We thank Dr Garg for his interest in our paper. He has raised a great number of issues, many of which were already covered in our discussion. However, we would like to respond to some of these points.
We were also surprised at the lack of antiemetic effect of dexamethasone and speculated on reasons for this in the discussion. The lack of additional analgesia is less surprising, as the study referred to only observed improved analgesia after 24 h and also used a dose of dexamethasone which was four times that used in our study.
‘Inhaled boluses’ were not used ‘frequently’, being required in 13 patients in the volatile induction and maintenance of anaesthesia (VIMA) group, six patients in the fentanyl–dexamethasone group and four in the fentanyl group. Although the ‘inhaled bolus’ would have transiently delivered more nitrous oxide, the possible emetic effect of this was obviously countered by the omission of fentanyl, as the overall results showed less severe postoperative nausea and vomiting (PONV) when fentanyl was omitted. We would like to correct one error in Dr Garg's letter, all patients in all groups were not fully satisfied with the control of PONV. It was only the median scores that were 10 in all three groups. There were more low scores indicating a lack of satisfaction in patients who received fentanyl. In the fentanyl groups combined, two patients scored 0, four scored 2, two scored 3, five scored 4 and five scored 5. In the VIMA group, one patient scored 3, two scored 4 and two scored 5. All other scores were of seven or more. This reflects the greater severity and need for antiemetics in those patients receiving fentanyl. As we mentioned in our discussion, patients knowing that they are participating in a trial of PONV who are regularly asked if they feel nauseous often give a positive response. In this context, moderate to severe symptoms or the need for antiemetics are probably better endpoints. The adverse effects of fentanyl were most clearly (but not exclusively) shown when the groups were combined, a common statistical method to increase power. The statistical tests used do not require equal group sizes to be valid.
We agree with Dr Garg, as mentioned in our discussion, that the adverse haemodynamic and respiratory effects of fentanyl were probably exaggerated by our sevoflurane administration regimen. This is one of the many compromises inherent in a randomized, double-blind study design. As the induction to incision interval was not standardized, providing this average time is not really helpful. However, even if the adverse events were discounted, there was no haemodynamic benefit to the addition of fentanyl. As all anaesthetics were titrated to consistent clinical endpoints by a single blinded anaesthetist, there was every opportunity to detect an agent-sparing effect, but this was not observed. As discussed in our paper, the majority of intraoperative movement was very minor and occurred far less frequently than with some other techniques that included opioids! The total surgical delay in the VIMA group was only 2 min and 15 s out of a total of 36 h of operating time in this group – hardly a clinical problem.
Finally, Dr Garg suggests that the analgesic effect of fentanyl was masked by the other analgesia we administered. This is entirely our point; surely, it is far better to provide adequate and long-lasting analgesia by nonopioid means rather than to rely on fentanyl that produces analgesia of far shorter duration than ibuprofen  and that also increases the incidence and severity of PONV. As mentioned in our discussion, several other studies have also shown that the use of modest doses of short-acting opioids in conjunction with a range of anaesthetic techniques can significantly increase PONV. We chose to give a weight-adjusted dose, which was obviously approximate for the reasons Dr Garg suggests. In routine practice, it is more common to give most patients 100 μg of fentanyl that, if anything, would increase the adverse effects. Fentanyl is still perceived as both innocuous and beneficial by a large proportion of anaesthetists. Our study provides further evidence of its detrimental effect, although showing no benefit in terms of analgesia (similar pain on awakening, followed by similar analgesic requirements with most patients requiring only one dose of postoperative analgesia prior to discharge) and minimal impact on intraoperative conditions. We would urge readers to consider abandoning the routine use of fentanyl to supplement day care anaesthesia.
1 Rosenblum M, Weller RS, Conard PL, et al
. Ibuprofen provides longer lasting analgesia than fentanyl after laparoscopic surgery. Anesth Analg 1991; 73:255–259.