So Eun-Young, Park Hyeoun-Ae
College of Nursing, Seoul National University, Seoul, Korea.
Stud Health Technol Inform. 2011;169:764-8.
The purpose of this study is to explore the ability of SNOMED CT to represent narrative statements of medical records. Narrative medical records of 281 hospitalization days of 36 patients with Gastrectomy were decomposed into single-meaning statements, and these single-meaning statements were combined into unique statements by removing semantically redundant statements. Concepts from the statements describing patients' problems and treatments were mapped to SNOMED CT concepts. A total 4717 single-meaning statements were collected and these single-meaning statements were combined into 858 unique statements. Out of 677 unique statements describing patients' problems and treatments, about 85.5% statements were fully mapped to SNOMED CT. The rest of the statements were partially mapped. This mapping result implies that physicians' narrative medical records can be structured and used for an electronic medical record system.
本研究的目的是探索SNOMED CT表示病历叙述性语句的能力。将36例胃切除术患者281个住院日的叙述性病历分解为单义语句，并通过去除语义冗余语句将这些单义语句合并为唯一语句。将描述患者问题和治疗的语句中的概念映射到SNOMED CT概念。共收集到4717条单义语句，并将这些单义语句合并为858条唯一语句。在677条描述患者问题和治疗的唯一语句中，约85.5%的语句完全映射到SNOMED CT。其余语句为部分映射。这一映射结果表明，医生的叙述性病历可以进行结构化处理并用于电子病历系统。
