FairTabGen: High-Fidelity and Fair Synthetic Health Data Generation from Limited Samples

TL;DR

FairTabGen is a novel LLM-based framework that generates high-quality, fair synthetic healthcare data from limited samples, addressing privacy concerns and reducing computational requirements in clinical research.

Contribution

It introduces a new method combining in-context learning, prompt curation, and structural constraints for efficient, fair synthetic health data generation from small datasets.

Findings

Uses 99% less data than traditional methods

Achieves 50% improvement in fairness metrics

Enhances fairness by 10% with bias mitigation techniques

Abstract

Synthetic healthcare data generation offers a promising solution to research limitations in clinical settings caused by privacy and regulatory constraints. However, current synthetic data generation approaches require specialized knowledge about training generative models and require high computational resources. In this paper, we propose FairTabGen, an LLM-based tabular data generation framework that produces high-quality synthetic healthcare data using only a small subset of the original dataset. Our method combines in-context learning, prompt curation and embedding structural constraints for data synthesis. We evaluate performance on MIMIC-IV dataset. Our method using 99% less data and achieving 50% improvement for fairness through unawareness while maintaining competitive predictive utility. However, we observe data distribution of racial groups is skewed affecting demographic parity. We thereafter apply bias mitigation algorithms in the pre-processing stage, improving overall fairness by 10% highlighting effectiveness of our approach.