American Board of Internal Medicine, Philadelphia, PA 19106, USA.
Eval Health Prof. 2010 Sep;33(3):302-20. doi: 10.1177/0163278710376400.
Much research has been devoted to addressing challenges in achieving reliable assessments of physicians' clinical performance but less work has focused on whether valid and accurate classification decisions are feasible. This study used 957 physicians certified in internal medicine (IM) or a subspecialty, who completed the American Board of Internal Medicine (ABIM) Diabetes Practice Improvement Module (PIM). Ten clinical and two patient-experience measures were aggregated into a composite measure. The composite measure score was highly reliable (r = .91) and classification accuracy was high across the entire score scale (>0.90), which indicated that it is possible to differentiate high-performing and low-performing physicians. Physicians certified in endocrinology and those who scored higher on their IM certification examination had higher composite scores, providing some validity evidence. In summary, it is feasible to create a psychometrically robust composite measure of physicians' clinical performance, specifically for the quality of care they provide to patients with diabetes.
大量研究致力于解决可靠评估医生临床绩效方面的挑战，但很少有工作关注于是否可行做出有效且准确的分类决策。本研究使用了 957 名在内科（IM）或亚专业领域获得认证的医生，他们完成了美国内科医师学院（ABIM）的糖尿病实践改进模块（PIM）。十个临床和两个患者体验指标被综合成一个综合指标。综合指标得分高度可靠（r =.91），整个评分范围内的分类准确性很高（>0.90），这表明可以区分高绩效和低绩效医生。在内分泌学领域获得认证的医生和在 IM 认证考试中得分较高的医生的综合得分较高，这提供了一些有效性证据。总之，为患有糖尿病的患者提供的护理质量，创建一个心理测量学上可靠的医生临床绩效综合指标是可行的。
