Methods for Generating and Evaluating Synthetic Longitudinal Patient Data: A Systematic Review

TL;DR

This paper reviews methods for creating and evaluating synthetic longitudinal patient data, highlighting gaps in privacy, evaluation standards, and real-world applicability.

Contribution

The study systematically maps methods for synthetic longitudinal patient data, identifying key challenges and gaps in privacy and evaluation.

Findings

Thirty-nine methods were identified, with only four addressing all key challenges in longitudinal data generation.

Most studies evaluated resemblance and utility, but few considered privacy and all three aspects together.

No methods incorporated privacy-preserving mechanisms, and effectiveness with small sample sizes remains unclear.

Abstract

The rapid growth in data availability has facilitated research and development, yet not all industries have benefited equally due to legal and privacy constraints. The healthcare sector faces significant challenges in utilizing patient data because of concerns about data security and confidentiality. To address this, various privacy-preserving methods, including synthetic data generation, have been proposed. Synthetic data replicate existing data as closely as possible, acting as a proxy for sensitive information. While patient data are often longitudinal, this aspect remains underrepresented in existing reviews of synthetic data generation in healthcare. This paper maps and describes methods for generating and evaluating synthetic longitudinal patient data in real-life settings through a systematic literature review, conducted following the PRISMA guidelines and incorporating data from five databases up to May 2024. Thirty-nine methods were identified, with four addressing all key challenges in longitudinal patient data generation: preserving temporal structure, heterogeneous variable types, missing values, and unbalanced data. Most studies assessed resemblance to real data, the majority evaluated utility, and just over half examined privacy. However, only a minority considered all three aspects together. While four methods addressed the key challenges in generating synthetic longitudinal patient data, none incorporated privacy-preserving mechanisms. Additionally, their effectiveness with small sample sizes remains unclear, raising concerns about their real-world applicability. The lack of standardized evaluation criteria further complicates comparison. Future research should focus on developing privacy-preserving methods, robust evaluation frameworks, and ensuring publicly accessible code. Clearer directives from data protection authorities are needed, as synthetic patient data availability lags behind method development.

The online version contains supplementary material available at 10.1007/s41666-025-00223-7.

Figures