Archer: Next-generation unstructured data access for hospitals and clinical trial sponsors, delivering efficiency, reducing costs and improving care

IgniteData是一家位于英国的创新型中小企业，其目标是通过其软件即服务平台Archer**改变医疗保健提供商（电子健康记录（EHR））和临床试验发起人（电子数据采集（EDC））系统的互操作性。**第一代Archer已成功开发并部署用于将结构化数据（例如EHR系统中的患者记录）从EHR传输到EDC。然而，**高达85% of actionable health data is stored in unstructured formats** (e.g. clinical narrative text found in Pathology reports, Radiology reports, Postoperative notes).This project will create the second-generation of Archer through the development of an AI-/NLP-driven module which answers trial-relevant questions from unstructured EHR data and inputs results into EDCs.**Market Opportunity/Need**The valuable data stored in unstructured formats within EHRs is currently resource-intensive and challenging to extract. Manual data transfer and reviews are still commonplace, despite being often time-consuming, slow-to-turnaround and error-prone since medical conditions and supporting evidence can potentially be missed \[Wong, 2016\]. Source-Document-Verification (required by manual data transfer to avoid errors) can alone account for 10-15% trial costs (e.g. ~£1-1.5m/PII trial).**Impacts/Project Outputs*** Directly extracting unstructured information from EHRs will enable exceptional reductions in the costs currently associated with on-site monitoring/auditing.* Second-generation Archer platform will deliver transformative growth within 5-years of launch.

IgniteData is an innovative, UK-based SME with a goal to transform the interoperability of healthcare provider (Electronic Health Record (EHR)) and clinical trial sponsor (Electronic Data Capture (EDC)) systems through its Software-as-a-Service platform **Archer**.**Project Objectives/Summary**The first generation of Archer has been successfully developed and deployed to transfer structured data (e.g. patient records in EHR systems) from an EHR to an EDC. However, **up to 85% of actionable health data is stored in unstructured formats** (e.g. clinical narrative text found in Pathology reports, Radiology reports, Postoperative notes).This project will create the second-generation of Archer through the development of an AI-/NLP-driven module which answers trial-relevant questions from unstructured EHR data and inputs results into EDCs.**Market Opportunity/Need**The valuable data stored in unstructured formats within EHRs is currently resource-intensive and challenging to extract. Manual data transfer and reviews are still commonplace, despite being often time-consuming, slow-to-turnaround and error-prone since medical conditions and supporting evidence can potentially be missed \[Wong, 2016\]. Source-Document-Verification (required by manual data transfer to avoid errors) can alone account for 10-15% trial costs (e.g. ~£1-1.5m/PII trial).**Impacts/Project Outputs*** Directly extracting unstructured information from EHRs will enable exceptional reductions in the costs currently associated with on-site monitoring/auditing.* Second-generation Archer platform will deliver transformative growth within 5-years of launch.