Uncovering Misattributed Suicide Causes through Annotation Inconsistency Detection in Death Investigation Notes

TL;DR

This paper presents an NLP-based method to detect annotation inconsistencies in death investigation notes, improving data quality for suicide cause attribution and highlighting the impact of data inconsistencies on research accuracy.

Contribution

The study introduces an empirical NLP approach for identifying annotation inconsistencies in NVDRS death notes, enhancing data reliability for suicide research.

Findings

Increased F-1 score by 5.4% when incorporating target state's data.

Detected annotation inconsistencies in NVDRS death investigation notes.

Identified problematic instances affecting data quality.

Abstract

Data accuracy is essential for scientific research and policy development. The National Violent Death Reporting System (NVDRS) data is widely used for discovering the patterns and causes of death. Recent studies suggested the annotation inconsistencies within the NVDRS and the potential impact on erroneous suicide-cause attributions. We present an empirical Natural Language Processing (NLP) approach to detect annotation inconsistencies and adopt a cross-validation-like paradigm to identify problematic instances. We analyzed 267,804 suicide death incidents between 2003 and 2020 from the NVDRS. Our results showed that incorporating the target state's data into training the suicide-crisis classifier brought an increase of 5.4% to the F-1 score on the target state's test set and a decrease of 1.1% on other states' test set. To conclude, we demonstrated the annotation inconsistencies in NVDRS's death investigation notes, identified problematic instances, evaluated the effectiveness of correcting problematic instances, and eventually proposed an NLP improvement solution.