Laaksonen Niina, Varjonen Juha-Matti, Blomster Minna, Palomäki Antti, Vasankari Tuija, Airaksinen Juhani, Huupponen Risto, Scheinin Mika
Institute of Biomedicine, University of Turku, Kiinamyllynkatu 10, FI-20520, Turku, Finland.
Auria Clinical Informatics, Hospital District of Southwest Finland, PO Box 52, FI-20521, Turku, Finland.
Contemp Clin Trials Commun. 2020 Dec 18;21:100692. doi: 10.1016/j.conctc.2020.100692. eCollection 2021 Mar.
Electronic health records (EHR) are a potential resource for identification of clinical trial participants. We evaluated how accurately a commercially available EHR Research Platform, InSite, is able to identify potential trial participants from the EHR system of a large tertiary care hospital. Patient counts were compared with results obtained in a conventional manual search performed for a reference study that investigated the associations of atrial fibrillation (AF) and cerebrovascular incidents. The Clinical Data Warehouse (CDW) of Turku University Hospital was used to verify the capabilities of the EHR Research Platform. The EHR query resulted in a larger patient count than the manual query (EHR Research Platform 5859 patients, manual selection 2166 patients). This was due to the different search logic and some exclusion criteria that were not addressable in structured digital format. The EHR Research Platform (5859 patients) and the CDW search (5840 patients) employed the same search logic. The temporal relationship between the two diagnoses could be identified when they were available in structured format and the time difference was longer than a single hospital visit. Searching for patients with the EHR Research Platform can help to identify potential trial participants from a hospital's EHR system by limiting the number of records to be manually reviewed. EHR query tools can best be utilized in trials where the selection criteria are expressed in structured digital format.
电子健康记录（EHR）是识别临床试验参与者的潜在资源。我们评估了一款商用EHR研究平台InSite从一家大型三级护理医院的EHR系统中准确识别潜在试验参与者的能力。将患者数量与为一项关于房颤（AF）与脑血管事件关联的参考研究进行的传统人工搜索结果进行了比较。图尔库大学医院的临床数据仓库（CDW）用于验证EHR研究平台的功能。EHR查询得出的患者数量比人工查询多（EHR研究平台为5859名患者，人工筛选为2166名患者）。这是由于搜索逻辑不同以及一些无法以结构化数字格式处理的排除标准所致。EHR研究平台（5859名患者）和CDW搜索（5840名患者）采用了相同的搜索逻辑。当两种诊断以结构化格式存在且时间差超过单次医院就诊时间时，可以识别它们之间的时间关系。使用EHR研究平台搜索患者有助于通过限制需人工审查的记录数量，从医院的EHR系统中识别潜在试验参与者。EHR查询工具在选择标准以结构化数字格式表示的试验中能得到最佳利用。
