A prognostic nomogram for 1-year mortality in ICU patients with heart failure and comorbid depression

Abstract

Heart failure (HF) is a common condition in the intensive care unit (ICU) and significantly impacts patients’ physical and mental health, with depression affecting up to 30% of this population. This study aims to analyze clinical characteristics and prognostic factors for HF patients with comorbid depression and to establish a multivariable nomogram for prognosis. Data of 957 eligible patients diagnosed with HF complicated by depression were extracted from the Medical Information Mart for Intensive Care-IV database. The least absolute shrinkage and selection operator (LASSO) regression was performed for variable screening, followed by multivariate Cox proportional hazards regression analysis. A prognostic nomogram was subsequently constructed, and its predictive performance was comprehensively evaluated from three perspectives: discrimination, calibration, and decision curve analysis (DCA). Finally, a web-based dynamic nomogram calculator was developed to facilitate rapid individualized prognostic risk assessment for mortality. Among the 957 enrolled patients, 351 (36.7%) died within one year. Nine independent predictors were screened out via LASSO regression. Following confirmation of the proportional hazards assumption and assessment of multicollinearity, a visualized nomogram was established. The nomogram achieved an area under the curve of 0.77 (95% confidence interval 0.74–0.80), exhibited satisfactory calibration, and demonstrated a positive net clinical benefit across a wide range of threshold probabilities (0.1–0.7) on DCA. Additionally, an online dynamic calculator was developed to enable real-time individualized risk estimation at the bedside. This study established a nomogram incorporating nine routine admission indicators to predict the 1-year all-cause mortality risk among ICU patients with HF complicated by depression. The supporting online dynamic prediction calculator allows rapid bedside individualized risk stratification and provides evidence to support individualized clinical decision-making.