建置一個隨選可擴增式臨床決策支援框架:以流感預警與用藥警示系統為例

Implementation of an on-demand and extendible clinical decision support framework: cases studies of early prediction of influenza epidemics and duplicate medication detection

楊承益(Cheng-Yi Yang)

指導教授 ： 羅友聲

臺北醫學大學/醫學科技學院/醫學資訊研究所/博士(2019年)

摘要

醫令輸入系統(Computerized physician order entry; CPOE)整合臨床決策支援系統(Clinical decision support systems; CDSS)可提供醫療人員即時的輸入檢核，並在適當的時機與地點發出提示，以促進病人安全與臨床照護品質。然而，醫令輸入系統整合臨床決策支援系統仍有許多缺口待解決，包括1.新的醫療知識、臨床指引、法規、政策以及相關決策功能或檢覈規則必須與時俱進，避免臨床決策時採用不合時宜的決策知識; 2.臨床決策支援功能大多與醫令輸入系統高度耦合，因此當臨床決策知識、判斷邏輯或演算法需要更新時，難以進行維護; 3.缺乏整合病人跨院病歷之架構，造成病人就醫紀錄不完整而發生決策錯誤; 4.缺乏彈性、可互通之雙向溝通架構，使公共衛生知識能整合至臨床流程中。有鑑於此，本研究發展一個隨選可擴增式臨床決策支援框架，透過去耦合、模組化的方式，讓醫療機構能夠彈性的設計、選用或更新這些獨立的CDSS模組來開發出符合各種醫療情境的臨床決策支援系統。
首先我們針對隨選可擴增式臨床決策支援框架中的核心跨院病歷交換模組進行上傳、查詢與下載等情境的效能測試。此模組主要符合Integrating the Healthcare Enterprise, Cross–Enterprise Document Sharing integrating profile (IHE XDS.b)，因此本研究導入OpenXDS 當作跨院病歷交換系統來進行跨院病歷資料交換之模擬，並提供OpenXDS系統的調校建議。
接著，我們將可擴增式臨床決策支援框架實作於知識基礎的CDSS系統(Knowledge based CDSS)，並稱為CDS engine，藉此整合雲端藥歷提供病人跨院用藥紀錄進行重複用藥偵測的可行性驗證。本架構導入後，我們發現民眾簽署授權同意CDS engine使用個人雲端藥歷用藥史進行決策支援的比率為24.59%，CDS engine即時偵測出5.83%重複開立藥物之醫囑。另外研究發現，導入CDS engine後，花費時間為4.3分鐘，較未導入者的3.6分鐘多出0.7分鐘。最後，從醫師對CDS engine警示回復的結果分析發現，CDS engine有效偵測出42.06%的潛在重複開立藥物事件。
最後，我們將可擴增式臨床決策支援框架實作於非知識基礎的CDSS系統(Non-knowledge based CDSS)稱為整合型流感監視系統(Integrated Influenza Surveillance System)，提供公共衛生開放數據與臨床資訊系統達到雙向資料互通應用。整合型流感監視系統將跨院資料與衛生福利部疾病管制署(Taiwan Centers for Disease Control; TWCDC)線上流感開放資料進行整合後，透過流感監視演算模型進行運算，即時主動式提供醫院感染控制團隊監視結果。本研究蒐集了過去三年流感季節的資料(2014年10月至2017年9月)，並透過本研究定義出的三個流感監視指標(TMUHcS-ILI, TMUHcS-rapid influenza laboratory tests with positive results (RITP)以及TMUHcS-influenza medication use (IMU))進行回溯性驗證。研究結果發現整合型流感監視系統導入後，能夠有效將流感預警時間提前3到4週。
本研究發展一隨選可擴增式臨床決策支援框架，透過實作CDS engine以及整合型流感監視系統進行效果實證，並應用於重複用藥偵測以及流感預警後發現，系統不僅能夠提升潛在重複用藥偵測能力，也能夠有效提前流感預警，增加其監視的能力。未來，透過隨選可擴增式臨床決策支援框架之彈性擴增之特性，將可持續擴充例如藥物交互作用偵測、藥物過敏、高風險用藥、懷孕用藥以及其他疾病監視之功能，提升醫療照護品質。

