Article ID: CJ-20-0032
Background: Transcatheter aortic valve replacement (TAVR) has been performed more and more frequently in elderly patients with aortic stenosis. We investigated the association of in-hospital availability of TAVR on outcomes of surgical aortic valve replacement (SAVR) in the era of TAVR.
Methods and Results: We utilized data from the Japan Adult Cardiovascular Surgery Database. Between October 2013 and December 2016, 9,330 patients aged ≥80 years underwent isolated SAVR or SAVR with coronary artery bypass grafting in 557 centers in Japan. We assessed the associations of in-hospital TAVR availability with operative mortality and composite complications adjusting for each patient’s characteristics, JapanSCORE predicted the risk scores, and hospital volumes of SAVR using generalized estimation equation methods. Observed operative mortality rates were 3.4% in all centers, 2.0% in TAVR centers and 4.0% in non-TAVR centers. The multivariable analyses showed that TAVR centers had statistically significantly lower operative mortality compared with non-TAVR centers among all patients (odds ratio 0.60, 95% confidence interval 0.41–0.89, P=0.01) and among intermediate/high-risk patients (odds ratio 0.52, 95% confidence interval 0.32–0.85, P<0.01) but not among low-risk patients (odds ratio 0.82, 95% confidence interval 0.44–1.51, P=0.52).
Conclusions: In-hospital TAVR availability was associated with better outcomes of SAVR among elderly patients. This association was statistically significant among intermediate/high-risk patients but not significant among low-risk patients.
Transcatheter aortic valve replacement (TAVR) has been more frequently performed in the elderly patients with severe aortic stenosis (AS) as an alternative to surgical aortic valve replacement (SAVR). In the guidelines, TAVR is recommended as a Class I or IIa recommendation for inoperable, high-risk and intermediate-risk patients with severe AS based on several clinical trials’ results.1–6 Additionally, recent randomized clinical trials have shown non-inferiority and/or superiority of TAVR over SAVR in early clinical outcomes among low-risk patients.7,8 Expanding the indication of TAVR may lead to a reduction in SAVR volume, as seen in Germany,9 and a reduction in SAVR volume may provoke more intense quality control of SAVR because the positive association of case volume with outcomes is well known.10
In Japan, TAVR centers are certified by the official accreditation organization based on strict criteria. On the other hand, there is no specific requirement for institutions to perform SAVR or other cardiac surgical operations. In-hospital availability of TAVR may contribute to appropriate TAVR/SAVR selection for AS patients and subsequently to improved outcomes of SAVR. Also, the requirements for TAVR center certification may contribute to improved outcomes of SAVR.
We hypothesized that in-hospital availability of TAVR is associated with better early operative outcomes of SAVR independent of patients’ operative risk and institutional SAVR volume. In this study, we only included elderly patients aged ≥80 years because such patients comprise most of the TAVR candidates in Japan.
This study was approved by the Data Utilization Committee of the Japan Cardiovascular Surgery Database (JCVSD: reference numbers: A0091 and A0097). We reviewed 9,930 elderly (≥80 years) SAVR or SAVR with CABG patients reported in the JCVSD from October 2013 through December 2016 by 577 hospitals in Japan. The JCVSD is a nationwide registry of patients undergoing cardiovascular surgeries in Japan. The data collection is conducted using the National Clinical database, which is a platform for nationwide registries. Registration to JCVSD is linked to the certification for surgery as well as for sub-specialty of cardiovascular surgery in Japan. Its coverage is assumed to exceed 95% of all major adult cardiovascular surgeries in Japan.11 It also performs routine on-site audits for data accuracy. The data have been used extensively for clinical and health policy research in the past decade.12
We included patients who were 80 years or older and underwent elective isolated SAVR or SAVR with CABG procedures. We excluded patients who underwent non-CABG concomitant procedures or non-elective surgery, those with infective endocarditis, those without 30-day postoperative status data or the data required for calculating JapanSCORE predicted risk scores.
The operative mortality and composite complication rates were the primary outcomes of this study. The operative mortality was defined as the rate of death before discharge or within 30 days after surgery. The composite complication included death, stroke, reoperation for bleeding, prolonged ventilation, renal failure and deep sternal infection before discharge or within 30 days after surgery. The JapanSCORE system provided the predicted operative mortality and composite complication rates. The observed to expected (O/E) ratios were calculated by dividing the actual operative mortality by predicted mortality.
