The Limits of Comorbidity Indices in TAVI: How Aggregation Dilutes the Prognostic Power of Individual Biomarkers

Objective. Heart-team decisions in transcatheter aortic valve implantation (TAVI) lean on EuroSCORE II, yet the comorbidity burden that often drives the final choice is rarely scored. We tested a specific two-part hypothesis: whether formally quantifying comorbidity burden (via the Charlson Comorbidity Index [CCI] and a modified CCI [mCCI]) recovers prognostic value beyond EuroSCORE II for Valve Academic Research Consortium-3 (VARC-3) outcomes, and whether aggregating comorbidities preserves or dilutes the prognostic signal of individually informative predictors. Methods. In 234 consecutive patients (mean age 76.6 ± 6.1 years) undergoing TAVI for severe aortic stenosis, EuroSCORE II, CCI, and mCCI were computed before the procedure. The primary endpoint was the 30-day VARC-3 composite; secondary endpoints were individual VARC-3 outcomes and one-year mortality. We compared discrimination (AUC, DeLong test) and incremental value with bootstrapping, and analyzed survival by Kaplan–Meier/Cox regression. Results. All three scores discriminated the 30-day composite (63 patients, 26.9%) poorly (AUCs 0.50–0.54), collapsing to chance after correction; neither comorbidity index improved on EuroSCORE II (all p > 0.40). EuroSCORE II’s only meaningful signal was stage 3 acute kidney injury (AKI; AUC 0.676), where it significantly exceeded mCCI (p = 0.048). Serum albumin and chronic kidney disease (HR 2.97 for one-year mortality) were informative individually but not within an aggregate score. Conclusions. In elderly TAVI, the actionable prognostic information resided in individual markers of organ reserve—serum albumin and chronic kidney disease—that count-based comorbidity aggregation dissipated rather than concentrated. EuroSCORE II retained a focused, mechanistically coherent association with stage 3 AKI, reflecting its renal components. These findings support a shift from disease-counting toward direct measurement of organ reserve and frailty in TAVI risk assessment, and the development of TAVI-specific tools that preserve dominant individual predictors rather than averaging them into a single score.