# Development and Evaluation of an Innovative Inflammatory Prognostic Score for Predicting Long‐Term Mortality in Patients With Pulmonary Embolism

## TL;DR

This study created a new inflammatory score to predict long-term death risk in patients with pulmonary embolism, improving on existing methods.

## Contribution

The novel inflammatory prognostic score (IPS) combines five biomarkers to enhance mortality prediction in acute pulmonary embolism patients.

## Key findings

- The IPS significantly predicted all-cause mortality with an adjusted hazard ratio of 2.55.
- IPS improved the 3-year AUC from 0.820 to 0.886 when added to clinical models.
- Random survival forest analysis confirmed IPS as the top inflammatory predictor of mortality.

## Abstract

Systemic inflammation is closely associated with adverse outcomes in pulmonary embolism (PE). This study aimed to develop and evaluate a novel inflammatory prognostic score (IPS) derived from multiple inflammation‐related biomarkers to predict long‐term all‐cause mortality in patients with acute PE.

This retrospective cohort study included 1642 patients with confirmed acute PE admitted between January 2016 and January 2024 at the First Affiliated Hospital of Ningbo University and the Affiliated People Hospital of Ningbo University. The primary outcome was all‐cause mortality. Follow‐up was conducted in January 2025. Fifteen inflammatory biomarkers were analyzed, including C‐reactive protein (CRP), white blood cell count (WBC), neutrophil count (NEU), lymphocyte count (LYM), monocyte count (MON), red cell distribution width (RDW), platelet count (PLT), and eight derived indices: neutrophil‐to‐lymphocyte ratio (NLR), derived NLR (dNLR), monocyte‐to‐lymphocyte ratio (MLR), neutrophil–MLR (NMLR), platelet‐to‐lymphocyte ratio (PLR), systemic immune‐inflammation index (SII), systemic inflammation response index (SIRI), and inflammatory burden index (IBI). IPS was constructed using least absolute shrinkage and selection operator (LASSO) Cox regression. Prognostic performance was assessed using Kaplan–Meier survival analysis, multivariable Cox regression, time‐dependent receiver operating characteristic (ROC) analysis, and random survival forest (RSF).

During a median follow‐up of 43.87 months, 262 patients (16.0%) died. The IPS was calculated as: IPS = 0.395 × IBI + 0.236 × RDW + 0.208 × SIRI + 0.115 × MLR + 0.040 × CRP. Patients with high IPS (≥0.55) had a significantly higher risk of all‐cause mortality compared to those with low IPS (adjusted hazard ratio [HR] = 2.55, 95% confidence intervals [CI]: 1.92–3.38; p < 0.001). Time‐dependent area under the curve (AUC) for IPS were 0.756, 0.768, and 0.773 at 1, 3, and 5 years, respectively. When added to the baseline clinical model—including age, length of hospital stay, deep vein thrombosis, diastolic blood pressure, body mass index (BMI), lactate, log‐transformed serum creatinine, blood urea nitrogen (BUN), log‐transformed D‐dimer, left ventricular ejection fraction (LVEF), pulmonary artery systolic pressure (PASP), and computed tomography pulmonary angiography (CTPA)‐based embolism location (main pulmonary artery)—IPS improved the AUC from 0.820 to 0.886 at 3 years (p < 0.001, DeLong test). RSF analysis further identified IPS as the most informative inflammatory predictor of mortality.

The IPS, derived from five routinely available inflammatory biomarkers, was independently associated with long‐term mortality and significantly enhanced risk stratification beyond traditional clinical predictors in patients with acute PE. This score may support early prognostic assessment and individualized management.