To the Editor:
We read with interest the investigation by Wuethrich et al.1 on the effects of epidural analgesic mixture on urodynamic parameters after open renal surgery. However, we wish to raise several methodologic and interpretative concerns which may undermine the clinical validity of the authors’ conclusions. First, the authors repeatedly use language when comparing the bupivacaine and bupivacaine plus fentanyl groups (“the addition of fentanyl enhances this effect” [Abstract, Conclusions], “was more pronounced in the bupivacaine/fentanyl group” [Results, first section], “is more pronounced if fentanyl is added” [Discussion, paragraph 3], “greater increase in PVRs” [Discussion, final paragraph]) that implies a significant difference among groups, when the intergroup differences to which these statements refer were not significant.
Second, change in postvoid residual (PVR) is not a clinically useful primary endpoint because the measurement of PVR is variable at different times and can reflect other factors such as rate of diuresis or psychological inhibition. Absolute PVR might be a relevant proxy for impending urinary retention or need for recatheterization, but the reference offered for an association between absolute PVR greater than 180 ml and bacteriuria is in uninstrumented, nonsurgical patients, and refers to a chronic increase in PVR, which has little relevance to the setting at hand. The authors state that a change in PVR of 230 ml is clinically relevant; however, this is highly dependent on factors such as bladder capacity and initial PVR. Furthermore, the same group has previously shown that even though a small difference in PVR may exist in patients with thoracic epidural analgesia (TEA), all patients were able to void, and none required recatheterization.2
The study design includes proxy outcomes for urinary retention, but by leaving urinary catheters in place until postoperative day 5 (POD5), this study does not allow for any direct estimation of the true rate of retention during TEA (which, in other studies, has been extremely low and not different in patients with TEA in whom the urinary catheter is removed earlier).3,4 In addition, the finding of decreased compliance associated with TEA is confounded by continuous catheter drainage, which is known to decrease bladder compliance. We agree with the fact that appropriate monitoring for retention always should occur after the removal of urinary catheter, but we do not agree with the authors’ conclusion that the urodynamic changes observed in this study should preclude any attempt to remove a urinary catheter until TEA is discontinued on POD5. Assuming adequate monitoring for retention, recatheterization is a more realistic adverse outcome than the “long-term debilitating morbidity, such as loss of bladder function after acute urinary retention” cited by the authors. The study was probably underpowered to examine urinary tract infection; none were observed in either group. Other studies have demonstrated a decreased rate of urinary tract infection with removal of urinary catheters on POD1 in patients receiving TEA with no difference in the incidence of retention.3,4
Ambiguity exists in how the authors report their primary endpoint, which relies on a measurement of postoperative PVR. Is this the PVR measured by urodynamics on POD2 or the PVR measured by noninvasive ultrasound after epidural removal on POD5? If the latter, the use of two different methods to calculate a Δ PVR fails to take into account the variability between methods. If the former, where are the results of the PVR measured on POD5? How many patients were in retention and required recatheterization after epidural removal?
Finally, the external validity of the study may be called into question by the observations that not a single patient required systemic or epidural opioids for breakthrough pain or experienced an episode of postoperative nausea and vomiting.