We investigated the association of in-hospital TAVR availability and these outcomes using multivariable analyses. We also conducted the subgroup analyses among low-risk patients (JapanSCORE predicted mortality <4%) and intermediate-/high-risk patients (JapanSCORE predicted mortality ≥4%).
TAVR Centers and Hospital VolumeIn Japan, TAVR centers are certified by an official accreditation organization based on strict criteria, which include minimal requirements of annual SAVR volume, number of full-time board-certified cardiovascular surgeons, and a fully equipped hybrid operating room. Certain numbers of coronary interventions, endovascular aortic repairs, board-certified interventional cardiologists, and anesthesiologists are also required for certification of TAVR centers. All TAVR center candidates are inspected in person by site visit and assessed by the accreditation organization before certification.
In Japan, TAVR was introduced in October 2013, and there were only 8 TAVR centers at that time; there were 105 TAVR centers in December 2016. Many institutions switched from non-TAVR to TAVR center during the study period. We obtained TAVR center certification dates for each TAVR center and defined each center as non-TAVR before the certification date and as a TAVR center after the certification date.
We defined hospital SAVR volume as the average annual number of isolated SAVR and SAVR with CABG, including all-ages patients, that each institution registered to the JCVSD during the study period. Hospital SAVR volume was categorized as follows: very low (<20 cases per year): low (≥20 and <50 cases per year): high (≥50 and <100 cases per year): very high (≥100 cases per year). We investigated the O/E ratios and 95% confidence intervals (CIs) for each hospital SAVR volume category.
Statistical AnalysisContinuous variables are expressed as median and 5–95th percentiles. Categorical variables are expressed as number and percentage.
We used generalized estimating equation methods with a compound symmetry correlation structure as the multivariable analyses to investigate the association of TAVR availability with outcomes with adjusting for hospital cluster effect. Each model included categorized hospital volume, patients’ ages, male sex, predicted risk, current smoking status, diabetes, hypertension, chronic obstructive pulmonary disease (moderate or severe), history of cerebrovascular disease, peripheral artery disease, previous CABG, previous PCI, previous heart valve surgery, New York Heart Association class III or IV, AS, aortic insufficiency, estimated glomerular filtration rate, impaired left ventricular function (ejection fraction <60%), chronic dialysis, and concomitant CABG as covariates. Chronic dialysis was not included in models for low-risk patients because there were few dialysis patients in the low-risk group.
All statistical analyses were performed with using SAS 9.4 (SAS institute, Cary, NC, USA). The significance level was defined as P<0.05.
Among 9,930 patients, there were 5,141 low-risk patients (51.8%) and 4,189 intermediate-/high-risk patients (48.2%); 2,484 patients (25.0%) underwent surgery at TAVR centers, and 6,846 patients (75.0%) underwent surgery at non-TAVR centers. Preoperative characteristics are summarized in Table 1. Median age was 83 years, and approximately 39% of patients were men in both the TAVR and non-TAVR center groups. Although there were some differences in preoperative risk factors, median predicted mortality scores and predicted composite complication scores were very similar between the TAVR and non-TAVR center groups.
| All patients | Low-risk patients | Intermediate-/high-risk patients | |||||||
|---|---|---|---|---|---|---|---|---|---|
| All centers (n=9,330) |
TAVR centers (n=2,484) |
Non-TAVR centers (n=6,846) |
All centers (n=5,141) |
