Abstract
BACKGROUND: Robotic assistance minimizes bony resection and surrounding soft tissue damage during total knee arthroplasty (TKA), potentially decreasing postoperative pain. The objective of this study was to evaluate in-hospital opioid consumption, 90-day opioid prescribing patterns, length of stay (LOS), and patient-reported outcome measures (PROMs) following robotic-assisted vs manual primary TKA. METHODS: Utilizing an institutional database, all patients undergoing primary unilateral TKA between 2019 and 2022 with 90-day minimum follow-up were retrospectively queried. Patients were excluded if they had another surgery within 90 days of the index TKA, were discharged to a rehabilitation center, or were prescribed opioid or benzodiazepine medications preoperatively. One-to-one propensity score matching identified 1476 patients undergoing robotic-assisted (n = 738) or manual (n = 738) TKA. Multivariable regression analysis assessed in-hospital morphine milligram equivalents (MMEs) consumption, LOS, 90-day opioid prescribing patterns (not necessarily consumption), and PROMs (preoperative, 6 weeks, and 3 months). RESULTS: The robotic-assisted cohort consumed on average 12 fewer in-hospital MMEs (P = .026) and had an average LOS that was 8 hours shorter than the manual cohort (P < .001). There was no difference in MMEs prescribed at discharge (P = .12), but the robotic-assisted cohort had on average 113 fewer 90-day postdischarge MMEs prescribed (P = .001) and 145 fewer total 90-day MMEs (consumed in-hospital plus prescribed) (P < .001). There was no difference in PROMs at 6 weeks or 3 months. CONCLUSIONS: Robotic-assisted TKA may confer shorter LOS with decreased 90-day opioid use patterns. This information is important given the increased scrutiny on opioid usage and recent focus on rapid recovery and ambulatory TKA pathways.