|Home | About | Journals | Submit | Contact Us | Français|
Survey studies have concluded that a lack of consensus exists between orthopaedic surgeons on indications for total hip and knee arthroplasty. Geographic variation in the rates of these operations has raised concerns that some surgeons inappropriately indicate healthier patients for surgery than others. The objective of this study was to compare primary hip and knee arthroplasty patients’pre-operative validated outcome scores between four orthopaedic surgeons operating at a single academic institution from 2003 to 2007.
A retrospective chart review was performed using CPT-4 codes to identify patients who underwent primary total hip or knee arthroplasty at our institution between June 2003 and June 2007. Pre-operative SF-36 and WOMAC scores were recorded for each patient Patient demographics including age, gender, body mass index (BMI), number of co-morbidities, life orientation score (a measure of patient optimism), smoking and alcohol use, education level, and occupation were also recorded. Statistical analysis using unbalanced analysis of variance (ANOVA) and Chi-Square test were used to compare data between the surgeons, with statistical significance set at P < 0.05.
There was no statistically significant difference in SF-36 or WOMAC stiffness and function scores between the surgeons. There was a small difference in WOMAC pain scores between the surgeons’total knee patients, but not total hip patients. The number of primary hip and total knee replacements performed by each surgeon ranged from 151 to 955, with a total of 1896 primary joint replacements by the four surgeons during the study period.
Patients undergoing primary total joint arthroplasty at our institution were equally disabled between four surgeons, despite the surgeons performing variable numbers of the procedures. Further comparative effectiveness research using validated outcome measures is warranted.
542,000 total knee replacements and 231,000 total hip replacements were performed in the United States in 2006.1,2 Despite these numbers and an increasing prevalence of these operations over recent years,3 at least three survey studies have demonstrated disagreement among orthopaedic surgeons on the indications for total hip and knee arthroplasty.4,6 Although orthopaedic surgeons may differ in their survey responses regarding the indications for these surgeries, no study has demonstrated a difference in patient-based measures of disability between different total joint arthroplasty surgeons. Variation in the rates of total hip and knee arthroplasty (and other common surgical procedures) between geographic regions is well documented in the literature, and has been one area of focus in the debate over Health Care Reform in the United States.7,8 An important question to inform this debate is:“Are surgeons who perform more of these operations indicating healthier patients for surgery than surgeons who do them less frequently?”The literature provides no clear answer to this question at present. The purpose of this study was to investigate this question at our institution by comparing pre-operative patient-based measures of disability between four total joint surgeons operating at a single university hospital over a four year period.
After obtaining IRB approval, a retrospective chart review was performed using CPT-4 codes to identify patients who underwent primary total hip or knee arthroplasty at our institution between June 2003 and surgeons were arbitrarily designated Surgeon A, B, C, or D. Surgeons A and D are primarily total hip and knee arthroplasty surgeons, while Surgeons B and C devote significant portions of their practice to other orthopaedic procedures.
Pre-operative SF-36 and WOMAC scores were recorded for each patient. Patient demographics including age, gender, body mass index (BMI), number of co-morbidities, life orientation score (a measure of patient optimism), smoking and alcohol use, education level, and occupation were also recorded. Education level was classified into one of three groups: some high school or less, high school graduate, or any post-graduate work. Occupation was also classified into three groups: home-maker/retired/unemployed, skilled labor, and professional/managerial/sales/clerical/student.
Separate analyses were undertaken for total hip patients and total knee patients. Unbalanced analysis of variance (ANOVA) and Chi-Square test were used to compare data between the surgeons. Fisher's exact test substituted Chi-Square test in case that some cell has an expected value fewer than 5. For all analyses, a P value less than 0.05 was considered statistically significant. Data were analyzed using SAS software (version 9.1.3; SAS Institute, Cary, NC).
Finally, we compared the age and gender data of our patient cohort with 2006 National Data available from the American Academy of Orthopaedic Surgeons website to see if any differences exist between our institution and the national average.
From 2003 to 2007, Surgeons A, B, C, and D performed a total of 646, 579,151 and 955 total joint replacements, respectively. Of these, 82, 81, 78 and 82 percent were primary procedures (see Table 1). Pre-operative SF-36 scores were available for 65% of patients, while WOMAC scores were available for 44% of patients.
Among total knee arthroplasty patients, there was no statistically significant difference between the surgeons in SF-36 scores (PCS and MCS components), or WOMAC stiffness and function scores. Surgeon C had significantly higher WOMAC pain scores than Surgeons D and B. In terms of patient demographics, no statistically significant difference was found between the surgeons in male to female ratio, education, occupation, smoking and number of musculoskeletal comorbidities. Life orientation scores were higher for Surgeon A than Surgeons C and D, and Surgeon B scores were higher than Surgeon C. Surgeon B had higher age, lower BMI, and higher rate of alcohol use compared to the other surgeons (Table 2).