TAVR centers (n=1,342) |
Non-TAVR centers (n=3,799) |
All centers (n=4,189) |
TAVR centers (n=1,142) |
Non-TAVR centers (n=3,047) |
|
| Age (years), median (p5–p95) |
83 (80–88) |
83 (80–88) |
83 (80–88) |
82 (80–88) |
82 (80–88) |
83 (80–88) |
83 (80–89) |
83 (80–89) |
83 (80–89) |
| Female, n (%) |
3,605 (61.4%) |
964 (61.2%) |
2,641 (61.4%) |
1,737 (66.2%) |
458 (65.9%) |
1,279 (66.3%) |
1,868 (55.4%) |
506 (55.7%) |
1,362 (55.3%) |
| Current smoking | 2,022 (21.7%) |
531 (21.4%) |
1,491 (21.8%) |
937 (18.2%) |
252 (18.8%) |
685 (18%) |
1,085 (20.9%) |
279 (24.4%) |
806 (26.5%) |
| COPD (moderate or severe) |
1,796 (19.2%) |
509 (20.5%) |
1,287 (18.8%) |
549 (10.7%) |
136 (10.1%) |
413 (10.9%) |
1,247 (29.8%) |
373 (32.7%) |
874 (28.7%) |
| Diabetes | 2,336 (25.0%) |
604 (24.3%) |
1,732 (25.3%) |
887 (17.3%) |
231 (17.2%) |
656 (17.3%) |
1,449 (34.6%) |
373 (32.7%) |
1,076 (35.3%) |
| Hypertension | 7,464 (80.0%) |
2,000 (80.5%) |
5,464 (79.8%) |
4,007 (77.9%) |
1,050 (78.2%) |
2,957 (77.8%) |
3,457 (82.5%) |
950 (83.2%) |
2,507 (82.3%) |
| Cerebrovascular disease |
807 (8.6%) |
201 (8.1%) |
606 (8.9%) |
322 (6.3%) |
71 (5.3%) |
251 (6.6%) |
485 (11.6%) |
130 (11.4%) |
355 (11.7%) |
| Peripheral artery disease |
810 (8.7%) |
216 (8.7%) |
594 (8.7%) |
152 (3.0%) |
34 (2.5%) |
118 (3.1%) |
658 (15.7%) |
182 (15.9%) |
476 (15.6%) |
| Previous CABG | 132 (1.4%) |
33 (1.3%) |
99 (1.4%) |
36 (0.7%) |
5 (0.4%) |
31 (0.8%) |
96 (2.3%) |
28 (2.5%) |
68 (2.2%) |
| Previous PCI | 978 (10.5%) |
296 (11.9%) |
682 (10.0%) |
262 (5.1%) |
63 (4.7%) |
199 (5.2%) |
716 (17.1%) |
233 (20.4%) |
483 (15.9%) |
| Preoperative eGFR, median (p5–p95) |
51.8 (8.2–84.8), missing 31 |
52.6 (8.2–84.8), missing 26 |
51.7 (8.3–84.8), missing 5 |
57.5 (33.2–87.8), missing 16 |
59.0 (34.0–88.4), missing 15 |
56.9 (32.9–87.4), missing1 |
42.1 (5.3–77.1), missing 15 |
41.9 (5.5–76.3), missing 11 |
42.2 (5.3–78.2), missing 4 |
| Dialysis | 613 (6.6%) |
181 (7.3%) |
432 (6.3%) |
4 (0.1%) |
3 (0.2%) |
1 (0.03%) |
609 (14.5%) |
178 (15.6%) |
431 (14.2%) |
| CAD requiring concomitant CABG |
2,572 (27.6%) |
750 (30.2%) |
1,822 (26.6%) |
1,081 (21%) |
317 (23.6%) |
764 (20.1%) |
1,491 (35.6%) |
433 (37.9%) |
1,058 (34.7%) |
| Previous valve operation |
166 (1.8%) |
53 (2.1%) |
113 (1.7%) |
40 (0.8%) |
15 (1.1%) |
25 (0.7%) |
126 (3%) |
38 (3.3%) |
88 (2.9%) |
| NYHA (III or IV) |
1,995 (21.4%) |
437 (17.6%) |
1,558 (22.8%) |
306 (6%) |
72 (5.4%) |
234 (6.2%) |
1,689 (40.3%) |
365 (32%) |
1,324 (43.5%) |
| Aortic stenosis | 8,300 (89.0%) |
2,211 (89.0%) |
6,089 (88.9%) |
4,585 (89.2%) |
1,196 (89.1%) |
3,389 (89.2%) |
3,715 (88.7%) |
1,015 (88.9%) |
2,700 (88.6%) |
| Aortic insufficiency | 1,642 (17.6%) |
424 (17.1%) |
1,218 (17.8%) |
790 (15.4%) |
201 (15%) |
589 (15.5%) |
852 (20.3%) |
223 (19.5%) |
629 (20.6%) |
| EF | 64.9 (37.2–78.0) |
63.8 (37–77.7) |
65 (38–78) |
66.5 (45–79) |
65.7 (46–78.7) |
67 (45–79) |
61 (32–77) |
60 (31–76) |
61.7 (32–77) |
| EF <60% | 2,964 (31.8%) |
791 (31.8%) |
2,173 (31.7%) |
1,036 (20.2%) |
256 (19.1%) |
780 (20.5%) |
1,928 (46.0%) |
535 (46.8%) |
1,393 (45.7%) |
| Predicted risk of operative death (%), median (p5–p95) |
3.6 (1.7–18.1) |
3.6 (1.7–18.1) |
3.6 (1.7–18.1) |
2.3 (1.7–3.8) |
2.4 (1.7–3.8) |
2.3 (1.7–3.8) |
7.2 (4.2–27.2) |
7.3 (4.1–25.6) |
7.2 (4.2–27.5) |
| Predicted risk of composite complication (%), median (p5–p95) |
16.1 (8.8–44.1) |
16.0 (8.8–44.3) |
16.3 (8.8–43.0) |
11.8 (8.1–19.9) |
11.8 (8.1–19.9) |
11.8 (8.1–20.0) |
25.8 (14.4–53.2) |
25.9 (14.3–52.0) |
25.8 (14.5–53.4) |
CABG, coronary artery bypass graft surgery; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease; eGFR, estimated glomerular filtration rate; NYHA, New York Heart Association; p5–p95, percentile range; PCI, percutaneous coronary intervention; TAVR, transcatheter aortic valve replacement.
