Skip to main content

The predicted body weight equation overestimates lung sizes of female, critically ill patients: an analysis of randomized, controlled trials and real-world clinical data.

Intensive care medicine2026-05-07PubMed
Total: 76.0Rigor: 7Innovation: 8Journal: 8Clinical: 8

Summary

Across 30,516 ventilated ICU patients, PBW-standardized tidal volumes were associated with a higher absolute risk of high driving pressures in women, mediating increased 28-day mortality. At the same PBW, women had smaller anatomical and aerated lung volumes than men, challenging current PBW-based dosing and supporting driving pressure-guided personalization.

Key Findings

  • With comparable ml/kg PBW, females had a 4.2% higher absolute risk of high driving pressures (aOR 1.26; 95% CI 1.19–1.33; p<0.001).
  • Mediation analysis indicated that excess high driving pressures accounted for 8.4% of the increased 28-day mortality in females (p<0.001).
  • At the same PBW, females had lower anatomical (-343 ml) and aerated (-188 ml) lung volumes than males (both p<0.001).
  • Analysis included 30,516 ventilated ICU patients from 10 RCTs and 2 real-world cohorts.

Clinical Implications

Consider prioritizing driving pressure-guided ventilation targets and reassessing PBW-based tidal volume prescriptions in women; implement bedside monitoring to avoid driving pressure ≥15 cmH2O.

Why It Matters

This large, methodologically rigorous synthesis links a ubiquitous bedside practice (PBW-based Vt) to sex-specific lung stress and mortality mediation, highlighting a modifiable source of harm.

Limitations

  • Secondary/observational analyses limit causal inference and are prone to residual confounding
  • PBW relies on anthropometry; measurement errors and heterogeneity across ICUs may affect generalizability

Future Directions

Prospective trials of driving pressure-guided or lung-volume-informed ventilation, and development of sex-adjusted PBW or alternative scaling metrics.

Study Information

Study Type
Cohort
Research Domain
Treatment
Evidence Level
II - Large pooled cohort analyses from RCTs and observational datasets with adjusted associations
Study Design
OTHER