update: a report from the American Heart Association. Circulation. 2021;143(8):e254-e743. Epub 2021 Jan 27. doi:10.1161/CIR.0000000000000950
3. Otto CM, Nishimura RA, Bonow RO, et al. 2020 ACC/AHA Guideline for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation. 2021;143:e72-e227. Epub 2020 Dec 17. doi:10.1161/CIR.0000000000000923
4. Valle JA, Li Z, Kosinski AS, et al. Dissemination of transcatheter aortic valve replacement in the United States. J Am Coll Cardiol. 2021;78(8):794-806. doi:10.1016/j.jacc.2021.06.028
5. Vemulapalli S, Thourani VH. Aortic valve replacement and patient-centered implementation: to boldly go where no device has gone before. J Am Coll Cardiol. 2021;78(22):2173-2176. doi:10.1016/j.jacc.2021.09.856
6. Reinohl J, Kaier K, Reinecke H, et al. Effect of availability of transcatheter aortic-valve replacement on clinical practice. N Engl J Med. 2015;373(25):2438-2447. doi:10.1056/NEJMoa1500893
7. Culler SD, Cohen DJ, Brown PP, et al. Trends in aortic valve replacement procedures between 2009 and 2015: has transcatheter aortic valve replacement made a difference? Ann Thorac Surg. 2018;105(4):1137-1143. Epub 2018 Feb 1. doi:10.1016/j.athoracsur.2017.10.057
8. Kundi H, Cohen DJ, Strom JB, et al. Trends in isolated aortic valve replacement in the United States in the early phase of expansion of TAVR. Int J Cardiol. 2019;292:68-72. Epub 2019 Jun 22. doi:10.1016/j.ijcard.2019.06.061
9. Mori M, Gupta A, Wang Y, et al. Trends in transcatheter and surgical aortic valve replacement among older adults in the United States. J Am Coll Cardiol. 2021;78(22):2161-2172. doi:10.1016/j.jacc.2021.09.855
10. Hannan EL, Cozzens K, King SB 3rd, Walford G, Shah NR. The New York State cardiac registries: history, contributions, limitations, and lessons for future efforts to assess and publicly report healthcare outcomes. J Am Coll Cardiol. 2012;59(25):2309-2316. doi:10.1016/j.jacc.2011.12.051
11. Salemi A, Sedrakyan A, Mao J, et al. Individual operator experience and outcomes in transcatheter aortic valve replacement. JACC Cardiovasc Interv. 2019;12(1):90-97. Epub 2018 Dec 12. doi:10.1016/j.jcin.2018.10.030
12. Feinstein AR, Sosin DM, Wells CK. The Will Rogers phenomenon. Stage migration and new diagnostic techniques as a source of misleading statistics for survival in cancer. N Engl J Med. 1985;312(25):1604-1608. doi:10.1056/NEJM198506203122504
13. Carroll JD, Mack MJ, Vemulapalli S, et al. STS-ACC TVT Registry of transcatheter aortic valve replacement. J Am Coll Cardiol. 2020;76(21):2492-2516. doi:10.1016/j.jacc.2020.09.595
14. Simpson TF, Kheiri B, Chadderdon S, et al. TAVR operator volumes, trends, and geographic variations amongst Medicare beneficiaries in the United States. Catheter Cardiovasc Interv. 2022;99(4):1181-1185. Epub 2022 Feb 21. doi:10.1002/ccd.30134
15. Sedrakyan A, Dhruva SS, Sun TY, Mao JL, Gaudino MFL, Redberg RF. Trends in use of transcatheter aortic valve replacement by age. JAMA. 2018;320(6):598-600. doi:10.1001/jama.2018.9938
16. Masri A, Svensson LG, Griffin BP, Desai MY. Contemporary natural history of bicuspid aortic valve disease: a systematic review. Heart. 2017;103(17):1323-1330. Epub 2017 May 10. doi:10.1136/heartjnl-2016-309916