Machine learning analysis of the UNOS database fails to predict lung transplant outcomes
Luo, L.; Mozejko, M.; Markov, N.; Peltekian, A.; Mohsin, S.; Carns, M.; Cooper, P.; Lysne, J.; Joudi, A.; Betensley, A.; Bemiss, B. C.; Myers, C.; Bharat, A.; Tomic, R.; Arunachalam, A.; Szczurek, E.; Budinger, G. S.; Misharin, A. V.; Subramani, M. V.
Show abstract
Lung transplantation is a life-saving therapy for end-stage lung disease but has the poorest survival among solid organ transplants. We analyzed standardized electronic health record (EHR) data from the United Network for Organ Sharing (UNOS) to predict one-, three-, and five-year survival and favorable long-term outcomes post-lung transplant. We applied two multivariate machine learning approaches, XGBoost or a tabular BERT model called EHRFormer, to data from 43,869 first-time lung transplant recipients (1987-2022). XGBoost and EHRFormer identified features that align closely with established risk factors for worse outcomes such as length of index stay, recipient age, and creatinine at the time of transplant. We developed a simple perturbation method with EHRFormer to probe in silico multivariate interactions between features that influence model prediction. Despite their attention to known risk factors, machine learning applied to EHR data collected by UNOS poorly predict one-, three-, and five-year mortality after lung transplant.
Matching journals
The top 6 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Integration of clinical characteristics, lab tests and a deep learning CT scan analysis to predict severity of hospitalized COVID-19 patients 94%
- Deep representation learning for clustering longitudinal survival data from electronic health records 93%
- GraftIQ: A Hybrid Multi-Class Neural Network Integrating Clinical Insight for Multi-Outcome Prediction in Liver Transplant Recipients 92%
Similar papers in this journal
- Predicting Post-Liver Transplant Outcomes in Patients with Acute-on-Chronic Liver Failure using Expert-Augmented Machine Learning 91%
- Tensor-Based Integration of Time-Series Measurements Reveals Relationships Between Underlying Disease, Early Injury, and CD4+ Polarization in Liver Transplantation 91%
- Th1/17 Cells Infiltrate Murine Cytomegalovirus-Infected Renal Allografts via Virus-Induced CCL20 and Promote Th1 Cells through IL-17A. 91%
Similar papers in this journal
- Mismatches in gene deletions and kidney-related proteins are novel histocompatibility factors in kidney transplantation 92%
- Changing of therapeutic trends between the 1st and 2nd wave did not reduce COVID-19 related mortality of renal transplant recipients: a national registry study 90%
- Urine single cell RNA-sequencing in focal segmental glomerulosclerosis reveals inflammatory signatures in immune cells and podocytes 90%
Similar papers in this journal
- Harnessing Deep Learning to Detect Bronchiolitis Obliterans Syndrome from Chest CT 94%
- Pretrained Patient Trajectories for Adverse Drug Event Prediction Using Common Data Model-based Electronic Health Records 92%
- Subpopulation-specific Machine Learning Prognosis for Underrepresented Patients with Double Prioritized Bias Correction 91%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.