Can large language models be privacy preserving and fair medical coders?
This addresses privacy-fairness trade-offs in healthcare AI deployment, though it is incremental as it analyzes existing methods on known datasets.
The study examined how differential privacy affects large language models for medical coding, finding that privacy-preserving models reduced micro F1 scores by over 40% on top labels and increased gender recall gaps by over 3%.
Protecting patient data privacy is a critical concern when deploying machine learning algorithms in healthcare. Differential privacy (DP) is a common method for preserving privacy in such settings and, in this work, we examine two key trade-offs in applying DP to the NLP task of medical coding (ICD classification). Regarding the privacy-utility trade-off, we observe a significant performance drop in the privacy preserving models, with more than a 40% reduction in micro F1 scores on the top 50 labels in the MIMIC-III dataset. From the perspective of the privacy-fairness trade-off, we also observe an increase of over 3% in the recall gap between male and female patients in the DP models. Further understanding these trade-offs will help towards the challenges of real-world deployment.