Reply to “Use DN4-T to rule out non-neuropathic pain”

Dear Editor,

We are grateful for Lin et al’s interest in our study,¹ as well as their valuable comments.² As pointed out, the distribution of the scores of the Taiwan version Douleur Neuropathique 4 (DN4-T) questionnaire was symmetric, which raised some concerns about the validity of our findings. In fact, it is not uncommon for the scores of many screening instruments to be normally distributed. Take the recently published ID Pain-T paper as an example.³ The scores appeared to normally distributed in the neuropathic pain patient group. In fact, this could reflect the wide range of severity of neuropathic pain symptoms among these patients, that is, most patients had relatively higher scores than controls, and a minority of patients had particularly higher or lower scores. However, the distribution alone may not give enough information with regard to its clinical utility. In our study, the optimum cut-off score of 3 was determined by the receiver operating characteristic (ROC) curve coupled with c index, which is a widely accepted approach, and its predictive power and reliability were confirmed by an area under the ROC curve of 0.83 and a Cronbach’s alpha coefficient of 0.7, respectively. We agree that a sensitivity of 0.77 and a specificity of 0.78 are far from perfect, although an optimum cut-off usually represents a compromise between sensitivity and specificity, which was well demonstrated by the ROC curve and c index. However, some of your point are well taken, language and culture backgrounds, as well as clinical needs, should be taken into consideration, and field testing will be needed to validate our findings.