Chernoff Miriam, Ford-Chatterton Heather, Crain Marilyn J
Harvard School of Public Health, USA.
Case Studies Bus Ind Gov Stat. 2012 Oct;5(1):17-31.
To demonstrate the utility of a medical terminology-based method for identifying cases of possible mitochondrial dysfunction (MD) in a large cohort of youths with perinatal HIV infection and to describe the scoring algorithms.
Medical Dictionary for Regulatory Activities (MedDRA) version 6 terminology was used to query clinical criteria for mitochondrial dysfunction by two published classifications, the Enquête Périnatale Française (EPF) and the Mitochondrial Disease Classification (MDC). Data from 2,931 participants with perinatal HIV infection on PACTG 219/219C were analyzed. Data were qualified for severity and persistence, after which clinical reviews of MedDRA-coded and other study data were performed.
Of 14,000 data records captured by the EPF MedDRA query, there were 3,331 singular events. Of 18,000 captured by the MDC query, there were 3,841 events. Ten clinicians blindly reviewed non MedDRA-coded supporting data for 15 separate clinical conditions. We used the Statistical Analysis System (SAS) language to code scoring algorithms. 768 participants (26%) met the EPF case definition of possible MD; 694 (24%) met the MDC case definition, and 480 (16%) met both definitions.
Subjective application of codes could have affected our results. MedDRA terminology does not include indicators of severity or persistence. Version 6.0 of MedDRA did not include Standard MedDRA Queries, which would have reduced the time needed to map MedDRA terms to EPF and MDC criteria.
Together with a computer-coded scoring algorithm, MedDRA terminology enabled identification of potential MD based on clinical data from almost 3000 children with substantially less effort than a case by case review. The article is accessible to readers with a background in statistical hypothesis testing. An exposure to public health issues is useful but not strictly necessary.
证明一种基于医学术语的方法在识别大量围产期感染艾滋病毒青年中可能存在的线粒体功能障碍（MD）病例方面的效用，并描述评分算法。
使用《监管活动医学词典》（MedDRA）第6版术语，通过两种已发表的分类方法，即法国围产期调查（EPF）和线粒体疾病分类（MDC），查询线粒体功能障碍的临床标准。分析了参与围产期感染艾滋病毒的2931名参与者在PACTG 219/219C研究中的数据。对数据的严重程度和持续性进行了评估，之后对MedDRA编码的数据和其他研究数据进行了临床审查。
在EPF MedDRA查询捕获的14000条数据记录中，有3331个单一事件。在MDC查询捕获的18000条记录中，有3841个事件。十位临床医生对15种不同临床情况的非MedDRA编码支持数据进行了盲审。我们使用统计分析系统（SAS）语言编写评分算法。768名参与者（26%）符合EPF可能的MD病例定义；694名（24%）符合MDC病例定义，480名（16%）符合两种定义。
编码的主观应用可能影响了我们的结果。MedDRA术语不包括严重程度或持续性指标。MedDRA 6.0版不包括标准MedDRA查询，而标准MedDRA查询本可减少将MedDRA术语映射到EPF和MDC标准所需的时间。
MedDRA术语与计算机编码的评分算法一起，能够根据近3000名儿童的临床数据识别潜在的MD，与逐案审查相比，所需工作量大大减少。有统计假设检验背景的读者可以读懂本文。了解公共卫生问题会有帮助，但并非严格必需。