關鍵字

臨床決策支援系統 ； IHE XDS.b ； 雲端藥歷 ； 重複用藥 ； 流感 監視

並列摘要

Computerized physician order entry (CPOE) systems coupled with Clinical decision support systems (CDSS), have been proposed as a key element of systems’ approaches to provide intelligent filter or presented reminder at appropriate times could improve patient safety and the quality of care. However, CPOE coupled with CDSS also have several issues to be solved, 1.new medical knowledge, clinical guidelines, regulations, policies, and CDSS functions must be kept updated to prevent the use of outdated knowledge; 2. CDSS rules are usually hard-coded or tight-bundled with CPOE or incorporated into CPOE, thus, the CPOE program has to be updated once the rules are updated; 3. Lack of an infrastructure supporting the integration of patients’ medical records prescribed by different healthcare facilities to prevent the medical error that occurred by incomplete medical records; 4. Lack of a flexible, interoperable infrastructure for bidirectional communications capable of integrating public health knowledge into clinical systems and workflows. In view of these issues, we developed an innovative on-demand and extendible clinical decision support framework which provides the decoupled decision support modules. Health care facilities can design or update these independent configurable CDSS modules separately with few configurations and invoke the needed data resources and CDSS function on-demanded.
Firstly, we conduct a performance testing on the required components for the Electronic Health Record (EHR) integration and transaction scenarios across healthcare facilities based on IHE XDS.b (Integrating the Healthcare Enterprise, Cross–Enterprise Document Sharing integrating profile). The performance testing was conducted for three use cases, EHR submissions, queries, and retrievals, based on the IHE XDS.b profile for EHRs sharing. In the performance testing of the EHR/OpenXDS system, the maximum affordable workload of the EHR submissions were 400 EHR submissions per hour. The maximum affordable workload of the EHR queries were 600 EHRs queries per hour. The maximum affordable workload of the EHR retrievals were 2000 EHR retrievals.
Secondly, we implement our framework in knowledge based CDSS and integrating the PharmaCloud to verify the feasibility of preventing the duplicate medication error. A CDS engine was developed which derived from this framework. In all, 43,844/178,300 (24.59%) patients signed the PharmaCloud consent form allowing their physicians to access their medication history in the PharmaCloud. The detection rate of duplicate medication was 5.83% (1843/31,614) of prescriptions. When prescribing using the CDS engine, the median encounter time was 4.3 (IQR 2.3-7.3) min, longer than that without using the CDS engine (median 3.6, IQR 2.0-6.3 min). From the physicians’ responses, we found that 42.06% (1908/4536) of the potential duplicate medications were recognized by the physicians and the medication orders were canceled.
Finally, we implement our framework in non-knowledge based CDSS and integrating the public health open data for bidirectional communications of public health knowledge into clinical systems and workflows for the influenza surveillances. An Integrated Influenza Surveillance System CDSS for integrating the EMRs of multiple hospitals with the influenza-like illness (ILI) data from the Taiwan Centers for Disease Control (TWCDC) website to proactively provide results of influenza epidemic monitoring to hospital infection control practitioners. We collected the EMRs and TWCDC ILI data of the past 3 influenza seasons (October 2014 to September 2017). On the basis of the EMRs of multiple hospitals, 3 surveillance variables, TMUHcS-ILI, TMUHcS-rapid influenza laboratory tests with positive results (RITP), and TMUHcS-influenza medication use (IMU), which reflected patients with ILI, those with positive results from rapid influenza diagnostic tests, and those treated with antiviral drugs, respectively, showed strong correlations with the TWCDC regional and national ILI data (r=.86-.98). The 2 surveillance variables—TMUHcS-RITP and TMUHcS-IMU—showed predictive power for influenza epidemics 3 to 4 weeks before the increase noted in the TWCDC ILI reports.
The implementation of this framework present a positive effect in the potential duplicate medication (PDM) detection and the early prediction of influenza epidemics. Firstly, the CDS engine could easily extend functions for detection of adverse drug reactions when more and more electronic health record systems are adopted. Moreover, the CDS engine can retrieve more updated and completed medication histories in the PharmaCloud, so it can have better performance for detection of duplicate medications. Secondly, the Integrated Influenza Surveillance System periodically integrated and compared surveillance data from multiple hospitals and the TWCDC website to maintain a certain prediction quality and proactively provide monitored results. This results can be extended to other infectious diseases, mitigating the time and effort required for data collection and analysis.

並列關鍵字

CDSS ； IHE XDS.b ； PharmaCloud ； Duplicate Medication ； Influenza surveillance

參考文獻

1. Berner, E.S., Clinical Decision Support Systems: Theory and Practice (Health Informatics) 3rd Edition. 2016: Springer.

2. Greenes, R.A., Clinical decision support: the road to broad adoption. 2014: Academic Press.

3. HealthIT.gov. Clinical Decision Support. 2018 [cited 2019 2019/5/25]; Available from: https://www.healthit.gov/topic/safety/clinical-decision-support.

4. Kohn, L.T., J. Corrigan, and M.S. Donaldson, To err is human: building a safer health system. Vol. 6. 2000: National academy press Washington, DC.

5. Baker, A.J.B., Crossing the quality chasm: a new health system for the 21st century. 2001, British Medical Journal Publishing Group. p. 1192.