从普通内科自由文本报告中自动编码医院转诊原因
Automatic coding of reasons for hospital referral from general medicine free-text reports.
作者信息
Letrilliart L, Viboud C, Boëlle P Y, Flahault A
机构信息
INSERM Unit 444, WHO Collaborating Center for Electronic Disease Surveillance, Paris, France.
出版信息
Proc AMIA Symp. 2000:487-91.
Although the coding of medical data is expected to benefit both patients and the health care system, its implementation as a manual process often represents a poorly attractive workload for the physician. For epidemiological purpose, we developed a simple automatic coding system based on string matching, which was designed to process free-text sentences stating reasons for hospital referral, as collected from general practitioners (GPs). This system relied on a look-up table, built up from 2590 reports giving a single reason for referral, which were coded manually according to the International Classification of Primary Care (ICPC). We tested the system by entering 797 new reasons for referral. The match rate was estimated at 77%, and the accuracy rate, at 80% at code level and 92% at chapter level. This simple system is now routinely used by a national epidemiological network of sentinel physicians.
尽管医学数据编码有望使患者和医疗保健系统都受益，但作为一种手动流程来实施时，它对医生来说往往是一项缺乏吸引力的繁重工作。出于流行病学目的，我们开发了一种基于字符串匹配的简单自动编码系统，该系统旨在处理从全科医生（GP）处收集的、陈述医院转诊原因的自由文本句子。该系统依赖于一个查找表，该表由2590份给出单一转诊原因的报告构建而成，这些报告已根据国际初级保健分类（ICPC）进行了手动编码。我们通过输入797条新的转诊原因对该系统进行了测试。匹配率估计为77%，准确率在代码级别为80%，在章节级别为92%。这个简单的系统现在被一个全国性的哨点医生流行病学网络常规使用。
Zotero 插件