Pain is reduced by transcutaneous cervical vagus nerve stimulation and correlated with cardiorespiratory variability measures in the context of opioid withdrawal

Figures

• 

  Figure 1. Protocol, stimulation, and physiological sensing. (A) Protocol overview. Patients with opioid use disorder (OUD) abstained from substance use for a minimum of eight hours prior to an in-lab protocol. The in-lab protocol consisted of opioid cue audio and videos meant to elicit opioid craving and active tcVNS or sham stimulation. (B) Physiological sensing and stimulation. The electrocardiogram (ECG) signal was sensed in a 3-lead configuration using Ag/AgCl electrodes placed as shown. Respiratory effort (RSP) was sensed using a resistive belt worn as shown. Transcutaneous cervical vagus nerve stimulation (tcVNS) or sham stimulation was delivered to the right side of the neck, targeting the carotid artery. (C) In-lab protocol. The in-lab protocol consisted of neutral videos, shown for control purposes, opioid cue audio and videos to elicit opioid craving, and tcVNS or sham stimulation. The protocol began and ended with pain measurements via the numerical rating scale for pain (NRS Pain).

• 

  Figure 2. CONSORT diagram. Of the 31 patients assessed for eligibility, 8 patients were excluded or declined to participate. The remaining 23 patients were randomized to exclusively receive either active transcutaneous cervical vagus nerve stimulation (tcVNS) or sham stimulation. Two participants’ data in the active group were unusable due to equipment malfunctions. An additional participant from the active group withdrew from the study. The analysis thus included data from 9 patients in the active group and 11 patients in the sham group.

• 

  Figure 3. NRS pain scores for the active (green) and sham (gray) groups. (A) Each patient's NRS Pain datapoints. “Pre” datapoints refer to the NRS Pain scores collected at the beginning of the protocol; “Post” datapoints refer to the NRS Pain scores collected at the end of the protocol. Each patient's pair of scores is connected using a line; if two patients reported the same pain levels for both pre and post, one patient's pre and post datapoints were shifted to the left for the ease of visualization. The summary statistics depict the mean and standard deviations for the pre and post data. (B) Comparison of active and sham Δ NRS Pain scores. The Δ denotes differences between the post and pre NRS Pain data. Note that this Δ NRS Pain data have been presented with alternative summary statistics in previous work (14). The box and whisker plots summarize the individual Δ NRS Pain datapoints shown, where the active and sham group's data are shown on the right and left, respectively. The x marker is used for outliers, and * indicates P < 0.05.

• 

  Figure 4. Plots of Δ NRS pain vs. Δ RPV and Δ HRV across all patients. The Spearman correlation values, ρ, are shown following adjustment for device group (i.e., active vs. sham). Gray is used to depict the sham group's datapoints, and green is used to depict the active group's datapoints. x indicates the outlier datapoint, and * indicates P < 0.05. (A) Plot of Δ NRS Pain vs. Δ coefficient of variation of inspiration time [CV(Ti)]. Note that CV(Ti) is unitless. (B) Plot of Δ NRS Pain vs. Δ standard deviation of Normal-to-Normal intervals (SDNN).