Among total hip arthroplasty patients, there were no statistically significant differences between the surgeons in SF-36 scores (PCS and MCS) or WOMAC pain, stiffness, and function scores. In terms of patient demographics, no statistically significant difference was found between the surgeons in male-to-female ratio, education, smoking, alcohol use, life orientation scores, and musculoskeletal co-morbidities. Surgeon C operated on a significantly higher proportion of homemaker/ retired/unemployed patients than the other surgeons, and also had a higher number of systemic comorbidities. Surgeon C also had a higher BMI than Surgeons A and B. Surgeon B's patients were significantly older than Surgeon D (Table 2).
Consensus statements on indications for total hip and knee arthroplasty have been developed and published by the National Institute of Health.9,10 In spite of these efforts, several studies have shown that disagreement exists between orthopaedic surgeons on indications for total hip and knee arthroplasty. In a survey of orthopaedists in the New York City area, Mancuso showed moderate agreement (greater than 50%) was noted on 5 of 9 variables for total hip arthroplasty, and 6 of 10 variables for total knee arthroplasty.6 Wright et al. surveyed 234 orthopaedic surgeons in Ontario, and found the respondents to disagree on how 20 of 34 patient characteristics affected their decision to perform knee replacement surgery.5 Tierney et al. showed strong agreement (>95%) on 7 of 33 surgical indications among 220 orthopaedic surgeons surveyed in the State of Indiana.4 Cross et al. performed a systematic review of the literature on indications for total knee arthroplasty and found that“pain not responsive to drug therapy”was the only patient factor of over 27 reviewed on which there was greater than 90 percent consensus.11
Wennberg and others have argued that,“in the absence of professional consensus based on outcomes, individual physicians can hold onto idiosyncratic clinical rules of thumb defining who needs surgery.”This“surgical signature”phenomenon has been postulated to explain dramatic variation in the rates of common surgical procedures, including total hip and knee arthroplasty between neighboring geographic regions.12,14 The assumption has been made that orthopaedic surgeons in regions with high rates of total hip and knee arthroplasty may be inappropriately indicating patients for surgery. This assumption has never been proven in the literature. In a retrospective chart review of elective primary hip and knee replacement patients from low-rate and high rate regions of Ontario, Canada, an expert panel of four orthopaedic surgeons, two rheumatologists, two family physicians, an internist, epidemiologist and a physiotherapist concluded that inappropriate use of these surgeries does not account for the high rate of total joint replacement in these regions.15 Interestingly, Hawker and others demonstrated that both the“potential need”for total joint arthroplasty and“patient willingness”to undergo surgery were greater in a region with a high rate of these surgeries when compared to a low rate region.16 Multiple authors have argued that total joint arthroplasty is, in fact, an underutilized procedure, since: a) there is a significant unmet need for total joint arthroplasty among those who might benefit, b) the surgeries carry a low risk of morbidity and mortality, c) over 90% of patients experience substantial pain relief and improvements in quality of life, and d) these procedures have been demonstrated to be cost-effective.16,17
An important question which, heretofore, has not been addressed in the literature, is:“Are total joint surgeons operating on the same patients?”Put another way, does disagreement in orthopaedic surgeons’survey responses about indications for total joint arthroplasty, or differences in the number of patients they indicate for these procedures mean that some surgeons are indicating‘healthier’patients for surgery than others? At several hospitals, including our own, valid, reliable patient-based measures of overall health and condition-specific disability such as the SF-36 and WOMAC are included in the medical record. These data provide a reasonable gauge for comparison of patients indicated for surgery by different orthopaedic surgeons.
At our institution, we found patients’pre-operative general health (SF-36) and hip and knee arthritis specific outcome measures (WOMAC scores) to be essentially the same between the four total joint surgeons. While the retrospective nature of this study prevents us from knowing which patient factors each surgeon used to indicate patients for total joint arthroplasty during the study period, or the relative importance of those factors to the surgeon, this study shows that the patients indicated for total joint arthroplasty by different surgeons at our institution were equally disabled.
There are several limitations to our study. First, this comparison was performed at a single academic institution; therefore, these results cannot be generalized to other institutions, nor have we compared surgeons operating in different institutions. Second, this study was a retrospective chart review, and a power analysis was not undertaken to determine the number of patients needed to detect clinical significance. Third, we had incomplete data on some patients in our study population, which may have biased our results. Finally, this study does not correct for any referral bias that may have existed between the surgeons.
Despite these limitations, this study is important for several reasons. First, this study demonstrates that orthopaedic surgeons who perform more total joint arthroplasty operations than their colleagues do not necessarily operate on healthier patients. Second, this study highlights the need for, and the research possibilities opened by, the creation of a National Total Joint Arthroplasty Registry. Finally, in the context of $1.1 billion dollars recently allocated for Comparative Effectiveness Research by the passage of the American Recovery and Reinvestment Act of 2009 , this study serves as a model for designing and funding future studies comparing patient-based measures of disability between healthcare providers.
This project was supported by the Bierbaum Research Fund.
The authors would like to thank Yubo Gao, PhD, for performing the statistical analysis for this project, and Alison Klaassen and Joyce Woody for their assistance in data collection.