Reducing Drug Name Confusion with Better Search Software

使用更好的搜索软件减少药物名称混淆

• 批准号: 6880562
• 负责人: King Lup Liu
• 金额: $ 12.53万
• 依托单位: PHARM I.R.
• 依托单位国家: 美国
• 财政年份: 2005
• 资助国家: 美国
• 起止时间: 2005-04-15 至 2005-10-14
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/6880562
• 关键词: Internet; behavior test; behavioral /social science research tag; chemical information system; chemical registry /resource; clinical research; computer human interaction; computer program /software; computer system design /evaluation; decision making; drug classification; health services research tag; human subject; information retrieval; memory; meta analysis; nomenclature; patient safety /medical error; perception; vocabulary development for information system

DESCRIPTION (provided by applicant): Confusions between drug names that look and sound alike (e.g., Indocid(r) and Endocet(r)) continue to occur frequently, and each error threatens patient safety. Drug companies and regulators try to avoid confusion by screening new names prior to approval, but the effectiveness of screening is limited by over-reliance on subjective assessments of confusability, small sample sizes, lack of expertise in psycholinguistics, and weak empirical validation of testing methods. The FDA recently adopted a system that can search separately for similar names based on spelling or phonetic similarity, but optimal results will be achieved by combining the results of multiple search engines, where each distinct search engine implements a different similarity measure. Our long term objective is to design, build, test and continuously improve tools that help decision makers minimize the incidence of drug name confusion errors. Our short-term goal is to develop and validate several new techniques for merging the results of distinct search engines. Given a name as input, the search engine will return a merged list of existing names ranked in descending order of confusability. Confusability ratings will be based on validated, objective criteria derived from studies of clinicians' and lay persons' memory errors, perceptual errors, and similarity judgments. To further these goals, we plan to test the following hypothesis: The performance of a drug name metasearch engine that merges the results of multiple, distinct first-order searches will be superior to a search engine using a single measure of similarity. To test these hypotheses, we propose studies with the following specific aims: 1. To design and implement several different algorithms for merging the ranked results of separate drug name search engines. 2. To evaluate the alternative merging techniques in relation to previously reported errors, results of behavioral tests, and expert similarity judgments. 3. To incorporate the merging algorithms into a web-accessible, user-friendly search engine that can be used to support decision making during the drug name approval process.

描述（由申请人提供）：外观和听起来相似的药物名称之间的混淆（例如 Indocid(r) 和 Endocet(r)）继续频繁发生，每个错误都会威胁患者的安全。制药公司和监管机构试图通过在批准前筛选新名称来避免混淆，但由于过度依赖对易混淆性的主观评估、样本量小、缺乏心理语言学专业知识以及测试方法的实证验证薄弱，筛选的有效性受到限制。 FDA 最近采用了一种系统，可以根据拼写或语音相似性单独搜索相似名称，但最佳结果将通过组合多个搜索引擎的结果来实现，其中每个不同的搜索引擎实施不同的相似性度量。我们的长期目标是设计、构建、测试和不断改进工具，帮助决策者最大限度地减少药物名称混淆错误的发生率。 我们的短期目标是开发和验证几种用于合并不同搜索引擎结果的新技术。给定一个名称作为输入，搜索引擎将返回按易混淆性降序排列的现有名称的合并列表。混淆性评级将基于对临床医生和非专业人士的记忆错误、知觉错误和相似性判断的研究得出的经过验证的客观标准。为了进一步实现这些目标，我们计划测试以下假设：合并多个不同一阶搜索结果的药物名称元搜索引擎的性能将优于使用单一相似性度量的搜索引擎。为了检验这些假设，我们提出了以下具体目标的研究： 1. 设计和实现几种不同的算法来合并单独药物名称搜索引擎的排名结果。 2. 根据先前报告的错误、行为测试结果和专家相似性判断来评估替代合并技术。 3. 将合并算法合并到可通过网络访问、用户友好的搜索引擎中，该搜索引擎可用于支持药物名称批准过程中的决策。