A Factored Generalized Additive Model for Clinical Decision Support in the Operating Room
This work addresses the need for interpretable yet powerful predictive models in clinical decision support, particularly in operating room settings, though it is incremental as it builds on existing generalized additive models.
The authors tackled the problem of limited expressive capability in clinical prediction models by proposing a factored generalized additive model (F-GAM) that preserves interpretability while allowing feature interactions, resulting in superior performance in predicting postoperative acute kidney injury and acute respiratory failure with higher AUPRC and AUROC compared to other methods.
Logistic regression (LR) is widely used in clinical prediction because it is simple to deploy and easy to interpret. Nevertheless, being a linear model, LR has limited expressive capability and often has unsatisfactory performance. Generalized additive models (GAMs) extend the linear model with transformations of input features, though feature interaction is not allowed for all GAM variants. In this paper, we propose a factored generalized additive model (F-GAM) to preserve the model interpretability for targeted features while allowing a rich model for interaction with features fixed within the individual. We evaluate F-GAM on prediction of two targets, postoperative acute kidney injury and acute respiratory failure, from a single-center database. We find superior model performance of F-GAM in terms of AUPRC and AUROC compared to several other GAM implementations, random forests, support vector machine, and a deep neural network. We find that the model interpretability is good with results with high face validity.