PROSPECT guideline for elective caesarean section: an update

The PROSPECT group recently added an update [1] to their original recommendations for the management of postcaesarean section pain [2]. Amongst the updates was a strengthened recommendation for the benefits of dexamethasone. The original guidelines recommended the use of dexamethasone on the basis of four trials [3–6] and this recommendation was graded as Grade A. The recent guidelines strengthened the recommendations because of a more recent systematic review [7]. Taking the four trials in order, I believe there are deficiencies in them that I wish to explore. “Use of i.v. dexamethasone was associated with better pain scores; prolongation of analgesic effect. . .” [3] Participants in this trial had less than the ED95 of bupivacaine [8] and no additional intrathecal opioid. There was no power calculation performed before the trial. Patients were allocated randomly but the clinicians were not blinded. There was a statistically significant increase in the duration of sensory block and an increase in time to first analgesia (not a prespecified outcome). There was nomention of total tramadol requirements other than a single sentence saying it was reduced (with no statistical analysis for this statement), despite that being a primary outcome. Visual analogue scores (VAS) were reported as being reduced by (approximately) 1 for the first 24 h at every time-point despite these not being a prespecified outcome. “. . .a reduction in opioid consumption” [4]. This was a better-conducted trial, although relevance to current UK practice is uncertain. During surgery the uterus was routinely exteriorised; this is not currently recommended in UK practice and is associated with increased pain [9]. The postoperative analgesic regimen was unusual with the only regular analgesic being ketorolac 30 mg every 6 h. Lowdose oxycodone/paracetamol was also given, depending on the pain score, and intravenous morphine was given for rescue analgesia. This analgesic regimen is probably not representative of best practice and does not fit with the PROSPECT recommendations of regular normal dose paracetamol and non-steroidal anti-inflammatory drugs with rescue opioid. There was no difference in opioid consumption at 0–6 h, 12–24 h or total of 24 h. There was a slight reduction at 6–12 h (5 mg) but this was not a prespecified outcome. The authors of this trial themselves stated that “administration of dexamethasone 8 mg did not decrease opioid consumption in the first 24 hours”. “. . .a reduced need for postoperative anti-emetics” [5]. This study was single-blinded. The spinal bupivacaine dose was again less than the ED95 (10 mg) and no long-acting opioid was given (fentanyl only). Again, the uterus was routinely exteriorised. Previous power analysis was only for detecting a difference in mean pain VAS at 24 h. There was no difference in morphine use at 24 h between the placebo and intravenous dexamethasone groups, and the VAS at 24 h was higher in the intravenous group compared with placebo (p = 0.07). The study was not powered to detect an effect on nausea and vomiting rates. “Intravenous dexamethasone was not as effective as i.v. tramadol” [6]. It is probably unsurprising that a non-standard analgesic is not as effective as a recognised analgesic. Although in this study time to first analgesia was similar (3 h) and the VAS were only different for the first 60 min. There was no opioid in the spinal anaesthetic. The systematic review [7] included three of the four trials mentioned above [3–5] and an additional 10 trials. No trials utilised neuraxial diamorphine and only three trials utilised neuraxial morphine. The authors concluded: “In the absence of long-acting neuraxial agents, dexamethasone modestly reduced pain scores at rest with a higher efficacy immediately postoperatively, but with beneficial effects lasting up to 24 hours.” They also concluded that after subgroup analysis of the three trials with neuraxial morphine that there was not a “statistically significant effect of dexamethasone on early pain scores or time to rescue analgesia. Thus, it appears that an analgesic effect of dexamethasone is not detectable in the simultaneous presence of long-acting neuraxial opioids.” Given that the PROSPECT recommendations are to utilise long-acting neuraxial opioids, do the authors still consider the recommendation to use intravenous dexamethasone to be a Grade A recommendation in this scenario?