An Ontology for the Social Determinants of Health Domain

TL;DR

This paper introduces SOHO, a comprehensive ontology for recording social determinants of health in electronic health records, aiming to standardize and improve data collection for better health outcomes.

Contribution

It presents the development and evaluation of SOHO, the first comprehensive ontology for SDOH terms in EHRs, filling a critical gap in health informatics.

Findings

SOHO covers relevant SDOH concepts and relationships.

Human expert review validated the ontology's comprehensiveness.

Algorithmic evaluation demonstrated SOHO's consistency and utility.

Abstract

Social determinants of health are societal factors, such as where a person was born, grew up, works, lives, etc, along with socioeconomic and community factors that affect individual health. Social Determinants of Health are correlated with many clinical outcomes, hence it is desirable to record SDOH data in Electronic Health Records (EHRs). Besides storing images, text, etc., EHRs rely on coded terms available in standard ontologies and terminologies to record observations and analyses. There is a substantial amount of research on understanding the clinical impact of SDOH, ranging from screening tools to practice based interventions. However, there is no comprehensive collection of terms for recording SDOH observations in EHRs. Our research goal is to develop an ontology that covers the terms describing SDOH. We present a prototype ontology called Social Determinant of Health Ontology (SOHO) that covers relevant concepts and IS--A relationships describing impacts and associations of social determinants. We describe the evaluation techniques that we applied to SOHO, including human experts review and algorithmic evaluation.