Does Fentanyl or Remifentanil Provide Better Postoperative Recovery After Laparoscopic Surgery?

A Randomized Controlled Trial

Ayako Asakura; Takahiro Mihara; Takahisa Goto


BMC Anesthesiol. 2018;18(81) 

In This Article


We have identified no significant differences in global QoR-40 score 24 h after surgery, between the fentanyl and the remifentanil groups. However, within the five dimensions, physical independence scores were significantly higher in the fentanyl group. Cortisol and ACTH measured during and at the end of the surgery showed significantly high plasma concentration values in the fentanyl group. No differences were observed in the incidence of PONV, the pain score, postoperative fentanyl consumption, and the time when patients started drinking or walking.

The fentanyl group showed a higher global QoR-40 score by 20 points than the remifentanil group, which is more than three times higher of minimal clinically important difference (MCID) for the QoR-40 that has been reported as 6.3.[17] Therefore, the difference between the groups is largely relevant. Nevertheless, no statistically significance was apparent. Two possible explanations for the results could be considered. First, despite the need for at least 66 subjects, primary outcome data were available for only 63 subjects. Three patients had some questions skipped, and 3 patients had lost the questionnaire at their home (Figure 1), which were unfortunately, more than we have expected. Second, the overall SD was 18 when we performed the sample size calculations with 28 subjects; however, it was 22 when the final analyses were performed with 63 subjects. Therefore, the allocation bias seems to be not minimized by the randomization. Overall, we could state that the current study was underpowered. In addition to these, we should discuss whether the MCID derived from the Australian study[17] could directly extrapolate to the Japanese population. The SD of the QoR40 score in the validation study of the Japanese version was also 22,[13] which was 1.6 times higher than that of the Australian study (i.e. 14).[17] Because in general, the MCID become higher with a higher SD, the MCID in Japanese population could be 1.6 times higher (i.e. 10) than the Australian population. Nevertheless, the difference of the QoR40 in our study still exceeded the MCID.

There was a significant difference in the physical comfort dimension. This dimension asks about breathing, sleeping, eating, resting, PONV, shivering, etc. As there were no differences in the PONV and the time of the patients' drinking, the significant difference must have occurred in one of the other remaining factors. There also was a significant difference in physical independence. Although the time when the patients started walking was not significantly different, their medians were 25.5 and 42.0 for the fentanyl and remifentanil groups, respectively. This is almost a day different and could have affected the score in the physical independence dimension, which was assessed on POD1.

With a longer period of time to recover from anesthesia and surgery, it is intriguing that the significant difference appeared in the GH domain of the SF-36. This domain mainly questions about whether the subjects are feeling healthy or not. Although it might not be related much to the time course of recovery, fentanyl might have a better effect on QoL a few months after surgery, compared with remifentanil.

The plasma concentration levels of cortisol and ACTH during surgery were higher in the fentanyl group, which is consistent with past study that remifentanil suppressed the increase in ACTH and cortisol during laparoscopic colectomy compared with epidural anesthesia.[18] The global QoR-40 score tended to be higher in the fentanyl group, nevertheless, no correlation between the global QoR-40 score and cortisol was apparent (r = 0.089). In the previous studies, premedication with ibuprofen improved the QoR-40 of POD1 in spite of intraoperative cortisol levels as high as control group.[8] Furthermore, intraoperative infusion of dexmedetomidine showed no difference in the QoR-40 score on POD1, although cortisol levels were significantly lower than control group after surgery.[7] Altogether, intraoperative cortisol seems not relevant to the QoR-40 score on POD1. Multiple factors are responsible for recovery from surgery, and thus we considered that some other factors beside hormone have affected the QoR-40 score, which, disappointingly, cannot be clarified in our study design.

There are limitations to our study. First, as previously discussed in detail, the study was underpowered. Second, the SF-36 is not specifically designed for use after surgery, so it may not be reliable for measuring the intermediate to the late phase postoperative recovery. Perhaps, we should have used other tools,[19] e.g. the functional recovery index,[20] the surgical recovery index,[21] or the postoperative quality of recovery score.[22] Third, small amount of fentanyl was used periodically throughout the management of remifentanil group, to base fentanyl for postoperative PCA. However, we considered that not using fentanyl might lead to lower QoR-40 score with stronger postoperative pain in remifentanil group. Owing to this reason, minimum amount of fentanyl was given in the remifentanil group. Fourth, external validity is low, and our results may not apply to patients with severe comorbidities or patients undergoing more invasive surgery, since the enrolled subjects were relatively healthy patients who underwent less invasive surgery. In addition, two thirds of the participants were male. Patient sex is known to affect QoR,[23] thus, we should have recruited patients of either sex.