Neuraxial Anesthesia
We suggest the use of neuraxial (such as intrathecal, epidural, and spinal) anesthesia for patients undergoing elective hip or knee arthroplasty to help limit blood loss, even though evidence suggests that neuraxial anesthesia does not affect the occurrence of venous thromboembolic disease.

Rationale
There is one high quality study and two moderate quality studies that addressed neuraxial anesthesia and VTE disease (Table 37 summarizes their results, Table 38 through Table 44 present a detailed description of their results, and Table 55 in Appendix XIII summarizes the results of our quality and applicability evaluations). None of these studies found a statistically significant difference in outcomes between regional (epidural or spinal) and general anesthesia.

Fifteen randomized controlled trials of high quality and moderate applicability compared peri-operative blood loss among patients receiving general, epidural, or a combination of general and epidural, or a combination of general anesthesia and lumbar plexus block.  There were eight high quality studies comparing epidural and general anesthesia.  Epidural anesthesia resulted in lower intra-operative blood loss. The combination of epidural and general anesthesia resulted in lower intra-operative blood loss compared to general anesthesia alone in two high quality studies. The combination of lumbar plexus block and general anesthesia resulted in lower intra- and post-operative blood loss compared to general anesthesia alone in two high quality studies. Hypotensive epidural anesthesia resulted in lower post-operative blood loss compared to spinal anesthesia in two high quality studies.

Table 64 in Appendix XIV summarizes the reasons for excluding some of the studies we initially considered for this recommendation.