Benaroia Mark, Elinson Roman, Zarnke Kelly
London Health Sciences Center, Department of Internal Medicine, University of Western Ontario, 339 Windermere Road, London, Ontario N6A 5A5, Canada.
Int J Med Inform. 2007 Apr;76(4):283-8. doi: 10.1016/j.ijmedinf.2006.01.006. Epub 2006 Feb 13.
Patients can be used as a resource to enter their own pertinent medical information. This study will evaluate the feasibility of an intelligent computer medical history-taking device directed at patients in the emergency department (ED).
Two of the authors (MB, RE) developed an expert system that can take patient-directed medical histories. Patients interacted with the computer in the ED waiting room while it gathered a medical history based on chief complaint (CC). A survey was completed post history. A sub-study assessed the computer's ability to take an adequate history for an index CC. We compared the computer and emergency physician histories for the presence or absence of important historical elements.
Sixty-seven patients used the interactive computer system. The mean time to complete the history was 5 min and 32s +/- 1 min and 21s. The patient response rate was 97%. Over 83% felt that the computer was very easy to use and over 92% would very much use the computer again. A total of 15 patients with abdominal pain (index CC) were evaluated for the sub-study. The computer history asked 90+/-7%, and the emergency physician asked 55+/-18%, of the important historical elements. These groups were statistically different with a p-value of <0.00001.
This feasibility study has shown that the computer history-taking device is well accepted by patients and that such a system can be integrated into the normal process of patient triage without delaying patient care. Such a system can serve as an initial mode for documentation and data acquisition directly from the patient.
患者可作为输入自身相关医疗信息的资源。本研究将评估一种针对急诊科患者的智能计算机病史采集设备的可行性。
两位作者（MB、RE）开发了一个能采集患者主导病史的专家系统。患者在急诊科候诊室与计算机互动，计算机根据主诉收集病史。病史采集完成后进行一项调查。一项子研究评估计算机针对主要主诉采集充分病史的能力。我们比较计算机采集的病史与急诊医生采集的病史中是否存在重要的历史要素。
67名患者使用了交互式计算机系统。完成病史采集的平均时间为5分32秒±1分21秒。患者回应率为97%。超过83%的患者认为计算机非常易于使用，超过92%的患者表示会非常愿意再次使用该计算机。共有15名腹痛患者（主要主诉）接受了子研究评估。计算机采集到重要历史要素的比例为90±7%，急诊医生采集到的比例为55±18%。这两组在统计学上存在差异，p值<0.00001。
这项可行性研究表明，计算机病史采集设备深受患者欢迎，且这样的系统可融入患者分诊的正常流程，而不会延误患者治疗。这样的系统可作为直接从患者获取病历记录和数据的初始模式。