The operative mortality rates were 3.5% in all patients, 1.8% in low-risk patients, and 5.6% in intermediate-/high-risk patients. In TAVR centers, the operative mortality rates were 2.0% among all patients, 1.3% among low-risk patients, and 2.9% among intermediate-/high-risk patients. In non-TAVR centers, the operative mortality rates were 4.0% among all patients, 2.0% among low-risk patients, and 6.6% among intermediate-/high-risk patients.
The composite complication rates were 12.4% in all patients, 8.3% in low-risk patients, and 17.4% in intermediate-/high-risk patients. In TAVR centers, the composite complication rates were 9.5% among all patients, 6.5% among low-risk patients, and 13.1% among intermediate-/high-risk patients. In non-TAVR centers, the composite complication rates were 13.4% among all patients, 9.0% among low-risk patients, and 19.0% among intermediate-/high-risk patients.
Hospital SAVR Volume and Operative MortalityThe O/E ratios of operative mortality and 95% CIs are plotted by hospital volume category in the Figure. Among all centers, the O/E ratio was significantly higher in the very low volume group than in the other 3 groups, and there was a decreasing trend of O/E ratio by hospital SAVR volume. Among the TAVR centers, there was no significant difference in the O/E ratio. However, as shown in the Figure, there was some decreasing trend of O/E ratio by SAVR volume. Among non-TAVR centers, there was no significant difference or decreasing trend by hospital SAVR volume.
Observed to expected ratio of operative mortality by hospital surgical aortic valve replacement volume among all centers (Left), transcatheter aortic valve replacement (TAVR) centers (Middle) and non-TAVR centers (Right). SAVR, surgical aortic valve replacement.
The results of the multivariable analyses are shown in Table 2. TAVR centers were significantly associated with lower operative mortality (odds ratio [OR] 0.60, 95% CI 0.41–0.89, P=0.01) and composite complications rate (OR 0.78, 95% CI 0.64–0.95, P=0.01) after adjusting patients’ risk factors and hospital SAVR volume. In the subgroup analysis of intermediate-/high-risk patients, TAVR centers were significantly associated with lower operative mortality (OR 0.52, 95% CI 0.32–0.85, P<0.01) and composite complications rate (OR 0.68, 95% CI 0.52–0.89, P=0.04). In the subgroup analysis of low-risk patients, TAVR centers were not significantly associated with lower operative mortality (OR 0.82, 95% CI 0.44–1.51, P=0.52) or composite complications rate (OR 0.93, 95% CI 0.70–1.25, P=0.64).
