Beyond LVEF: morbidity burden (GMA) as the main predictor of mortality and outcomes in heart failure

Abstract

Heart failure (HF) is a major contributor to mortality and remains a clinical and public health priority. Left ventricular ejection fraction (LVEF) is widely used for HF classification and therapeutic decision-making, yet its prognostic value for mortality at the population level is limited. Adjusted Morbidity Groups (GMA) provide a comprehensive measure of comorbidity and clinical complexity using real-world healthcare data. This study aims to evaluate whether GMA is a stronger predictor of all-cause mortality than LVEF in the Catalan HF population. This population-based observational study included all individuals diagnosed with HF who were alive on December 31, 2023, in Catalonia. HF was defined either by hospital admission with an HF diagnosis (ICD-10-CM codes I09.81, I11.0, I13.0, I13.2, I50.x) or, in the absence of hospitalisation, by a clinical diagnosis plus at least one dispensation of loop diuretics (ATC code C03C*). Morbidity burden and clinical complexity were assessed using Adjusted Morbidity Groups (GMA), a validated risk stratification tool based on diagnostic and pharmacy data across care settings. All-cause mortality during 2024 was analysed using Cox proportional hazards regression models. Predictive performance of GMA and LVEF was compared in the overall HF population and in subgroups according to LVEF categories: reduced, preserved, or not recently evaluated. A total of 149,703 patients with HF were identified in 2023 (median age 81 [IQR 73–88], 53.6% women). Median time since diagnosis was 3.96 years [IQR 1.6–7.6]. Socioeconomic status was low or very low in 71.1% of patients. LVEF was preserved in 20.6%, reduced in 40.2%, and not recently evaluated in 39.1%. GMA was classified as low in 3.3%, moderate in 29.8%, high in 45.5%, and very high in 21.4%. In multivariable models, independent predictors of all-cause mortality included age >65 years (HR 1.70, 95% CI 1.50–1.80), female sex (HR 0.79, 95% CI 0.77–0.82), very low socioeconomic status (HR 0.99, 95% CI 0.90–1.03), reduced LVEF (HR 1.33, 95% CI 1.27–1.38), unavailable LVEF (HR 1.20, 95% CI 1.20–1.30), primary care–based diagnosis (HR 0.80, 95% CI 0.77–0.83), and very high GMA (HR 4.99, 95% CI 4.20–5.80). The GMA score demonstrated similar predictive power for mortality than age, and was the most powerful predictor for urgent admission, heart failure readmission and healthcare costs. In this large population-based cohort of patients with heart failure, the Adjusted Morbidity Group (GMA) emerged as the strongest independent predictor of all-cause mortality, reamission for heart failure and healthcare costs in the multivariable model. These findings support the integration of morbidity-based stratification tools such as GMA into population health strategies and clinical decision-making in heart failure care.