Comorbidity burden drives hard outcomes in a cardiovascular cardiac ICU cohort
B F O Gomes, G P Dutra, H A Parente, T M B Silva, L S M Pereira, N J Kehdi, L Z Peres, L C V Lima, F C Oliveira, D V Dias, J L F Petriz, G M M OliveiraAbstract
Background
Contemporary cardiac intensive care units (CICUs) increasingly care for patients with substantial chronic disease burden. The Charlson Comorbidity Index (CCI) is a validated, weighted score summarizing comorbid conditions (e.g., renal disease, diabetes, malignancy) and serves as an objective proxy of baseline health status and physiological reserve.
Purpose
To evaluate the association between CCI and in-hospital mortality in cardiovascular CICU admissions and to describe its relationship with cardiovascular phenotype (LVEF), biomarkers, organ-support requirements, ICU readmission, and length of stay.
Methods
Retrospective single-center cohort (July 2011–February 2025). We included admissions with primary cardiovascular diagnosis (ICD-10 chapter I: I00–I99). CCI was analyzed as categories (0, 1–2, ≥3) and continuous (per point). Primary outcome was in-hospital death. Secondary outcomes included post-discharge ICU readmission during index hospitalization, use of invasive mechanical ventilation (MV), and vasopressors, and CICU/hospital length of stay (LOS). Cardiovascular phenotype/biomarkers included LVEF, peak high-sensitivity troponin, and peak creatinine. Associations were assessed using Spearman correlations and multivariable logistic regression (mortality adjusted for age, sex, and LVEF category).
Results
The analytic cohort comprised 4,850 cardiovascular admissions (median age 69 years [IQR 57–80], 60.5% male); in-hospital mortality was 4.8% (231/4,850). Mortality increased stepwise across CCI strata: 1.6% (32/2,037) for CCI=0, 4.8% (97/2,013) for CCI 1–2, and 12.8% (102/800) for CCI≥3. After adjustment for age, sex, and LVEF, CCI remained strongly associated with mortality (CCI≥3 vs 0: OR 5.11, 95% CI 3.35–7.81; CCI 1–2 vs 0: OR 2.00, 1.32–3.02; OR 1.33 per CCI point, 1.25–1.42; all p≤0.001). CCI correlated inversely with LVEF (ρ = −0.21; p<0.001) and HFrEF prevalence (LVEF<40) increased across CCI categories (5.5%, 15.8%, 21.1%). CCI correlated with hospital LOS (ρ = 0.33; p<0.001) and CICU LOS (ρ = 0.21; p<0.001); median hospital LOS was 4 (IQR 3–7), 6 (4–12), and 9 (5–20) days across CCI categories. Post-discharge ICU readmission increased with CCI (5.0%, 8.9%, 14.9%), with adjusted OR 2.98 (2.21–4.02) for CCI≥3 vs 0 (p<0.001). Higher CCI was associated with greater organ-support use (adjusted OR for CCI≥3 vs 0: vasopressors 2.95 [1.57–5.51], MV 2.23 [1.00–4.97]). Peak high-sensitivity troponin was available in 80.6% and showed higher values in patients with greater comorbidity burden (P75 6.77 in CCI≥3 vs 3.80 in CCI=0), while peak creatinine increased with CCI (median 0.91, 0.97, 1.10).
Conclusion
In a large cardiovascular CICU cohort, Charlson comorbidity burden identifies a worse cardiovascular phenotype, higher resource utilization, longer hospitalization, higher ICU readmission, and markedly higher in-hospital mortality, adding substantial prognostic information beyond age, sex, and LVEF.