Analgesia for Total Knee Arthroplasty: What Is the Best Option?
Our concern, and one that has been raised before,2 is that the current body of literature has not resolved this question, as results of studies comparing different techniques are variable and, seemingly, dependent on study design. In the words of Kehlet and Andersen,2 many of the studies, “lack sufficient blinding, have methodological inadequacies with method of pain assessment and the quality and composition of the continuous epidural or peripheral nerve analgesia technique or lack of comparable systemic analgesia between the groups.” The concerns related to study design revolve around 2 issues: not controlling for the systemic analgesic effects of medications in the periarticular infiltration (PI) and/or not controlling for the effect of the sciatic nerve. As noted by Abdallah et al,3 not controlling for the sciatic nerve will lead to an increase not only in posterior knee pain but also in anterior knee pain. This can confound the results of studies because without controlling for analgesia of the sciatic nerve, the PI (providing analgesia in both femoral and sciatic territories) is being compared with a technique (either femoral or adductor) that only provides analgesia in the femoral nerve distribution.
Unfortunately, this study appears to suffer from both of these issues. First, the PI group received 10 mg of morphine (equivalent to 2 mg of hydromorphone) and 30 mg of ketorolac that was not given to the AC-only group. If the amount of morphine in the PI is added to the PCA totals, then the total hydromorphone dose would be equal (7 mg each) for the PI alone and AC-only groups. If the morphine-sparing effects of ketorolac (equivalent to 5–10 mg of morphine) are added to the PI alone group, then the PI alone group needed more analgesics than the ACB-alone group—the opposite of the conclusion presented. Second, the design of the study did not control for the effect of the sciatic nerve because the ACB alone group received neither a sciatic block nor a posterior injection.
Moving forward, perhaps more studies based on the design of Carli et al,4 which control for the systemic analgesic effects of medication in the PI as well as for the effect of the sciatic nerve, will be able to give us the answer to this important question.