| Variables | All patients | Intermediate-/high-risk patients | Low-risk patients | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | Lower 95% CI |
Upper 95% CI |
P value | OR | Lower 95% CI |
Upper 95% CI |
P value | OR | Lower 95% CI |
Upper 95% CI |
P value | |
| Performed at TAVR center | 0.60 | 0.41 | 0.89 | 0.01 | 0.52 | 0.32 | 0.85 | <0.01 | 0.82 | 0.44 | 1.51 | 0.52 |
| Very low hospital volume category* |
1.76 | 1.08 | 2.84 | 0.02 | 1.82 | 0.87 | 3.82 | 0.11 | 1.75 | 0.87 | 3.50 | 0.11 |
| Low hospital volume category* |
1.24 | 0.79 | 1.94 | 0.36 | 1.30 | 0.64 | 2.62 | 0.47 | 1.14 | 0.61 | 2.14 | 0.67 |
| High hospital volume category* |
1.10 | 0.69 | 1.78 | 0.68 | 1.26 | 0.61 | 2.58 | 0.53 | 0.83 | 0.45 | 1.54 | 0.56 |
| Age | 1.04 | 0.99 | 1.08 | 0.09 | 1.03 | 0.98 | 1.09 | 0.22 | 1.04 | 0.97 | 1.12 | 0.26 |
| Female | 1.09 | 0.84 | 1.42 | 0.53 | 1.16 | 0.85 | 1.57 | 0.34 | 1.04 | 0.65 | 1.67 | 0.86 |
| Predicted risk of operative death |
1.04 | 1.02 | 1.06 | <0.01 | 1.05 | 1.03 | 1.06 | <0.01 | 1.35 | 0.94 | 1.95 | 0.10 |
| Current smoking | 1.24 | 0.93 | 1.65 | 0.14 | 1.23 | 0.88 | 1.71 | 0.23 | 1.22 | 0.71 | 2.09 | 0.47 |
| COPD (moderate or severe) | 1.04 | 0.77 | 1.39 | 0.80 | 0.95 | 0.70 | 1.30 | 0.76 | 0.78 | 0.36 | 1.67 | 0.52 |
| Diabetes | 1.02 | 0.77 | 1.35 | 0.89 | 0.90 | 0.66 | 1.23 | 0.51 | 1.15 | 0.68 | 1.95 | 0.61 |
| Hypertension | 0.68 | 0.51 | 0.90 | <0.01 | 0.62 | 0.44 | 0.86 | <0.01 | 0.75 | 0.48 | 1.17 | 0.20 |
| Cerebrovascular disease | 1.10 | 0.78 | 1.56 | 0.58 | 1.09 | 0.73 | 1.61 | 0.68 | 1.03 | 0.46 | 2.27 | 0.95 |
| Peripheral vascular disease | 1.09 | 0.76 | 1.57 | 0.63 | 0.92 | 0.63 | 1.35 | 0.68 | 1.55 | 0.56 | 4.30 | 0.40 |
| Previous CABG | 1.35 | 0.66 | 2.76 | 0.41 | 1.13 | 0.52 | 2.46 | 0.75 | 2.85 | 0.65 | 12.43 | 0.16 |
| Previous PCI | 1.00 | 0.71 | 1.39 | 0.98 | 0.96 | 0.67 | 1.37 | 0.82 | 0.84 | 0.33 | 2.16 | 0.72 |
| CAD requiring concomitant CABG (concomitant CABG) |
1.47 | 1.14 | 1.89 | <0.01 | 1.37 | 1.01 | 1.87 | 0.04 | 1.37 | 0.85 | 2.21 | 0.19 |
| Previous valve operation | 0.91 | 0.43 | 1.92 | 0.81 | 0.63 | 0.28 | 1.40 | 0.25 | 3.35 | 0.72 | 15.60 | 0.12 |
| NYHA (III or IV) | 1.15 | 0.85 | 1.57 | 0.37 | 0.91 | 0.65 | 1.27 | 0.56 | 1.04 | 0.44 | 2.41 | 0.94 |
| Aortic stenosis | 1.01 | 0.65 | 1.58 | 0.97 | 1.10 | 0.68 | 1.77 | 0.70 | 0.73 | 0.30 | 1.78 | 0.49 |
| Aortic insufficiency | 1.06 | 0.75 | 1.49 | 0.76 | 1.16 | 0.82 | 1.64 | 0.41 | 0.68 | 0.29 | 1.63 | 0.39 |
| Preoperative eGFR | 0.99 | 0.98 | 1.00 | <0.01 | 0.99 | 0.98 | 1.00 | 0.17 | 0.99 | 0.97 | 1.01 | 0.10 |
| EF <60% | 1.23 | 0.94 | 1.60 | 0.13 | 1.16 | 0.86 | 1.57 | 0.33 | 1.04 | 0.60 | 1.79 | 0.90 |
| Dialysis | 1.29 | 0.79 | 2.10 | 0.30 | 1.31 | 0.80 | 2.12 | 0.28 | Not applicable | |||
*Very high hospital volume category is a reference. CI, confidence interval; EF, ejection fraction; OR, odds ratio. Other abbreviations as in Table 1.
Our study showed that TAVR centers demonstrated lower operative mortality and composite complications rate of SAVR in elderly patients than did non-TAVR centers. The multivariable analyses showed that in-hospital TAVR availability was associated with a 40% reduction in operative mortality and 22% reduction in composite complications rate after adjusting patient risk factors and hospital SAVR volume. These associations were prominent in the group of intermediate-/high-risk patients, but not in the group of low-risk patients.
