Abstract
Objectives
We examined the utility of quality indicators (QIs) for transcatheter aortic valve implantation (TAVI) in the assessment of TAVI care quality and variation in practice.
Design
We performed a retrospective population-level cohort study in Ontario, Canada, where all residents receive publicly funded universal medical care. We examined the association between QI attainment and outcomes using multivariable hierarchical logistic models. We used median ORs to understand if variation in clinical outcomes between hospitals was attributable to variation in QI attainment.
Setting
We used all-comer registry data from the provincial CorHealth registry in Ontario, with linkage to administrative datasets using unique patient encoded identifiers.
Participants
TAVI recipients between 2018 and 2023 in Ontario, Canada.
Intervention
We derived a unique set of QIs from internationally agreed ones.
Primary and secondary outcome measures
The primary endpoint was a composite of all-cause mortality or rehospitalisation at 1 year from the date of TAVI.
Results
Data from 9748 TAVI procedures were included between 2018 and 2023. We identified five feasible QIs, the majority of which had high compliance and minimal variation; the lone exception was the performance of transfemoral TAVI without general anaesthesia (median 0.87; IQR 0.78–0.93). Adherence to QIs was associated with a reduction in the composite endpoint. The strongest association was observed with multidisciplinary heart team involvement (defined as the presence of an interventional cardiologist and a cardiac surgeon) in the TAVI procedure (OR 0.67, 95% CI 0.50 to 0.90, p=0.007), the performance of transfemoral TAVI without general anaesthesia (OR 0.80, 95% CI 0.71 to 0.91, p
Conclusions
We developed a feasible set of QIs for TAVI and validated these QIs using routinely collected data from a large all-comer registry in Ontario. We have identified overall high quality of care for TAVI, but with some variation in practice, which in part is attributable to differences in patient factors. Our work suggests that QIs can inform quality improvement efforts by prompting future work into discretionary versus non-discretionary variation.