Reduced Early Revision Surgery and Medical Complications in Computer-assisted Knee Arthroplasty Compared With Non–computer-assisted Arthroplasty

BACKGROUND: Computer assistance can help surgeons achieve mechanical accurate alignment, but the clinical effect of this technology in different arthroplasty types remains controversial because of conflicting functional outcomes, revision rates, and complication rates. The goal of this study was to compare 90-day medical complications and 1 and 2-year revision surgeries after computer-assisted patellofemoral arthroplasty, unicompartmental knee arthroplasty (CA-UKA), bicompartmental knee arthroplasty (CA-BKA), and total knee arthroplasty (CA-TKA) with non-computer-assisted procedures. METHODS: A retrospective cohort analysis was conducted using the Mariner data set of the PearlDiver Patient Records Database from 2010 to 2018. Current Procedural Terminology codes were used to identify patients who underwent patellofemoral arthroplasty, unicompartmental knee arthroplasty, bicompartmental knee arthroplasty, and total knee arthroplasty with or without computer assistance for osteoarthritis. All included patients were followed up for 2 years. Univariate and multivariable analyses were conducted. RESULTS: On multivariable analysis, revision within 1 year was decreased in patients who underwent computer-assisted patellofemoral arthroplasty (odds ratio [OR]: 0.541, P = 0.031), CA-UKA (OR: 0.798, P = 0.019), and computer-assisted bicompartmental knee arthroplasty (OR: 0.186, P = 0.025) compared with the same surgeries without technology assistance. CA-TKA had decreased odds of revision for aseptic loosening at 2 years (OR: 0.789, P < 0.001). CA-UKA and CA-TKA had decreased risk of overall 90-day medical complications (OR: 0.838, P < 0.001; OR: 0.903, P < 0.001, respectively) and major complications (OR: 0.750, P = 0.004; OR: 0.822, P < 0.001, respectively). DISCUSSION: Complication rates and revision surgeries for all arthroplasty types were equivalent or more favorable when computer assistance was used during surgery. Our results quantify some early benefits to using computer assistance in total and partial knee arthroplasties.