Singer P A, Robb A, Cohen R, Norman G, Turnbull J
Department of Medicine, University of Toronto Faculty of Medicine, Ontario, Canada.
Acad Med. 1996 May;71(5):495-8. doi: 10.1097/00001888-199605000-00021.
To further examine the objective structured clinical examination (OSCE) as a performance-based assessment method for clinical ethics.
In the spring of 1993, a volunteer sample of 88 final-year medical students from all five Ontario medical schools took a four-station OSCE that used standardized patients and involved decisions to forego life-sustaining treatment. Performance was scored on a checklist of behaviors unique to each case. Data were analyzed for reliability using intraclass correlation coefficients and the Spearman-Brown prophecy formula.
Reliability of the test was only .28 as a result of a low average inter-station correlation of .07. To achieve a test reliability of .8, 41 stations (almost seven hours of testing time) would be required.
Because of its low test reliability, the OSCE is not a feasible stand-alone method for summative evaluation of clinical ethics. This performance-based evaluation method should be combined with other, more reliable evaluation methods. The OSCE has promise for formative evaluation.
进一步检验客观结构化临床考试（OSCE）作为一种基于表现的临床伦理评估方法。
1993年春季，来自安大略省所有五所医学院的88名最后一年医学生组成的志愿者样本参加了一个四站式OSCE，该考试使用标准化患者并涉及放弃维持生命治疗的决策。根据每个案例特有的行为清单对表现进行评分。使用组内相关系数和斯皮尔曼 - 布朗预测公式对数据进行可靠性分析。
由于站间平均相关性较低，仅为0.07，该测试的可靠性仅为0.28。要使测试可靠性达到0.8，则需要41个站（近七个小时的测试时间）。
由于其测试可靠性较低，OSCE不是对临床伦理进行总结性评估的可行独立方法。这种基于表现的评估方法应与其他更可靠的评估方法相结合。OSCE在形成性评估方面具有前景。
