RETAIN: An Interpretable Predictive Model for Healthcare using Reverse Time Attention Mechanism

TL;DR

RETAIN is a novel neural network model that balances high predictive accuracy with interpretability for healthcare data, specifically Electronic Health Records, by using reverse time attention to highlight influential visits and variables.

Contribution

The paper introduces RETAIN, a new attention-based neural network that improves interpretability in healthcare predictions without sacrificing accuracy.

Findings

RETAIN achieves accuracy comparable to RNNs on large EHR datasets.

RETAIN provides interpretability similar to traditional models.

RETAIN is scalable to millions of patient visits.

Abstract

Accuracy and interpretability are two dominant features of successful predictive models. Typically, a choice must be made in favor of complex black box models such as recurrent neural networks (RNN) for accuracy versus less accurate but more interpretable traditional models such as logistic regression. This tradeoff poses challenges in medicine where both accuracy and interpretability are important. We addressed this challenge by developing the REverse Time AttentIoN model (RETAIN) for application to Electronic Health Records (EHR) data. RETAIN achieves high accuracy while remaining clinically interpretable and is based on a two-level neural attention model that detects influential past visits and significant clinical variables within those visits (e.g. key diagnoses). RETAIN mimics physician practice by attending the EHR data in a reverse time order so that recent clinical visits are likely to receive higher attention. RETAIN was tested on a large health system EHR dataset with 14 million visits completed by 263K patients over an 8 year period and demonstrated predictive accuracy and computational scalability comparable to state-of-the-art methods such as RNN, and ease of interpretability comparable to traditional models.