Muzoora Michael Rusongoza, Haffer Nina, Thun Sylvia, Vorisek Carina Nina
Berlin Institute of Health at Charité - Universitätsmedizin Berlin.
Charité - University Medicine Berlin.
Stud Health Technol Inform. 2025 Aug 7;329:1530-1534. doi: 10.3233/SHTI251095.
Societal or institutional identity assignments often create a disconnect between patients and healthcare providers, perpetuating biases in medical treatment. Addressing this gap is essential for fostering equitable and respectful care for all individuals, transcending societal norms. This study evaluates the limitations and ambiguities in existing semantic terminologies, specifically within LOINC and SNOMED CT, as applied to the Diversity Minimal Item Set (DiMIS) diversity domains. Suitability of SNOMED CT and LOINC was assessed using the scoring system ISO/TS 21564, and intercoder reliability was evaluated between two independent mapping specialists. The analysis revealed that while most data items had either a complete or partial equivalent in SNOMED CT or LOINC, 34% and 19% of items, respectively, could not be mapped. Intercoder reliability was low, potentially due to the limited percentage of concepts with equivalent meaning (30% for LOINC and 36% for SNOMED CT). The DiMIS used ambiguous terms, underscoring the necessity for using terminology to ensure clarity and accuracy. These findings highlight critical gaps to represent diverse identities, emphasizing the need for updates and enhancements to support evolving, patient-defined descriptors within a rapidly changing medical data landscape.
社会或机构的身份认定往往会造成患者与医疗服务提供者之间的脱节，使医疗中的偏见长期存在。弥合这一差距对于促进对所有人的公平和尊重的护理至关重要，这超越了社会规范。本研究评估了现有语义术语中的局限性和模糊性，特别是在应用于多样性最小项目集（DiMIS）多样性领域的LOINC和SNOMED CT中。使用评分系统ISO/TS 21564评估了SNOMED CT和LOINC的适用性，并评估了两位独立映射专家之间的编码员间信度。分析表明，虽然大多数数据项在SNOMED CT或LOINC中有完全或部分等效项，但分别有34%和19%的项无法映射。编码员间信度较低，可能是由于具有等效含义的概念百分比有限（LOINC为30%，SNOMED CT为36%）。DiMIS使用了模糊的术语，强调了使用术语以确保清晰度和准确性的必要性。这些发现突出了在表示不同身份方面的关键差距，强调了在快速变化的医疗数据环境中更新和增强以支持不断发展的、患者定义的描述符的必要性。
