The validation of pain index extracted from electroencephalogram in chronic pain patients

Purpose A non-invasive tool is needed to help objectively assess pain. We aimed to validate the pain index (PI) extracted from electroencephalograms (EEG) to detect moderate-to-severe pain. Methods The participants were 18- to 80-year-old outpatients with moderate-to-severe chronic pain (NRS ≥ 4) who were eligible to receive injection therapy or shockwave therapy. The numerical rating scale (NRS) and PI extracted from the EEG signals were recorded before and 30 minutes after treatments. Anxiety and depression statuses were evaluated by the Hospital Anxiety and Depression Scale before treatments. The association between NRS and PI was evaluated using Pearson's correlations and regression analysis. The sensitivity and specificity of PI were analyzed using the area under the receiver operating characteristic (ROC) curves. Multiple regression analysis was used to analyze the associations of PI with anxiety and depression. Results A total of 111 patients were enrolled. A positive correlation was found between PI and NRS and between the changes in NRS (ΔNRS) and the changes in PI (ΔPI) before versus after treatments. The sensitivity and specificity of PI for moderate and severe pain were 80.4% and 86.5%, respectively, and the diagnostic threshold was 10.65. The area under the ROC curves of PI was 0.912. PI appeared to be slightly affected by anxiety or depression. Conclusion PI extracted from EEG appears to be a feasible objective tool to assess moderate-to-severe pain.