Department of Surgery, Seinäjoki Central Hospital, Seinäjoki, Finland.
Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland.
J Med Syst. 2022 Aug 26;46(10):63. doi: 10.1007/s10916-022-01852-w.
Structured medical records improve readability and ensure the inclusion of information necessary for correct diagnosis and treatment. This is the first study to assess the quality of computer-generated structured medical records by comparing them to conventional medical records on patients with acute abdominal pain.
A prospective double-blinded study was conducted in a tertiary referral center emergency department between January 2018 and June 2018. Patients were examined by emergency department physicians and by experience and inexperienced researcher. The researchers used a new electronical medical records system, which gathered data during the examination and the system generate structured medical records containing natural language. Conventional medical records dictated by physician and computer-generated medical records were compared by a group of independent clinicians.
Ninety-nine patients were included. The overall quality of the computer-generated medical records was better than the quality of conventional human-generated medical records - the structure was similar or better in 99% of cases and the readability was similar or better in 86% of cases, p < 0.001. The quality of medical history, current illness, and findings of physical examinations were likewise better with the computer-generated recording. The results were similar when patients were examined by experienced or inexperienced researcher using the computer-generated recording.
The quality of computer-generated structured medical records was superior to that of conventional medical records. The quality remained similar regardless of the researcher's level of experience. The system allows automatic risk scoring and easy access for quality control of patient care. We therefore consider that it would be useful in wider practice.
结构化病历可提高可读性，并确保纳入正确诊断和治疗所需的信息。这是第一项评估计算机生成的结构化病历质量的研究，研究通过比较其与急性腹痛患者的常规病历来进行。
这是一项 2018 年 1 月至 6 月在三级转诊中心急诊科进行的前瞻性双盲研究。由急诊科医生和有经验和无经验的研究人员对患者进行检查。研究人员使用了一种新的电子病历系统，该系统在检查过程中收集数据，并生成包含自然语言的结构化病历。由一组独立的临床医生比较由医生口述的常规病历和计算机生成的病历。
共纳入 99 例患者。计算机生成的病历的整体质量优于常规的人工生成病历——在 99%的情况下，结构相似或更好，在 86%的情况下，可读性相似或更好，p<0.001。病史、当前疾病和体格检查结果的质量也更好，计算机生成的记录。当使用计算机生成的记录由有经验或无经验的研究人员检查患者时，结果相似。
计算机生成的结构化病历的质量优于常规病历。无论研究人员的经验水平如何，质量都保持相似。该系统允许自动风险评分，并便于进行患者护理质量控制。因此，我们认为它在更广泛的实践中会很有用。
