基于网络的电子应用在内镜术前抗血栓药物管理中的开发与验证
研究开发了基于ASGE 2016指南算法的网络应用ENDOAID,用于内镜术前抗血栓药物管理决策。
要点
- ENDOAID基于ASGE 2016指南算法开发,可生成抗血栓药物围术期管理建议。
- 验证研究中初始一致率99.03%(206/208次模拟),经专家复核后达100%。
核心结论
- ENDOAID是一个易于使用的网络应用,可辅助抗血栓药物的围术期管理,与ASGE指南高度一致。
研究设计与样本
- 基于JavaScript软件开发网络应用ENDOAID,依据2016年更新的ASGE指南算法将患者和操作分为低风险和高风险。
- 验证研究纳入52例转诊内镜并服用抗血栓药物的患者,分为5组,每组患者模拟4种操作,共208次模拟场景,涵盖26种操作。
- ENDOAID结果与2名内镜医生共识确定的ASGE指南推荐(金标准)进行比较,操作ENDOAID和评估ASGE指南的人员不同以避免偏倚。
主要发现
- ENDOAID推荐与金标准初始一致率为99.03%(206/208次模拟场景)。
- 2次需进一步复核的场景涉及Factor V Leiden突变和抗磷脂抗体综合征所致深静脉血栓患者接受高风险操作且指南模糊,ENDOAID建议术前咨询专家,该建议获专家内镜医生共识同意。
关键数值与推荐
- ENDOAID依据2016年更新的ASGE指南将患者和操作分为低风险和高风险。
- ENDOAID根据患者心血管风险和已发表的ASGE指南生成抗血栓药物管理建议。
临床怎么用
- ENDOAID可辅助内镜术前抗血栓药物的围术期管理决策。
- 该应用有潜力简化抗血栓药物管理,并可能避免操作延迟、取消或不必要的会诊。
局限性
- 原文未述局限性
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
BACKGROUND AND AIMS: Antithrombotic therapy among patients undergoing GI procedures is frequently encountered and can impact the procedure and patient outcomes. The American Society for Gastrointestinal Endoscopy (ASGE) guidelines help to manage these medications before endoscopy depending on the patient's clinical status and the type of GI procedure. However, currently there is no readily available electronic tool that can assist in decision-making regarding preprocedural management of these agents. Our aim was to develop an electronic application, endoscopy + aid (ENDOAID), to help manage antithrombotic agents before endoscopy and to perform a validation study to test its accuracy. METHODS: ENDOAID, a web-based application, was developed using JavaScript software (Oracle Corporation, Redwood Shores, Calif, USA) based on an algorithm to categorize patients and procedures into low and high risk as outlined in the updated ASGE guidelines published in 2016. Once pertinent information regarding a patient's clinical status and the procedure are entered, the application generates recommendations for the management of antithrombotic agents based on their cardiovascular risk and published ASGE guidelines. We performed a validation study with 52 patients who were referred to endoscopy and were taking antithrombotic agents. The patients were divided into groups of 5, and in the simulation each patient had 4 procedures. Different GI procedures were used in the simulation for each group of patients to ensure the entire spectrum of procedures were covered for analysis. Every simulation was then run through ENDOAID. The results from ENDOAID were compared with recommendations based on ASGE guidelines. The latter was derived by consensus between 2 endoscopists (the criterion standard). The personnel using the ENDOAID and those using the ASGE guidelines were different to avoid bias. Any clinical scenario that was unclear or not clearly outlined in ASGE guidelines was discussed with expert endoscopists for a final decision. We evaluated ASGE recommendations and calculated concordance rates between guidelines and ENDOAID results. The Pearson correlation coefficient (r) was calculated to assess the correlation between ENDOAID results to guidelines. RESULTS: There was a total of 208 simulated encounters, including 26 procedures. Initial concordance between ENDOAID recommendations and the criterion standard was seen in 206 encounters (99.03%). The 2 encounters that needed further review occurred among patients with Factor V Leiden mutation and deep vein thrombosis from antiphospholipid antibody syndrome and who were undergoing high-risk procedures that had ambiguous guidelines. ENDOAID suggested consultations with an expert before the elective procedure. This suggestion was agreed on by expert endoscopist consensus. Thus, ENDOAID showed a 100% concordance with the ASGE guideline for managing antithrombotics. There was a high degree of correlation (r = .996, P < .01) between ENDOAID results with ASGE. CONCLUSIONS: We have developed and validated an easy-to-use web-based application that can help in periprocedural management of antithrombotics. Such an application has the potential to simplify the management of these agents and potentially prevent procedural delays, cancellations, or unnecessary consults.
引用
- Nutalapati V, Tokala KT, Desai M, et al. Development and validation of a web-based electronic application in managing antithrombotic agents in patients undergoing GI endoscopy. Gastrointest Endosc. 2019. PMID: 31228431. doi:10.1016/j.gie.2019.06.015
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。