Abstract
Introduction
Robotic-arm assisted knee arthroplasty (rKA) has been associated with improved clinical, radiographic, and patient-reported outcomes. There is a paucity of literature, however, addressing its cost effectiveness. In the context of an integrated health system with an insurance plan and single source comprehensive data warehouse for electronic health records and claims data, we present an evaluation of healthcare costs and utilization associated with manual knee arthroplasty (mKA) versus rKA. We also examine the influence of rKA technology on surgeons’ practice patterns.
Methods
Practice patterns of KA were assessed 18 months before and after introduction of robotic technology in April 2018. For patients also insured through the system's health plan, inpatient costs (actual costs recorded by health system), 90-day postoperative costs (allowed amounts paid by insurance plan), and 90-day postoperative utilization (length of stay, home health care visits, rehabilitation visits) were compared between mKA and rKA patients, stratified by total (TKA) or unicompartmental (UKA) surgery. Linear regression modeling was used to compare outcomes between the two pairs of groups (mKA vs. rKA, for both UKA and TKA). Log-link function and gamma error distribution was used for costs. All analyses were done using SAS statistical software, with p<0.05 considered statistically significant.
Results
Overall KA volume increased 21%, from 532 cases in the pre-rKA period to 644 post-rKA introduction, with UKA surgeries increasing from 38 to 97 (155%). Of these KAs, 218 patients were insured through our system's health plan (38 rUKAs, 9 mUKAs, 91 rTKAs, and 80 mTKAs), allowing precise insurance claims analysis for postoperative utilization and cost. Patients with rKA had significantly lower mean home health costs (-90% difference for UKA, −79% difference for TKA, p<0.02) and home rehab costs (-64% difference for UKA, −73% difference for TKA, p≤0.007) than mKA patients. No significant differences were observed in outpatient rehab (visits or costs), total rehab costs, or length of stay. Mean total postoperative costs were significantly lower for rUKA than mUKA (-47% difference, p=0.02) but similar for TKA (p>0.05). There were no significant differences in total inpatient costs between MAKO and non-MAKO patients.
Conclusion
Robotic-arm assisted KA can allow for increased UKA volume and potential for substantial cost savings over the total episode of care by reducing postoperative utilization and costs.
