Revision Rates After Knee Replacement in the United States

OBJECTIVES: Each year approximately 100,000 Medicare patients undergo knee replacement surgery. Patients, referring physicians, and surgeons must consider a variety of factors when deciding if knee replacement is indicated. One factor in this decision process is the likelihood of revision knee replacement after the initial surgery. This study determined the chance that a revision knee replacement will occur and which factors were associated with revision. METHODS: Data on all primary and revision knee replacements that were performed on Medicare patients during the years 1985 through 1990 were obtained. The probability that a revision knee replacement occurred was modeled from data for all patients for whom 2 full years of follow-up data were available. Two strategies for linking revisions to a particular primary knee replacement for each patient were developed. Predictive models were developed for each linking strategy. ICD-9-CM codes were used to determine hospitalizations for primary knee replacement and revision knee replacement. RESULTS: More than 200,000 hospitalizations for primary knee replacements were performed, with fewer than 3% of them requiring revision within 2 years. The following factors increase the chance of revision within 2 years of primary knee replacement: (1) male gender, (2) younger age, (3) longer length of hospital stay for the primary knee replacement, (4) more diagnoses at the primary knee replacement hospitalization, (5) unspecified arthritis type, (6) surgical complications during the primary knee replacement hospitalization, and (7) primary knee replacement performed at an urban hospital. CONCLUSIONS: Revision knee replacement is uncommon. Demographic, clinical, and process factors were related to the probability of revision knee replacement.