What Has Happened to Evidence-based Medicine?

Abstract

Wake Forest University School of Medicine, Winston-Salem, North Carolina. rweller@wfubmc.eduIn Reply:—We appreciate the opportunity to respond to Dr. Chelly and his expert coauthors to correct an apparent misinterpretation of our discussion of retroperitoneal hematoma following posterior lumbar plexus blockade (LPB). Most important, we would like the readers of Anesthesiology to understand that our intent was to notify clinicians of the serious complication we observed, rather than to promulgate guidelines for clinical practice. Because the use of perineural catheters and lower extremity nerve block has increased substantially in recent years, 1we believe it is critical that complications of newer techniques are reported so that clinicians use appropriate caution as they introduce techniques novel to their practice. Because there are no data on the frequency of this complication, we believe no individual should suggest practice guidelines for others. To suggest we “insinuated malpractice” is ludicrous. We believe we were quite cautious in our discussion to emphasize that anesthesiologists must weigh the risks of concealed bleeding from deep blocks and anticoagulation with the benefits of this (and any) regional anesthetic technique. As proponents of regional anesthesia with considerable experience with continuous LBP, we simply stated how we are managing such patients at our own institution. Ultimately, experts may be able to develop a consensus statement with respect to peripheral nerve block and anticoagulation, but there are insufficient data at this time for any such statement.We would also like to respond to Chelly et al.’ s interpretation of the details of these two cases. We certainly considered the apparent vascular trauma in the first case a contributing factor in the complication. Repositioning of the catheter and a negative test dose suggested to us that we could provide satisfactory analgesia with the block and catheter already in place. We believe this is no different from similar practice with epidural anesthesia in which blood is initially aspirated, but the catheter is then readjusted or replaced and then used for anesthesia. Despite the authors’ reference to “evidence-based medicine,” they provide none for their recommendation that we should have “aborted” the procedure with the LPB catheter already in place. In this case report, we believe the important issue was the removal of the catheter within 2 h of an enoxaparin dose. We believe this timing (as well as the known vascular trauma) contributed to the complication. In the second case, it is true that anticoagulation was excessive, but it is also important to acknowledge that unintentional excessive anticoagulation sometimes occurs in clinical medicine. The authors seem to imply that this patient's retroperitoneal hemorrhage was spontaneous (i.e. , due to anticoagulation without relation to the LPB.) This denies the fact that the computed tomographic scan pinpointed the location of the hematoma to the site of needle placement. Once again, needle trauma and anticoagulation (excessive in this case) both seemed to have contributed to the complication.We feel compelled to respond to Chelly et al.’ s characterization of retroperitoneal hematoma as “potential and rare.” Certainly, retroperitoneal hematoma has been reported less frequently following LPB than spinal hematoma has been reported following epidural blockade, but none would debate that the number of continuous epidural anesthetics nationwide is substantially higher than the number of continuous LPB. Retroperitoneal hematoma following LPB remains a complication with unknown incidence. Although we were fortunate that neither of our patients suffered neurologic injury, 2we would like to reemphasize that retroperitoneal hematoma should be considered a significant complication—one that can result in major morbidity and death. 3In addition to retroperitoneal hematoma following LPB, Auroy et al. 4reported a rate of other serious complications following LPB that exceeds that of other peripheral nerve blocks. Finally, Chelly et al. may have collectively performed a large number of continuous LPBs, but they provide no evidence to support their “more measured” recommendation that the lumbar plexus approach should be aborted if “significant bleeding” occurs during placement. How is “significant bleeding” defined, and how is this recommendation any more measured than our current management? We stand by our conclusion that retroperitoneal hematoma following LPB is a serious complication whose incidence is unknown.