Development and Internal Validation of a Nomogram Model to Predict Invasive Pulmonary Aspergillosis Occurrence Risk...

Area of Science:

• Critical Care Medicine
• Infectious Diseases
• Pulmonology

Background:

• Sepsis-induced immune dysregulation increases susceptibility to secondary infections like invasive pulmonary aspergillosis (IPA).
• IPA is a significant complication in intensive care unit (ICU) patients with sepsis, associated with adverse outcomes.
• Early identification of IPA predictors is crucial for risk assessment and timely intervention.

Purpose of the Study:

• To develop and internally validate a prediction model for secondary IPA in ICU patients with sepsis.
• To identify independent predictors associated with IPA occurrence in this patient population.
• To facilitate early risk stratification and management of sepsis patients at risk for IPA.

Main Methods:

• A retrospective cohort study of 574 adult ICU patients with sepsis was conducted.
• Patients were divided into training (n=459) and validation (n=115) sets.
• Least Absolute Shrinkage and Selection Operator (LASSO) regression and multivariable logistic regression were used to develop a nomogram-based prediction model.

Main Results:

• 14.98% of sepsis patients developed IPA.
• Independent predictors identified: smoking history, corticosteroid therapy, SFTS/COVID-19, mechanical ventilation duration, NRS 2002 score, and SOFA score.
• The nomogram demonstrated good discrimination (AUC training: 0.883, validation: 0.844) and calibration, with clinical utility shown by Decision Curve Analysis (DCA).

Conclusions:

• A nomogram was developed and validated for predicting IPA risk in ICU sepsis patients.
• Nutritional risk (NRS 2002 score) emerged as a potential contributor to IPA risk.
• External validation is recommended prior to widespread clinical implementation.