Achieving Inclusivity by Design: Social and Contextual Information in Medical Knowledge

Abstract

Objectives: To select, present, and summarize the most relevant papers published in 2020 and 2021 in the field of Knowledge Representation and Knowledge Management, Medical Vocabularies and Ontologies, with a particular focus on health inclusivity and bias. Methods: A broad search of the medical literature indexed in PubMed was conducted. The search terms 'ontology'/'ontologies' or 'medical knowledge management' for the dates 2020-2021 (search conducted November 26, 2021) returned 9,608 records. These were pre-screened based on a review of the titles for relevance to health inclusivity, bias, social and contextual factors, and health behaviours. Among these, 109 papers were selected for in-depth reviewing based on full text, from which 22 were selected for inclusion in this survey. Results: Selected papers were grouped into three themes, each addressing one aspect of the overall challenge for medical knowledge management. The first theme addressed the development of ontologies for social and contextual factors broadening the scope of health information. The second theme addressed the need for synthesis and translation of knowledge across historical disciplinary boundaries to address inequities and bias. The third theme encompassed a growing interest in the semantics of datasets used to train medical artificial intelligence systems and on how to ensure they are free of bias. Conclusions: Medical knowledge management and semantic resources have much to offer efforts to tackle bias and enhance health inclusivity. Tackling inequities and biases requires relevant, semantically rich data, which needs to be captured and exchanged.

Enriching Real-world Data with Social Determinants of Health for Health Outcomes and Health Equity: Successes, Challenges, and Opportunities

“…It is challenging to consolidate the concepts and measures in these frameworks and define relationships between SDoH concepts. Second, even though existing standard terminologies such as ICD-10, SNOMED CT, and LOINC have added certain concepts for SDoH, their coverage is limited, the actual use of these codes in EHR is low [17], and the semantic alignment between these coding schemes are challenging [76]. Researchers often reported a lack of granularity in these terminologies.…”

Inclusive Digital Health

“…The survey paper of the Knowledge and Representation Management (KRM) section presents a selection of 22 papers from 2020 and 2021 focusing on contributions to health inclusivity and bias [29]. The author, Janna Hastings, grouped these works according to their contribution, namely: (i) semantic resources developed to represent knowledge about health inequalities and biases, (ii) efforts to enable data sharing across different health information systems, communities, cultures, and disciplines while ensuring data privacy, and (iii) solutions that take into account the importance of reducing biases that exist in the data on which medical AI algorithms are trained in order to achieve inclusiveness.…”