Predictive algorithms from Electronic Health Records

This repository hosts code for the working paper: Exploring a complexity gradient in representation and predictive algorithms for EHRs

Documentation

Source Code

Working Paper repository

Abstract

Electronic Health Records contain time-varying features with high cardinality. Current state-of-the-art predictive models build on increasingly elaborated pipelines –based on transformers– to handle the complexity of these data. Acknowledging the complexity to deploy, transfer and adapt these models on local care environments, we explore a complexity-benefit tradeoff by comparing them to simple aggregation of events. We use three clinical tasks involving time-varying structured Electronic Health Records (EHRs) and increasingly clinically relevant problems. We show that these benchmarking tasks display heterogeneous predictive difficulties. We introduce a simple aggregation of static embeddings –transferred from national claims and publicly available–, showing that it outperforms transformer-based models on simple tasks with medium sample sizes. We highlight the sample and computing resource efficiency of these models. Finally, clinically relevant problems generally present a strong class imbalance, which complicates models development and undermines their performances. Further work is needed to understand if transformer-based models perform well in these scenarios where the number of cases requires good sample efficiency.

Usage

See the usage page on the documentation