Adverse Events Associated With Robotic-Assisted Joint Arthroplasty: An Analysis of the US Food and Drug Administration MAUDE Database.

ABSTRACT

BACKGROUND: The use of robotic assistance in arthroplasty is increasing; however, the spectrum of adverse events potentially associated with this technology is unclear. Improved understanding of the causes of adverse events in robotic-assisted arthroplasty can prevent future incidents and enhance patient outcomes. METHODS: Adverse event reports to the US Food and Drug Administration Manufacturer and User Facility Device Experience database involving robotic-assisted total hip arthroplasty (THA), total knee arthroplasty (TKA), and partial knee arthroplasty were reviewed to determine causes of malfunction and related patient impact. RESULTS: Overall, 263 adverse event reports were included. The most frequently reported adverse events were unexpected robotic arm movement for TKA (59/204, 28.9%) and retained registration checkpoint for THA (19/44, 43.2%). There were 99 reports of surgical delay with an average delay of 20 minutes (range 1-120). Thirty-one cases reported conversion to manual surgery. In total, 68 patient injuries were reported, 7 of which required surgical reintervention. Femoral notching (12/36, 33.3%) was the most common for TKA and retained registration checkpoint (19/28, 67.9%) was the most common for THA. Although rare, additional reported injuries included femoral, tibial, and acetabular fractures, MCL laceration, additional retained foreign bodies, and an electrical burn. CONCLUSION: Despite the increasing utilization of robotic-assisted arthroplasty in the United States, numerous adverse events are possible and technical difficulties experienced intraoperatively can result in prolonged surgical delays. The events reported herein seem to indicate that robotic-assisted arthroplasty is generally safe with only a few reported instances of serious complications, the nature of which seems more related to suboptimal surgical technique than technology. Based on our data, the practice of adding registration checkpoints and bone pins to the instrument count of all robotic-assisted TJA cases should be widely implemented to avoid unintended retained foreign objects.