Prediction models for the occurrence and mortality of sepsis-associated lung injury: a systematic review and meta-analysis - Takeaways - MDSpire

Prediction models for the occurrence and mortality of sepsis-associated lung injury: a systematic review and meta-analysis

  • By

  • Chen Liu

  • Jian Huo

  • Yan-Song Li

  • An-Min Hu

  • Ting-Ting Ao

  • June 9, 2026

  • 0 min

Share

  • 1

    Sepsis is a life-threatening syndrome that often leads to multiple organ dysfunction, particularly affecting the lungs.

  • 2

    Current predictive models for sepsis-related lung injury and mortality show moderate discrimination but are limited by bias and low certainty.

  • 3

    The review included nine studies, primarily from China, assessing 68 model phase units for ARDS occurrence and short-term mortality.

  • 4

    Pooled test-phase AUC for ARDS occurrence was 0.749, while short-term mortality AUCs were 0.800 in training and 0.815 in testing.

  • 5

    Future predictive models for sepsis-associated lung injury should be developed and validated separately for ARDS and mortality outcomes.

Original Source(s)

Prediction models for the occurrence and mortality of sepsis-associated lung injury: a systematic review and meta-analysis

  • by Chen Liu, Jian Huo, Yan-Song Li, An-Min Hu, Ting-Ting Ao

  • June 9, 2026

Related Content

Effect modification by chronic obstructive pulmonary disease, anion gap, and serum creatinine on the association between invasive mechanical ventilation and 28-day mortality in intensive care unit sepsis patients: a retrospective cohort study

Citrullinated Histone H3 and Procalcitonin as Biomarkers for Diagnosing and Predicting Outcomes in Septic Shock

Artificial intelligence based predictive models for early sepsis detection in intensive care units: a scoping review