The effect of combining dexamethasone with ondansetron for nausea and vomiting associated with fentanyl‐based intravenous patient‐controlled analgesia

Abstract

Song et al. reported that the addition of dexamethasone to ondansetron failed to reduce the incidence of nausea and vomiting experienced by patients using fentanyl patient-controlled analgesia (PCA) after thoracoscopic surgery, although it appeared to lessen the severity [1]. We find the incidences of nausea and vomiting both with and without dexamethasone reported in this study (45% and 56%, respectively) surprisingly high and far in excess of those seen in our institution where fentanyl PCA is frequently used after kidney transplant. Other studies have reported considerably lower incidences of nausea and/or vomiting (30% or less) associated with the use of fentanyl PCA after arthroplasty [2], colorectal surgery [3], thoracoscopy [4] and thoracotomy [5], and during labour [6], even without any antiemetic prophylaxis. Incidences exceeding 50% have indeed been reported but are related to higher basal infusion rates [7]. Interestingly, in comparison with the above studies, the regimen employed by Song et al. consisted of a relatively high basal infusion rate (0.4 μg. kg−1.h−1) and low patient demand capacity (0.1 μg.kg−1 bolus every 15 min). It seems counterintuitive to use a high background infusion rate if this could result in a high incidence of nausea and vomiting. Moreover, in the presence of an insufficient fentanyl bolus on demand, we would question the validity of administering pethidine for breakthrough pain. Although the authors state that cumulative pethidine consumption was similar between groups, they did not provide the doses used over 48 hours. Without the patients’ pain scores, which would inform the reader about the efficacy of the PCA regimen, we find it difficult to put the severity of the side effects into the context of the benefits of any pain relief. We would suggest that the need for frequent pethidine administration suggests that the analgesic regimen is unsatisfactory. Finally, the practice of adding an antiemetic to PCA mixtures has declined in popularity because the dose received by patients is typically sub-therapeutic if PCA usage is reduced or stopped, and especially if boluses of an alternative opioid are then administered. We agree that it may be valid to state that dexamethasone failed to reduce the incidence of postoperative nausea and vomiting in this study; however, without reference to the cumulative doses of fentanyl, pethidine or ondansetron, we question the conclusion that these high incidences of nausea and vomiting, apparently refractory to 5-HT3 receptor-blocking drugs, are typical of fentanyl PCA in general.