Detection of Neuropathic Pain in End-Stage Cancer Patients: Diagnostic Accuracy of Two Questionnaires.

Abstract

Neuropathic pain is a frequent consequence of cancer pain. Quite often, in the end stage, it is difficult to discern its presence and delineate its characteristics in the context of painful cancer complications. The aim of this study was to compare the diagnostic accuracy of the Douleur Neuropathique en 4 Questions (DN4) and painDETECT questionnaires, which were translated to the patient's native language, for the diagnosis of peripheral neuropathic pain in oncology patients. End-stage cancer patients who presented to the outpatient pain clinic were prospectively followed. At presentation, all patients completed the DN4 and painDETECT questionnaires, which had been translated to their native language, and the output was compared to the pain specialist's diagnosis of the neuropathic or non-neuropathic nature of the pain, which was considered as the gold standard. The diagnostic accuracy of both questionnaires was tested with receiver operating characteristic curves plotted from the data collected. Ninety patients (48.5% of 185 in total) presented with severe pain. Seventy-six had neuropathic pain (41.1%) and 109 had non-neuropathic pain. Of those with neuropathic pain, most had a mixed pain (bone or visceral in addition to neuropathic pain). The DN4 questionnaire had a sensitivity of 71.1% and a specificity of 88.7% in detecting neuropathic pain, with a cutoff value of ≥ 4, while the painDETECT questionnaire had a sensitivity of 26.3% and a specificity of 100%, with a cutoff value of ≥ 19. At standard cutoff values, the DN4 and painDETECT questionnaires, despite having been translated to the patient's native language, failed to adequately discriminate between neuropathic and non-neuropathic pain in our end-stage cancer patients.