Diagnostic accuracy of the Central Sensitization Inventory for detection of descending nociceptive inhibitory system impairment in patients with chronic musculoskeletal pain

Abstract

Introduction: The persistent aspect of musculoskeletal pain has been associated with central sensitization (CS). The inefficiency of descending nociceptive inhibitory system (DNIS) is a component of CS, and the Central Sensitization Inventory (CSI) is the most widely used questionnaire for the clinical identification of patients with CS. However, an investigation of the criterion validity and the diagnostic accuracy of the CSI against a psychophysical test to measure the efficiency of DNIS in patients with chronic musculoskeletal pain is lacking. Objective: To evaluate the concurrent validity and the diagnostic accuracy of the CSI for the detection of the impairment of the DNIS in patients with chronic musculoskeletal pain. Methods: Concurrent validity and diagnostic accuracy study were conducted in 267 consecutive patients with chronic musculoskeletal pain enrolled prospectively in an outpatient department. The concurrent validity and diagnostic accuracy of the CSI (index method) were compared with the cold pressor test, which was the psychophysical measure of the DNIS (reference standard). Spearman’s correlations assessed the concurrent validity of the CSI, and measurements of the diagnostic accuracy were performed. Results: Ninety- three (34.83%) patients had CSI scores  40, while 204 (76.40%) patients had CSI scores  22. A negative but non-significant correlation between CSI findings and the results of the CPT for the dorsal forearm site (rs = -0.009, p = .886) and tibialis anterior site (rs = -0.119, p = .0518) were found. The cutoff point of 22 of the CSI showed high sensitivity (70.37%, 95%IC 56.39, 82.02), high negative predictive value (74.60%,
95%IC 64.46, 82.63), and low accuracy (31.84%, 95%IC, 26.29, 37.79). The cutoff point of 40 of the CSI showed values of sensitivity (35.19%, 95%IC 22.68, 49.38) and specificity (65.26%, 95%IC 58.45, 71.63) below 70%, and high negative predictive value (79.89%, 95%IC 76.12, 83.18) when compared to the CPT. The ROC curve analysis yielded an area under the curve of 0.54 (95% CI 0.45 – 0.63, p = .395). Conclusion: The CSI is a useless instrument for detection of the impairment in the DNIS in patients with chronic musculoskeletal pain due to the absence of correlation with the cold pressor test result and the insufficient measurements of diagnostic accuracy.