Predictors of post-cesarean delivery pain in a prospective, randomized controlled trial of dexamethasone vs. placebo
Abstract Number: SAT-39
Abstract Type: Original Research
Introduction: Previous work has suggested that certain patient characteristics, responses to standardized questions, and to mechanical temporal summation (MTS) predict pain control after CD. Most of those results have not been validated in other centers. Therefore, we performed this analysis to assess if quantitative sensory testing and questionnaires could predict postoperative pain scores and opioid consumption in women undergoing CD.
Methods: Data were collected as part of a prospective, randomized controlled trial for post-CD pain control. Following IRB approval and informed consent, ASA 1-3 women scheduled to undergo primary or repeat CD were randomized to receive either IV dexamethasone 8mg or placebo after administration of a standardized spinal anesthetic. but prior to skin incision. Randomization was stratified by MTS response. Prior to surgery, subjects underwent MTS testing and were asked to rate their pain (NRS 0-10), anxiety (0-100), anticipated pain after surgery (NRS 0-100), and anticipated pain medication consumption after surgery (0-5). Postoperatively, subjects received pain medication according to a standardized protocol. Opioid consumption and pain scores at rest and with movement were collected at 2, 24, and 48-h after surgery. All postoperative opioids were converted to IV morphine equivalents. We performed multivariable regression analysis for 3 outcome variables: total postoperative opioid consumption at 24-h; postoperative pain at rest and with movement at 24-h. A set of candidate variables were included in the initial model on the criteria of p<0.15 and some clinically meaningful covariates. We performed backward variable selection by eliminating the variable with the largest p-value over 0.05 at each step until we reached the smallest AIC or BIC. Parameters in the final model were estimated and reported for each outcome separately.
Results: Data from 47 subjects were obtained and analyzed. Higher postoperative opioid consumption at 24-h was associated with older age (p = 0.01) and anticipated medication consumption (p = 0.01). Higher anticipated medication consumption was also associated with higher 24-h postoperative pain scores at rest (p = 0.04) and with movement (p = 0.05). Increasing gravidity was associated with higher pain scores at 24-h at rest (p < 0.0001), and with movement (p = 0.004). MTP-status was not associated with opioid consumption or pain scores.
Conclusions: Subjects’ predictions of post-op opioid consumption prior to CD were positively associated with actual post-op opioid consumption at 24-h. Older age was also associated with higher postoperative opioid consumption at 24-h. Increasing gravidity and anticipated medication consumption were both associated with higher 24-h pain scores at rest and with movement. We did not find that other predictors of postoperative pain, such as anxiety and anticipated pain after surgery, were significantly associated with postoperative pain control.