LUNDON

LUNDON predicts donor lung acceptability for pulmonary transplantation using a multivariable model developed from Scientific Registry of Transplant Recipients (SRTR) donor data (2006–2019).

Key Features:

• Input variables: Age, maximum creatinine, arterial partial pressure of oxygen to fraction of inspired oxygen ratio (PaO2/FiO2), mechanism of death (including asphyxiation or drowning), cigarette use ≥20 pack-years, history of myocardial infarction, chest x-ray appearance, bloodstream infection, and cardiac arrest after brain death are used as predictors.
• Model type: Multivariable prediction model producing the LUNDON score to estimate probability of lung acceptance.
• Training cohort: Developed from a retrospective cohort of 83,219 brain-dead donors in the SRTR (2006–2019) randomly split into derivation (n = 58,314; 70%) and validation (n = 24,905; 30%) cohorts.
• Outcome baseline: Overall lung acceptance rate in the study cohort was 27.3% (22,767 of 83,219 donors).
• Performance: Model discrimination in the validation cohort yielded a C statistic of 0.891 (95% confidence interval, 0.886–0.895).

Scientific Applications:

• Donor acceptance prediction: Provides a predicted probability of donor lung acceptance to inform organ selection decisions.
• Recipient outcome association: LUNDON score is associated with recipient survival, particularly among recipients with high lung allocation scores.
• Standardization of evaluation: Enables standardized lung donor assessment to support efforts to improve lung utilization rates and optimize outcomes for pulmonary transplantation candidates.

Methodology:

Multivariable prediction model developed from a retrospective SRTR donor cohort (2006–2019) using a random 70% derivation and 30% validation split and evaluated by C statistic (reported 0.891, 95% CI 0.886–0.895).