Approximately 300,000 persons in the US receive total knee replacement (TKR) annually, and more than 90% experience dramatic symptom relief. These excellent results have been reported from high volume centers, yet research on cardiac and other procedures shows that centers and surgeons with lower caseloads have higher rates of complications and mortality. We showed that patients who received primary and revision total hip replacement (THR) in high volume hospitals have lower rates of postoperative mortality and complications than those treated in lower volume settings. Also, patients who had revision THR in higher volume hospitals had significantly better functional status than those treated in low volume hospitals. Associations between total knee replacement volume and outcomes have not been studied extensively, yet plans to regionalize, or restrict TKR to select high volume centers are underway.
We aim to determine whether surgeon and hospital volume influence the outcomes of TKR and to identify aspects of the process of care or care setting that explain the effect of volume on outcome. We will analyze Medicare claims data for all beneficiaries who had primary or revision TKR in 2000. We will determine whether patients operated upon in higher volume centers or by higher volume surgeons had lower rates of perioperative mortality and complications, after adjusting for demographic factors and co-morbidity. We will survey a random sample of patients two and five years post-operatively to determine whether hospital or surgeon volume are associated with functional status, pain relief or satisfaction following TKR. We will also review these patients' medical records and survey the acute and rehabilitation hospitals where they received care. Using these data we will determine whether aspects of the process or care setting explain the effect of volume on outcome. Our findings will provide an evidence base for clinical policy decisions regarding regionalization f TKR and for quality improvement initiatives to optimize TKR outcomes.
Showing the most recent 10 out of 444 publications
