Leveraging Language Models and Machine Learning in Verbal Autopsy Analysis

TL;DR

This paper demonstrates that pretrained language models using verbal autopsy narratives significantly improve cause of death classification accuracy, especially when combined with question data, informing better health policy in data-scarce settings.

Contribution

It introduces the use of transformer-based language models for verbal autopsy narratives and explores multimodal fusion strategies, advancing automated cause of death classification methods.

Findings

Narratives alone with PLMs outperform question-only algorithms.

Multimodal fusion improves classification accuracy.

Classification accuracy varies with information sufficiency levels.

Abstract

In countries without civil registration and vital statistics, verbal autopsy (VA) is a critical tool for estimating cause of death (COD) and inform policy priorities. In VA, interviewers ask proximal informants for details on the circumstances preceding a death, in the form of unstructured narratives and structured questions. Existing automated VA cause classification algorithms only use the questions and ignore the information in the narratives. In this thesis, we investigate how the VA narrative can be used for automated COD classification using pretrained language models (PLMs) and machine learning (ML) techniques. Using empirical data from South Africa, we demonstrate that with the narrative alone, transformer-based PLMs with task-specific fine-tuning outperform leading question-only algorithms at both the individual and population levels, particularly in identifying non-communicable diseases. We explore various multimodal fusion strategies combining narratives and questions in unified frameworks. Multimodal approaches further improve performance in COD classification, confirming that each modality has unique contributions and may capture valuable information that is not present in the other modality. We also characterize physician-perceived information sufficiency in VA. We describe variations in sufficiency levels by age and COD and demonstrate that classification accuracy is affected by sufficiency for both physicians and models. Overall, this thesis advances the growing body of knowledge at the intersection of natural language processing, epidemiology, and global health. It demonstrates the value of narrative in enhancing COD classification. Our findings underscore the need for more high-quality data from more diverse settings to use in training and fine-tuning PLM/ML methods, and offer valuable insights to guide the rethinking and redesign of the VA instrument and interview.