Lung Injury Prediction Model in Bone Marrow Transplantation: A Multicenter Cohort Study.

Area of Science:

• Hematology
• Pulmonology
• Critical Care Medicine

Background:

• Pulmonary complications are a major cause of mortality after hematopoietic stem cell transplantation (HCT).
• Early identification of high-risk patients is crucial for managing post-transplantation respiratory failure.
• Existing risk stratification methods for post-HCT respiratory complications require refinement.

Purpose of the Study:

• To develop and validate a predictive model for acute respiratory failure following HCT.
• To identify key pre-transplantation and early post-transplantation predictors of respiratory complications.
• To improve risk stratification for patients undergoing HCT.

Main Methods:

• A predictive model was developed using data from a derivation cohort (Rochester, MN) of 1,718 HCT hospitalizations.
• The model incorporated 11 pre-HCT predictors and 11 early in-hospital (<24h) predictors.
• Validation was performed on a separate cohort (Scottsdale, AZ, and Jacksonville, FL) of 1,005 HCT hospitalizations with diverse characteristics.

Main Results:

• The 22-point prediction model demonstrated strong performance in predicting acute respiratory distress syndrome (ARDS) (C-statistic, 0.905) and need for ventilation (C-statistic, 0.863) in the derivation cohort.
• The model maintained good predictive accuracy in the distinct validation cohort for ARDS (C-statistic, 0.841) and need for ventilation (C-statistic, 0.872).
• The model effectively integrated pre-transplantation, post-transplantation, and early in-hospital data.

Conclusions:

• A novel, reliable prediction model for post-HCT acute respiratory failure has been developed and validated.
• This model can accurately identify patients at increased risk of developing ARDS or requiring mechanical ventilation after HCT.
• The model facilitates targeted enrollment in clinical studies and proactive management of respiratory complications.