Retrospective cohort study of the economic value of providing microprocessor knees to the population of Medicare fee-for-service K2 beneficiaries with a knee disarticulation/above knee amputation.

The study used Medicare claims, not clinical records, so it lacked direct measures of function, social determinants of health, and poverty level; billing codes can be incomplete or wrong. It did not compare K2 patients with microprocessor knees to K2 patients with nonmicroprocessor knees. Because Medicare did not cover microprocessor knees for K2 users, the comparison was K2 nonmicroprocessor users versus K3 microprocessor users, so K-level differences may remain. Statistical weighting can only balance variables available in claims data, not unmeasured differences between the groups. The 10-year budget savings are model projections based on assumed take-up, mortality, and replacement rates, not observed Medicare savings. The fall outcome counted beneficiaries with an injurious fall that led to a medical claim, not the number of falls or less serious falls. Bilateral prosthetic users and individuals covered by Medicare Advantage, commercial insurance, or Medicaid were not included.

Authors at Dobson|DaVanzo & Associates, LLC reported financial support from Otto Bock HealthCare Products GmbH, a company that sells microprocessor knees. The authors state the company had no role in study design, data collection, analysis, interpretation, writing, or submission, and the authors declared no conflicts of interest.

Do not read the $410.3 million 10-year savings as proof that covering microprocessor knees for K2 patients will save Medicare money. The study did not observe K2 patients using microprocessor knees; it compared K2 nonmicroprocessor users with K3 microprocessor users and used modeled assumptions about adoption, mortality, and device replacement.