Department of Pain Medicine, BG University Hospital Bergmannsheil GmbH, Ruhr University, Bochum, Germany Centre of Biomedicine and Medical Technology Mannheim (CBTM), Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany Department of Neurology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany Division of Neurological Pain Research and Therapy, Department of Neurology, University Hospital Schleswig-Holstein, Campus Kiel, Kiel, Germany Department of Neurology, BG University Hospital Bergmannsheil GmbH, Ruhr-University Bochum, Germany Department of Neurology, Klinikum rechts der Isar, Technische Universität München, Munich, Germany.
Pain. 2015 Dec;156(12):2423-2430. doi: 10.1097/j.pain.0000000000000300.
Quantitative Sensory Testing (QST) is a psychophysical method assessing the somatosensory nervous system. A premise for comparable results between laboratories is standardized testing. Its quality can be proven by analyzing healthy subjects, because their results should lie within confidence intervals estimated from large database samples. However, it is unclear how many abnormal values can be tolerated. Based on a binomial distribution, a tool for assessing samples of healthy subjects was developed to detect inclusion errors (inclusion of nonhealthy subjects) or measuring errors (inaccuracies in single QST parameters). Sensitivity and specificity of detecting inclusion errors were assessed in 431 healthy subjects and 833 patients with neuropathic pain syndromes from the German Research Network on Neuropathic Pain (DFNS) database. Measuring errors were simulated by raising all absolute values in a single parameter by 0.5 SD. We calculated optimal cutoff values for group sizes of 16 healthy subjects, as implemented in the DFNS certification procedures. The algorithm was applied in the certification process of 18 European QST laboratories. With a specificity of 95% and a sensitivity of 60%, inclusion errors can be assumed for ≥4 abnormal values per subject, whereas ≥6 abnormal values per QST parameter and laboratory indicate measuring errors. Subsequently, in the certification process of 5 of 18 centers, inclusion or measuring errors were detected. In most cases, inclusion errors were verified and reasons for measuring errors were illuminated by the centers. This underlines the usefulness and validity of our tool in quality assurance of QST laboratories using the DFNS protocol.
定量感觉测试（QST）是一种评估躯体感觉神经系统的心理物理学方法。实验室之间进行可比结果的前提是标准化测试。通过分析健康受试者，可以证明其质量，因为他们的结果应该落在从大型数据库样本估计的置信区间内。然而，尚不清楚可以容忍多少异常值。基于二项式分布，开发了一种用于评估健康受试者样本的工具，以检测纳入误差（纳入非健康受试者）或测量误差（单个 QST 参数的不准确性）。在德国神经病理性疼痛网络（DFNS）数据库中的 431 名健康受试者和 833 名神经病理性疼痛综合征患者中评估了检测纳入误差的敏感性和特异性。通过将单个参数中的所有绝对值提高 0.5 SD 来模拟测量误差。我们计算了在健康受试者组大小为 16 的情况下的最佳截止值，这是在 DFNS 认证程序中实现的。该算法应用于 18 个欧洲 QST 实验室的认证过程中。特异性为 95%，敏感性为 60%时，可以假设每个受试者有≥4 个异常值存在纳入误差，而每个 QST 参数和实验室有≥6 个异常值则表明存在测量误差。随后，在 5 个中心中的 18 个认证过程中，检测到了纳入或测量误差。在大多数情况下，纳入误差得到了验证，中心揭示了测量误差的原因。这强调了我们的工具在使用 DFNS 协议的 QST 实验室质量保证中的有用性和有效性。
