The Journal of Bone and Joint Surgery

HOME	HELP	FEEDBACK	SUBSCRIPTIONS	ARCHIVE	SEARCH

Commentary & Perspective

Commentary & Perspective on
"Association Between Hospital and Surgeon Procedure Volume and the Outcomes of Total Knee Replacement"
by Jeffrey N. Katz, MD, MS, et al.

Commentary & Perspective by
Daniel J. Berry, MD*,
Mayo Clinic, Rochester, Minnesota

Katz et al. report important and valuable information about the relationship between the number of knee arthroplasties performed in hospitals or by surgeons and the risk of early complications related to primary total knee arthroplasty performed on patients receiving Medicare in the United States. The authors reported that, for many complications, it was the patients who were treated in hospitals and by surgeons performing a lower annual volume of total knee arthroplasties in the Medicare population who experienced a higher average rate of many early complications. These findings parallel the findings that the same group previously reported with regard to total hip arthroplasty and suggest that patients who are treated in hospitals and by surgeons performing less than a critical number of these procedures annually, on average, are at higher risk of experiencing some early complications of total knee arthroplasty.

The study's conclusions are powerful, in part because of the strong methodology used. Eighty thousand, nine hundred and four patients were included in the study, all of whom had surgery in a short time frame between January 1 and August 31, 2000. The study was inclusive, and patients who were analyzed were treated in all geographic areas of the United States. The large number of patients allowed the authors to perform a meticulous multivariate statistical analysis that allowed them to adjust for some comorbidities and for age, gender, and type of arthritis.

The important findings of the study can be summarized as follows:

Eleven percent of the primary total knee arthroplasties performed in Medicare beneficiaries took place in hospitals that performed twenty-five or fewer such procedures annually, and 25% of the primary total knee arthroplasties performed in Medicare beneficiaries were done by surgeons who performed twelve or fewer such procedures annually.

Effect of hospital volume: The risk of mortality was about 30% higher for the patients who had total knee arthroplasty in hospitals that performed twenty-five or fewer Medicare total-knee arthroplasties per year compared with hospitals that had higher volumes. Above a threshold of twenty-five cases, there was no statistical association between mortality and increasingly higher volume. However, the risk of postoperative pneumonia and early deep infection declined steadily with increasing volumes, even >25 per year. Patients who had the procedure in the highest volume hospitals had about a 25% lower risk of experiencing myocardial infarction, pneumonia, or pulmonary embolism compared with the patients at lower-volume centers.

Effect of surgeon volume: Patients who were treated by the highest volume surgeons (those performing >50 total knee arthroplasties per year in Medicare beneficiaries) had lower rates of knee infection (by approximately 40%) compared with patients who were treated by surgeons who performed twelve or fewer such procedures per year. The risk of pneumonia also was lower for patients treated by high-volume surgeons, as was the combined rate of any major medical complication. Surgeon volume was not associated with the risk of mortality, acute myocardial infarction, or pulmonary embolism, after adjustment for hospital volume.

Notably, mortality was associated with hospital but not surgeon volume, while other complications, such as pneumonia and deep infection, were associated with both hospital and surgeon volume independently. This led the authors to conclude, probably correctly, "These observations suggest that the rate of mortality may reflect hospital factors such as the quality and intensity of anesthesia care, nursing, and other services. Deep infection and pneumonia, on the other hand, may arise both from hospital factors and from processes largely under the surgeon's control, such as the surgical technique and the duration of the procedure."

Methodologic limitations of the study preclude analysis of many important aspects of the effects of hospital and surgeon volume on the outcome of total knee arthroplasty. Even though the paper is entitled "Association Between Hospital and Surgeon Procedure Volume and the Outcomes of Total Knee Replacement," in actuality the authors have examined the association between volume and early complications, not outcome of the procedure in a more general sense. The study does not provide information on the effect of volume on functional outcome of the procedure, or on longer-term durability of the procedure clinically or radiographically, or on the rate of reoperation after the procedure—all of which are outcomes of great importance to the patient and the surgeon. While the data presented in this paper hint that those outcomes might also be linked to surgeon and/or hospital volume, that hypothesis remains unproved. Additionally, an important limitation of the study is the assumption that any differences among patients who have been treated at high and low-volume centers and by high and low-volume surgeons can be accounted for in a multivariate model. The model used in this paper is limited by variables collected in the Medicare dataset; hence, subtle differences between patients who were treated in different settings may not be considered, thus biasing the results.

The authors rightly point out that even though the statistical associations they identified were strong, the overall rate of early complications resulting from primary total knee arthroplasty in all settings was low. Thus the actual (not percent) differences from low to high-volume surgeons and hospitals were small in magnitude and, consequently, the number of patients that needed to be treated to produce a single extra complication in a low compared with a high-volume center was large.

The data published in this paper, taken together with the data published previously by these and other authors regarding hip and knee arthroplasty, make a compelling case that total joint replacement can be performed with a lower risk of early complications in centers and by surgeons performing the operations with sufficient frequency to gain and maintain expertise in the operative procedure and perioperative care of these patients. How these data are used may have important implications for patients seeking hip or knee arthroplasty, for hospitals and surgeons performing these operations, and for government and private insurers that pay for these operations.

The authors, wisely, are careful to avoid overinterpreting their data and do not make suggestions about how their data should be used in healthcare policy. In the closing paragraph of their article, the authors comment on the unknown potential burden for patients who must travel longer distances to higher-volume centers. "In fact, we do not know whether patients who shift from low to high-volume centers decrease their risk to the extent suggested by our data. There may be disadvantages associated with having surgery in a distant and unfamiliar center that attenuate these risk differences." This paper provides valuable information that informs surgeons, health-policy makers, and patients and suggests important areas for further discussion and study.

*The author did not receive grants or outside funding in support of his research or preparation of this manuscript. The author received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity (DePuy). In addition, a commercial entity (DePuy) paid or directed, or agreed to pay or direct, benefits to a research fund, foundation, educational institution, or other charitable or nonprofit organization with which the author is affiliated or associated.