Measurement Error of a Simplified Protocol for Quantitative Sensory Tests in Chronic Pain Patients.

Abstract

Large-scale application of Quantitative Sensory Tests (QST) is impaired by lacking standardized testing protocols. One unclear methodological aspect is the number of records needed to minimize measurement error. Traditionally, measurements are repeated 3 to 5 times, and their mean value is considered. When transferring QST to a clinical setting, reducing the number of records would be desirable to meet the time constraints encountered in a routine clinical environment and to reduce the testing burden to chronic pain patients. However, there might be a trade-off between measurement error and number of records. We determined the measurement error of a single versus the mean of 3 records of pressure pain detection threshold (PPDT), electrical pain detection threshold (EPDT), and nociceptive withdrawal reflex threshold (NWRT) in 429 chronic pain patients recruited in a routine clinical setting. We calculated intraclass correlation coefficients and performed a Bland-Altman analysis. Intraclass correlation coefficients were all clearly greater than 0.75, and Bland-Altman analysis showed minute systematic errors with small point estimates and narrow 95% confidence intervals. Reducing the number of records from traditionally 3 to only 1 did not lead to relevant measurement error in PPDT, EPDT, or NWRT. This study contributes to a standardized QST protocol, and based on the minimal measurement error of 1 single record of PPDT, EPDT, and NWRT, we submit to reduce the testing burden. This would allow saving time, resources, and patient discomfort.