Development and validation of a nomogram for predicting long-term survival in patients with infective endocarditis: a single-center retrospective study.
To develop and internally validate a nomogram for predicting long-term survival in patients with infective endocarditis (IE), focusing on microbiological characteristics and surgical treatment.
We retrospectively analyzed 210 patients with IE. Clinical characteristics, laboratory data, echocardiographic findings, microbiological profiles, and treatment information were collected. LASSO regression was used for variable selection, followed by multivariable Cox regression to identify independent prognostic factors and construct the nomogram. Model performance was evaluated using time-dependent ROC curves, concordance index, calibration curves, and decision curve analysis. Internal validation was performed using bootstrap resampling.
Among 210 patients (67.1% male; median age, 55 years), 125 (59.5%) underwent surgery. Seven independent prognostic factors identified from 11 LASSO-selected candidates were incorporated into the nomogram. The model showed good discrimination, with AUCs of 0.872, 0.840, and 0.846 for 1-, 3-, and 5-year survival prediction, respectively. The concordance index was 0.826. Calibration and decision curve analyses demonstrated good predictive accuracy and clinical utility. Blood culture-negative IE and Gram-negative bacterial infection were associated with poorer survival, whereas surgical treatment predicted improved outcomes.
We developed and internally validated a nomogram integrating clinical, microbiological, and treatment-related variables to predict long-term survival in IE patients, which may support individualized risk assessment and prognostic stratification.
We retrospectively analyzed 210 patients with IE. Clinical characteristics, laboratory data, echocardiographic findings, microbiological profiles, and treatment information were collected. LASSO regression was used for variable selection, followed by multivariable Cox regression to identify independent prognostic factors and construct the nomogram. Model performance was evaluated using time-dependent ROC curves, concordance index, calibration curves, and decision curve analysis. Internal validation was performed using bootstrap resampling.
Among 210 patients (67.1% male; median age, 55 years), 125 (59.5%) underwent surgery. Seven independent prognostic factors identified from 11 LASSO-selected candidates were incorporated into the nomogram. The model showed good discrimination, with AUCs of 0.872, 0.840, and 0.846 for 1-, 3-, and 5-year survival prediction, respectively. The concordance index was 0.826. Calibration and decision curve analyses demonstrated good predictive accuracy and clinical utility. Blood culture-negative IE and Gram-negative bacterial infection were associated with poorer survival, whereas surgical treatment predicted improved outcomes.
We developed and internally validated a nomogram integrating clinical, microbiological, and treatment-related variables to predict long-term survival in IE patients, which may support individualized risk assessment and prognostic stratification.