EVALUATING PROFICIENCY OF SKIN TESTING

为了在临床上有用，诊断测试需要评估其 准确度、精密度和灵敏度。 一旦这些测试属性被 指定，然后可以通过确定来评估人员的熟练程度 他们的测试结果是否符合这些规格。 本研究探讨了 是否可以修改用于ID检测的FDA能力验证方法 在实践中 改良的方法采用两种剂量的FDA许可的 组胺碱（H）：1.8 mg/ml和0.1 mg/ml。 相对效价（RP） 这两个H浓度为18（1.8/0.1）。 通过ID测试的RP为 测定如下：终点稀释度和下3个系列的3倍 在注射0.05 μ l后，在每个受试者中测定每个H的稀释度。 ml. 终点为浓缩液的首次稀释液（稀释液#0） 其中红斑反应不存在或等同于风团。 可接受 滴定具有分级红斑反应，稀释度为50 mm 红斑总和（SE）。 稀释#产生45- 55 mm之间的SE或通过 在50 mm范围内的2个稀释度之间插值是估计的D50。 每个H的两个估计的D50之间的差是RP。 的 6名受试者的估计D50的平均差异为2.5（2 - 3）。 这是 测试准确度的估计。 该平均值的标准差小于 然后是0.6。 这是测试精度的估计值。 平均估计D50 1.8H的分子量为6（5.5-7）。 0.1 H的平均估计D50为4（3-4.5）。 这些 后两个平均值是测试灵敏度规格。 该方法 提供了一种方法来评估熟练的ID测试在实践中。

To be useful clinically, diagnostic tests require evaluation of their accuracy, precision, and sensitivity. Once these test properties have been specified, then personnel can be evaluated for proficiency by determining whether their test results meet these specifications. This study addressed whether an FDA proficiency method for ID testing could be modified for use in practice. The modified method employs two doses of FDA licensed histamine base (H): 1.8 mg/ml and 0.1 mg/ml. The relative potency (RP) of these two H concentrations is 18 (1.8/0.1). The RP by ID test was determined as follows: The endpoint dilution and the next 3 serial 3 fold dilutions of each H were determined in each subject after injection of 0.05 ml. The endpoint was the first dilution from the concentrate (Dilution #0) in which an erythema response was absent or equal to the wheal. Acceptable titrations had graded erythema responses with dilutions bracketing 50 mm sum of erythema (SE). The Dilution# producing a SE between 45-55mm or by interpolating between the 2 dilutions bracketing 50mm is the estimated D50. The difference between the two estimated D50s of each H is the RP. The mean difference in estimated D50 in six subjects is 2.5 (2 to 3). This is the estimate of test accuracy. The standard deviation of this mean is less then 0.6. This is an estimate of test precision. The mean estimated D50 of 1.8H is 6 (5.5-7). The mean estimated D50 of 0.1 H is 4 (3-4.5). These latter two means are test sensitivity specifications. The proposed method provides an approach to assess proficiency of ID testing in practice.