Brennan et al have shown that overall AVR (including SAVR and TAVR) procedural mortality has declined significantly since the introduction of TAVR, especially in high-risk SAVR patients.13 This trend was more prominent in the TAVR centers than in non-TAVR centers. Hawkins et al have compared outcomes of SAVR with or without CABG in 3 different periods (pre-TAVR era, early-TAVR era and commercial-TAVR era), and observed the lowest O/E ratios of mortality and major morbidities in the commercial TAVR era.14 These findings indicate that the introduction of TAVR has contributed to an improvement in SAVR outcomes. However, no studies have compared SAVR outcomes between TAVR and non-TAVR centers. Singh et al compared SAVR outcomes between TAVR and non-TAVR centers using a national administrative database, and their multivariable and propensity-matched analyses showed a benefit in terms of mortality and morbidity when SAVR was performed in a TAVR center.15 Our study results strengthen their findings further because we used a clinical database that included more detailed clinical information than an administrative database. Also, we adjusted hospital SAVR volume, which is an important confounding factor, and defined a TAVR center more thoroughly by identifying the date of TAVR introduction in each institution.
There are several presumable rationales for the fact that TAVR centers have demonstrated better outcomes of SAVR than non-TAVR centers. First, TAVR centers generally have more human and material resources than non-TAVR centers, which may contribute to the better surgical outcomes. Second, patient selection is a potential reason for differences in surgical outcomes. Because TAVR has demonstrated better early outcomes than SAVR and has become a primary treatment option for intermediate-/high-risk patients in many countries,1,16 not having the in-hospital TAVR option may prevent the heart team from selecting the right treatment, especially for intermediate-/high-risk patients. In our study, some very high-volume non-TAVR centers demonstrated a high O/E ratio, which might have resulted partly from patient selection in those centers. Although we adjusted many patients’ risk factors and operative risk score as covariates in the models, we did not include some other risk factors such as frailty, calcification of the ascending aorta, history of chest irradiation and chest wall deformity because that data was not available in the JCVSD. Those risk factors generally favor TAVR rather than SAVR.16 Third, TAVR centers might simply have higher surgical and perioperative management performance than non-TAVR centers.
Another important thing to be aware of is that the average O/E ratio of operative mortality is lower than 1 in this cohort (Figure), naturally because the risk model was based on past cases, and this means that Japanese cardiac surgical centers have been improving SAVR outcomes in elderly patients.
The strengths of our study include the large sample size and rich clinical information. Including hospital SAVR volume as a covariate is unique and was important to assess the association of in-hospital TAVR availability and SAVR outcomes in this study. We categorized SAVR cases performed before TAVR introduction in each TAVR center into the non-TAVR center group, which is also unique and important in the assessment of the association between in-hospital TAVR availability and SAVR outcomes in this study.
On the other hand, there are several limitations. We did not review the appropriateness of the operative indication of aortic valve disease in each case. Also, we did not review some operative risk factors such as frailty, calcification of the ascending aorta and chest wall deformity, which are potential confounding factors. As we described, however, old patients with those risk factors are supposed to be referred to TAVR. Furthermore, we did not review institutional resource data, which is also a potential confounding factor. To adjust for these factors, we included institutional SAVR volume as a covariate in our multivariable models because case volume reflects institutional human and material resources to a certain degree. Lastly, we limited the study population to patients aged 80 years or older, and this population might not be the primary candidates of SAVR once TAVR for low-risk patients is approved. We should conduct further investigation in all age groups based on the results of this study.
Our findings did not show that all non-TAVR centers had poor outcomes of SAVR. Some non-TAVR centers have demonstrated excellent outcomes. However, our findings suggest clinical benefits for elderly patients with aortic valve disease of undergoing SAVR at TAVR centers.
In conclusion, in-hospital TAVR availability was associated with lower operative mortality and composite complication rates after SAVR among elderly patients. This association was statistically significant among intermediate/high-risk patients but not significant among low-risk patients.
We express our sincere gratitude to Dr. Shinichi Takamoto and all those involved in the organization, management and data collection of JCVSD. This research was supported by the JADECOM grant to the Department of Cardiovascular Surgery at Tokyo Bay Urayasu Ichikawa Medical Center.
The deidentified participant data will not be shared.
This research was supported by the JADECOM grant to the Department of Cardiovascular Surgery at Tokyo Bay Urayasu Ichikawa Medical Center.
This study was approved by the Data Utilization Committee of the JCVSD (A0091, A0097).
