Diagnostic accuracy of two questionnaires for the detection of neuropathic pain in the Spanish population

BackgroundSeveral questionnaires have been developed for the detection of neuropathic pain. ObjectivesThis study aimed to compare the diagnostic accuracy of the Leeds Assessment of Neuropathic Symptoms and Signs (LANSS) pain scale and the Douleur Neuropathique en 4 questions (DN4) questionnaire for the detection of peripheral neuropathic pain in the Spanish population, and to analyse in detail the diagnostic quality of each item in these questionnaires. MethodsA total of 192 patients were enrolled. We compared the validity of the DN4 and LANSS questionnaires by studying sensitivity and specificity and using receiver operating characteristic (ROC) curve analysis. We also analysed the validity of each item in the questionnaires. ResultsThe sensitivity of the DN4 questionnaire with an accepted cut-off value of 4 was 95.04% and that of the LANSS questionnaire with an accepted cut-off value of 12 was 80.17%. The specificity of the DN4 instrument was 97.18% and that of the LANSS instrument was 100%. The estimated area under the ROC curve (95% confidence interval) was 0.989 (0.977-1) for the DN4 instrument and 0.973 (0.956-0.991) for the LANSS questionnaire. The area under the ROC curve was significantly larger for the DN4 than the LANSS questionnaire (p<0.05). Analyses of specific items showed that tingling and numbness in the DN4 tool, and light touch pain and altered pinprick threshold in the LANSS scale, were the most important features of neuropathic pain. ConclusionsThese results show that although both questionnaires are good screening tools, the DN4 questionnaire is particularly recommended for identifying patients with neuropathic pain in clinical practice and research studies.