The Choice of Mitral Valve Surgery Type and Mid-term Outcomes in Patients up to 70-years: Results of the AUTHEARTVISIT Study

This nationwide, registry-based, retrospective cohort study is part of the AUTHEART VISIT project, which investigates long-term outcomes of cardiac surgery in Austria. The present analysis focuses on patients aged 70 years or younger undergoing surgical mitral valve replacement (MVR) between January 1, 2010 and December 31, 2020. The aim is to evaluate differences in survival, reoperation rates, and cardiovascular outcomes between patients receiving mechanical mitral valve replacement (SMMVR) and those receiving biological mitral valve replacement (SBMVR).

Data source and registry procedures:

Data were obtained from the Austrian Health Insurance Funds, which cover approximately 98% of the Austrian population. The registry includes detailed patient-level information on diagnoses, procedures, and outcomes based on mandatory coding for reimbursement (ICD-10 for diagnoses and MEL codes for surgical procedures). The data are anonymized prior to analysis.

Quality assurance:

Data integrity is ensured by several mechanisms:

Data validation: Automatic data checks are performed to confirm plausibility, completeness, and consistency with predefined coding rules.

Source verification: Information is cross-checked against external records where available (e.g., insurance billing and hospital discharge data).

Data dictionary: A standardized coding framework is maintained, including MEL codes for procedures and ICD-10 for diagnoses. Definitions of outcomes are harmonized with prior AUTHEART VISIT studies.

Standard operating procedures: Established registry protocols govern data management, statistical analysis, and reporting.

Sample size and follow-up:

The study population includes all eligible patients undergoing mitral valve surgery in Austria during the study period (N=3520), with 413 undergoing sM-MVR, 487 undergoing sB-MVR, and 2620 undergoing mitral valve repair (MVRe). Median follow-up times were 6.6 years for SMMVR, 7.4 years for SBMVR, and 5.8 years for MVRe.

Handling of missing data:

Cases with implausible entries (e.g., death date before index surgery) are excluded. Missing values in baseline covariates are addressed through exclusion or handled in the propensity score weighting model. Outcome variables are based on administrative coding, minimizing missingness.

Statistical analysis plan:

The primary outcome is all-cause mortality. Secondary outcomes include reoperation, reoperation-free survival, major adverse cardiac events (MACE), myocardial infarction, heart failure, stroke, and bleeding. Time-to-event analyses are conducted using Cox proportional hazards models adjusted for age, sex, comorbidities, and concomitant procedures. To account for potential imbalances between groups, inverse probability of treatment weighting (IPTW) is applied based on a multinomial logistic regression model. Weighted and unweighted Kaplan-Meier curves are generated, with subgroup analyses by age. Cause-specific hazard models are used for secondary outcomes. Proportional hazards assumptions are evaluated using Schoenfeld residuals. A p-value <0.05 is considered statistically significant.

This registry-based approach provides comprehensive, real-world evidence on prosthesis choice in surgical MVR, with rigorous quality assurance procedures to ensure reliability and generalizability of results.

Inclusion criteria

• Mitral valve surgery between 01.01.2010 and 31.12.2020

Exclusion criteria

• concomitant procedures on other valves