Donor ventilation parameters as predictors for length of mechanical ventilation after lung transplantation: Results of a prospective multicenter study.

Background: The evaluation of donor lungs heavily depends on the subjective judgment of the retrieval surgeon. As a consequence, acceptance rates vary significantly among transplant centers. We aimed to determine donor ventilation parameters in a prospective study and test if they could be used as objective quality criteria during organ retrieval.

Methods: A prospective evaluation of lung donors was performed in 3 transplant centers. Ventilation parameters were collected at the time of retrieval using a standardized ventilation protocol. Recipient length of mechanical ventilation (LMV) was defined as the primary end point, and collected data was used to build linear models predicting LMV.

Results: In total, 166 donors were included in this study. Median LMV after transplantation was 32 hours (interquartile range: 20-63 hours). Peak inspiratory pressure and dynamic compliance (C) at the time of retrieval, but not the partial pressure of oxygen/fraction of inspired oxygen (P/F) ratio, correlated with recipient LMV in Spearman correlations (r = 0.280, p = 0.002; r = -0.245, p = 0.003; and r = 0.064, p = 0.432, respectively). Linear models were built to further evaluate the impact of donor ventilation parameters on LMV. The first model was based on donor P/F ratio, donor age, donor intubation time, donor smoking history, donor partial pressure of carbon dioxide, aspiration, chest trauma, and pathologic chest X-ray. This model performed poorly (multiple R-squared = 0.063). In a second model, donor ventilation parameters were included, and C was identified as the strongest predictor for LMV. The third model was extended by recipient factors, which significantly improved the robustness of the model (multiple R-squared = 0.293).

Conclusion: In this prospective evaluation of donor lung parameters, currently used donor quality criteria poorly predicted recipient LMV. Our data suggest that C is a strong donor-bound parameter to predict short-term graft performance; however, recipient factors are similarly relevant.