Transcatheter aortic valve implantation (TAVI) has transformed the management of severe aortic stenosis, particularly as the procedure has expanded beyond patients at high surgical risk into younger and lower-risk populations. With this growth comes an increasing need to measure whether patients receive consistent, evidence-based care and whether differences between hospitals represent meaningful opportunities for quality improvement. Aktaa and colleagues examined this question using a large population-based cohort from Ontario, Canada, evaluating whether internationally developed TAVI quality indicators (QIs) can be measured with routinely collected healthcare data and whether adherence to those indicators is associated with improved clinical outcomes.
The investigators conducted a retrospective observational cohort study using Ontario’s CorHealth clinical registry linked with provincial administrative databases. The analysis included patients undergoing TAVI for aortic stenosis between April 2018 and March 2023. Because reporting TAVI procedures to CorHealth is required for provincial funding, the registry provides broad real-world coverage of procedures performed in Ontario. A total of 9,748 TAVI procedures across 11 hospitals were included. The median patient age was 82 years, 42.8% were women, and approximately three-quarters had a low frailty score.
The researchers began with quality indicators established by the European Society of Cardiology (ESC) and American College of Cardiology/American Heart Association (ACC/AHA). They assessed which indicators could realistically be measured using available registry and administrative data. Only five were sufficiently feasible: procedural heart-team involvement, gated cardiac CT before TAVI, transfemoral vascular access, avoidance of general anesthesia during transfemoral TAVI, and transthoracic echocardiography after TAVI. These measures were also combined into an opportunity-based composite quality indicator.
Overall adherence suggested a high standard of TAVI care in Ontario. Provincial attainment ranged from 81% for transfemoral TAVI without general anesthesia to 98% for procedural multidisciplinary heart-team involvement. Gated cardiac CT was performed in 94%, transfemoral access was used in 95%, and post-TAVI transthoracic echocardiography was achieved in 96%. The article’s Figure 1 on page 4 provides a visual summary of these high provincial attainment rates.
Despite generally high adherence, meaningful variation existed for some measures. The greatest hospital-level difference involved performing transfemoral TAVI without general anesthesia, which ranged from approximately 32% to 93% between hospitals. There was also substantial variation in the composite measure: the proportion of eligible patients receiving all five quality indicators ranged from approximately 20% to 90% across centers. Figure 2 on page 5 illustrates these hospital-level differences particularly clearly.
Clinical outcomes were then examined. At one year after TAVI, overall mortality was 10.2%, while rehospitalization occurred in 37.2% of patients. Thirty-day bleeding or stroke complications occurred in 5.1%, and mean hospital stay was 5.4 days. In general, patients receiving care consistent with the measured quality indicators experienced more favorable outcomes.
After adjustment for measured baseline characteristics, procedural heart-team involvement showed the strongest association with a lower risk of the combined endpoint of death or rehospitalization at one year, with an odds ratio of 0.67 (95% CI 0.50–0.90). Transfemoral TAVI without general anesthesia was also associated with better outcomes, with an odds ratio of 0.80 (95% CI 0.71–0.91). Transfemoral access produced an odds ratio of 0.84, although its confidence interval crossed 1.0. Figure 3 on page 5 graphically summarizes these adjusted associations.
The mortality analysis further supported an association between quality-indicator attainment and survival. Transfemoral access was associated with an adjusted hazard ratio for one-year mortality of 0.56, while transfemoral TAVI without general anesthesia had a hazard ratio of 0.66 and procedural heart-team involvement had a hazard ratio of 0.67. The composite indicator also showed an apparent dose-response relationship: compared with patients achieving all five eligible QIs, those achieving three or fewer had approximately twice the risk of death at one year.
These associations should not be interpreted as proof that the quality indicators themselves caused improved survival. This is one of the study’s most important findings. The researchers performed falsification analyses using outcomes such as pneumonia and gastrointestinal bleeding that would not necessarily be expected to result directly from these TAVI care processes. Several quality indicators were nevertheless significantly associated with these outcomes. This finding indicates substantial residual confounding: patients who receive certain procedural approaches may systematically differ from patients who do not.
The analysis of hospital variation reinforced that concern. Once baseline patient characteristics were incorporated into the statistical models, much of the apparent between-hospital variation disappeared. Adding individual quality indicators produced relatively little additional reduction. Therefore, differences in TAVI outcomes among Ontario hospitals could not simply be attributed to differences in QI attainment. Patient selection, comorbidities, anatomical characteristics, frailty, and other measured or unmeasured factors may influence both treatment decisions and outcomes.
This distinction has important implications for healthcare policy. The authors caution against using these administrative-data-derived quality indicators simply to rank hospitals or determine pay-for-performance incentives. A hospital performing fewer TAVI procedures without general anesthesia, for example, is not necessarily providing inferior care. Some variation may represent appropriate clinical decision-making based on patients’ anatomy, health status, procedural complexity, or other non-discretionary factors.
Instead, the indicators are best viewed as signals for further investigation. When substantial variation appears, health systems can determine whether it reflects unavoidable patient differences or discretionary differences in clinical practice. The latter may identify genuine opportunities for quality improvement.
The study demonstrates that large clinical registries can support practical measurement of TAVI quality across an entire healthcare system. At the same time, only five established ESC and ACC/AHA indicators could be reliably measured with the available datasets, illustrating important limitations of routinely collected administrative information. The observational study design also prevents causal conclusions.
Overall, Ontario demonstrated high adherence to measurable standards of TAVI care, with some important practice variation between hospitals. Better QI attainment correlated with favorable clinical outcomes, but the falsification analysis shows that these relationships are substantially influenced by residual confounding. The study therefore supports quality indicators as tools for identifying areas worthy of investigation rather than simplistic hospital scorecards. As TAVI continues to expand, integrating high-quality registry data with thoughtful interpretation of patient-level factors may provide a more meaningful path toward improving TAVI quality, consistency, and outcomes.
