Williams B C, Philbrick J T, Becker D M, McDermott A, Davis R C, Buncher P C
Department of Internal Medicine, University of Virginia Health Sciences Center, Charlottesville.
J Gen Intern Med. 1991 Jan-Feb;6(1):57-63. doi: 10.1007/BF02599394.
To develop a patient-based classification system to describe the clinical content of ambulatory medicine practices.
A system of 100 diagnosis clusters was developed based on retrospective review of computerized problem lists of patients from a university practice, and then applied to the problem lists of patients in a community practice. Chart review of a 5% random sample (n = 184) of university practice patients who had problem lists was carried out to assess the accuracy of the computerized problem lists.
A university ambulatory medicine practice and a community ambulatory medicine practice.
PATIENTS/PARTICIPANTS: For the same one-year period, all 4,490 patients seen in the university practice and all 1,294 patients seen two or more times in the community practice.
None.
Of the 27,634 problems listed for university patients and the 5,648 problems listed for community patients, 22,629 (82%) and 4,924 (87%), respectively, were assigned to diagnosis clusters. For the university and community practices, the mean numbers of problems per patient were 6.1 (SD 5.4) and 4.4 (SD 3.7), and the mean numbers of diagnosis clusters per patient were 4.5 (SD 3.7) and 3.6 (SD 3.0), respectively. Among the ten most common diagnosis clusters in both practices were HYPERTENSION, SYMPTOM OR SIGN, OBESITY, and DIABETES. Only 18% (SD 3%) of patient problem lists in the university practice omitted one or more chronic, important medical problems (e.g., hypertension, dementia, COPD).
This system of diagnosis clusters effectively and efficiently described the clinical content of two types of internal medicine practices, and has important applications in medical education, epidemiology, clinical and health services research, and public policy.
开发一种基于患者的分类系统，以描述门诊医疗实践的临床内容。
基于对一所大学医疗机构患者计算机化问题列表的回顾性分析，开发了一个包含100个诊断类别的系统，然后将其应用于社区医疗机构患者的问题列表。对有问题列表的大学医疗机构患者的5%随机样本（n = 184）进行病历审查，以评估计算机化问题列表的准确性。
一所大学门诊医疗实践机构和一个社区门诊医疗实践机构。
患者/参与者：在同一为期一年的时间段内，大学医疗机构诊治的所有4490名患者以及社区医疗机构中就诊两次或以上的所有1294名患者。
无。
大学患者列出的27634个问题和社区患者列出的5648个问题中，分别有22629个（82%）和4924个（87%）被归入诊断类别。对于大学和社区医疗机构，每位患者的平均问题数分别为6.1（标准差5.4）和4.4（标准差3.7），每位患者的平均诊断类别数分别为4.5（标准差3.7）和3.6（标准差3.0）。两种医疗机构中最常见的十个诊断类别包括高血压、症状或体征、肥胖症和糖尿病。大学医疗机构中只有18%（标准差3%）的患者问题列表遗漏了一个或多个慢性、重要的医疗问题（如高血压、痴呆、慢性阻塞性肺疾病）。
该诊断类别系统有效且高效地描述了两种内科医疗实践的临床内容，在医学教育、流行病学、临床和卫生服务研究以及公共政策方面具有重要应用。
