Infective Endocarditis Mortality Prediction Nomogram Data

A 2026 study by Zhaojun Yu developed a nomogram for predicting 1-year mortality in 383 patients with infective endocarditis, using a 7:3 training-validation split. The model identified five independent prognostic factors and achieved a C-index of 0.879 in training and an AUC of 0.939 in validation.

Use Cases

• Validate the published nomogram's performance using the reported age, heart_failure, embolic_events, vegetation_diameter, and surgical_treatment factors.
• Replicate the multivariate Cox regression analysis to identify independent prognostic factors from the clinical data of 383 patients.
• Assess model calibration by comparing predicted 1-year mortality probabilities against observed outcomes as in the study's calibration curves.

Strengths

• Model demonstrated high discriminative ability with a validation cohort AUC of 0.939.
• Analysis is based on clinical data from 383 confirmed infective endocarditis patients.
• Study provides hazard ratios and confidence intervals for five identified independent prognostic factors.

Limitations

• The dataset is very small at 14.9 KB, likely containing only summary statistics or the nomogram itself, not the underlying patient-level data.
• Sample size of 383 patients limits the generalizability of the predictive model.
• Temporal coverage and specific patient geography are not stated, potentially introducing bias.

Provenance

Source

Zhaojun Yu via figshare.

Collection Method

Retrospective analysis of clinical data from patients with confirmed infective endocarditis.

Freshness

Last updated March 2026.