Almost 1,000,000 total hip arthroplasty (THA) and total knee arthroplasty (TKA) procedures, costing over $25 billion, are performed annually in the U.S. for end-stage arthritis of the hip and knee1. The number of procedures is expected to increase exponentially over the next several decades2 because of an aging population and the growth of THA and TKA utilization rates among younger patients3.
Despite the well-documented success of these procedures, complications do occur, and the rates of certain complications vary by practice setting. Cram et al. reported that unadjusted ninety-day composite complication rates (death, sepsis, hemorrhage, pulmonary embolism, deep venous thrombosis, or wound infection) in orthopaedic (specialty) and nonspecialty general hospitals in the U.S. were 2.8% and 6.2%, respectively, for THA and 2.1% and 3.8%, respectively, for TKA4. These differences remained significant after risk adjustment. In addition, there is documented variation in practice patterns, patient outcomes, and adherence to payer-defined and evidence-based practice guidelines for THA and TKA5, suggesting that there is room for quality improvement6.
Although these studies point out variations in THA and TKA complication rates among hospitals, prior investigators4,6 focused on comparing hospitals with specific characteristics (e.g., specialty compared with general) and did not study the overall variation in complication rates across U.S. hospitals. There is a need for a scientifically valid, risk-adjusted measure of complications that accurately reflects quality of care in all hospitals that perform THA and TKA procedures. Furthermore, because the results of some studies suggest that disparities in outcomes exist for disadvantaged patients7-9, it is important to illuminate whether minority and lower-income patients tend to undergo these elective procedures in hospitals with higher complication rates.
“Quality” measures currently used to evaluate elective THA and TKA procedures have generally been limited to processes of care, such as administration and discontinuation of antibiotics, administration of venous thromboembolism prophylaxis, and removal of indwelling urinary catheters following surgery10,11. Although such measures assess adherence to commonly accepted, evidence-based processes of care, they have limited discriminatory power for evaluating quality, given that providers typically achieve uniformly high performance.
Accordingly, we developed a hospital-level measure of the risk-standardized complication rate (RSCR) following THA and TKA through a transparent process involving formal review by a technical expert panel and public comment. The measure conforms to standards for publicly reported outcome measures12,13 and was reviewed and endorsed by the National Quality Forum (NQF)13,14. Beginning in December 2013, the Centers for Medicare & Medicaid Services (CMS) began publicly reporting hospitals’ RSCRs on its Hospital Compare web site15 as part of its Inpatient Quality Reporting program16.
The goals of this study were to use this NQF-endorsed quality measure to calculate hospital-level RSCRs for elective primary THA and TKA procedures performed in U.S. hospitals, describe variation in RSCRs, and explore how hospitals serving vulnerable populations perform on the measure relative to other hospitals.
Materials and Methods
The methodological approach for developing the RSCR measure is consistent with that used to develop other publicly reported CMS outcome measures17-21. The development and validation of the measure are described fully in a technical report22, and this study was approved by the Yale University institutional review board.
We used 2008 to 2010 Medicare Standard Analytic Files to identify index (eligible) admissions and complications (using inpatient claims) and comorbidities for risk adjustment (using inpatient and outpatient claims). We used the Medicare Enrollment Database to determine Medicare Fee-for-Service enrollment status, death, and race (black or other), and we used the American Hospital Association Annual Survey Data for the 2009 fiscal year to determine the proportion of Medicaid beneficiaries at each hospital.
The measure combines THA and TKA procedures because both are performed in clinically similar patient cohorts and for similar indications (lower-extremity arthritis), the same surgeons frequently perform both procedures, our analyses revealed similar rates and types of complications and duration of stay among the two patient cohorts, and combining procedures provides greater statistical power.
The measure includes admissions of patients, sixty-five years of age and older, who underwent elective primary THA (ICD-9-CM [International Classification of Diseases, 9th Revision, Clinical Modification] 81.51) and/or TKA (ICD-9-CM 81.54). To identify a homogeneous cohort of patients, we excluded patients with nonelective procedures such as those to treat a fracture of the femur, hip, or pelvis or a malignant neoplasm involving the bone; patients admitted with a mechanical complication (e.g., mechanical loosening, dislocation, implant breakage, or periprosthetic fracture); and patients undergoing an arthroplasty-related procedure with an associated complication risk that differs substantially from that in an elective THA or TKA population (e.g., patients undergoing revision, partial hip replacement, hip resurfacing, or removal of implanted devices or prostheses) (as detailed in Appendix B of the technical report22). The measure also excludes admissions of patients who had incomplete administrative data, left the hospital against medical advice, transferred into the hospital that performed the procedure, or had more than two THAs or TKAs coded during the index admission. If patients had more than one primary THA or TKA in one year, we randomly selected one admission for inclusion.
On the basis of the published literature23-35 and expert clinical input, we selected complications that could be attributable to THA or TKA procedures and subsequent hospital care and that were identifiable in claims data (Table I). The binary outcome for the measure is one or more (compared with none) of the following medical or surgical complications occurring during the index admission or resulting in a readmission within the specified time period: acute myocardial infarction, pneumonia, or sepsis within seven days of admission; pulmonary embolism, surgical site bleeding, or death within thirty days of admission; or mechanical complications, periprosthetic joint, or wound infection within ninety days of admission. All complications were identified with use of ICD-9-CM diagnosis codes (as shown in Table 1 in the technical report22). We counted a periprosthetic joint infection, wound infection, or surgical site bleeding only when at least one accompanying ICD-9-CM procedure code for an intervention (e.g., irrigation and debridement or evacuation of a hematoma) was also recorded during the same admission22,36.
Risk Adjustment Variables
To account for a hospital’s case mix, the measure adjusts for patient age, sex, procedure type (THA or TKA) and number (one or two), history of comorbidities, and comorbidities present during the index admission that do not reflect a potential complication (Table II). To examine the ability of the model to accurately predict the likelihood that an event will occur for patients with various levels of risk (predictive ability), we compared the observed complication rate with the predicted complication rate across ten deciles of risk.
The measure uses hierarchical logistic regression to estimate hospital-level complication rates as a function of the risk variables. This approach accounts for the clustering (nonindependence) of patient outcomes within the same hospital and for variation in sample size across hospitals while risk-adjusting for differences in the patient case-mix37. It simultaneously models two levels (patient and hospital) to account for the variance in patient outcomes within and between hospitals.
The RSCR is the ratio of the number of “predicted” patients with one or more complications to the number of “expected” patients with one or more complications for each hospital, multiplied by the national complication rate (among all patients included in the measure). The predicted-to-expected ratio is similar to an observed-to-expected ratio. For ease of interpretation, we multiply the ratio by the national complication rate to transform the ratio into a rate. A hospital’s RSCR value can best be interpreted by comparing it with the national complication rate; if the RSCR for that hospital is 2% and the national rate is 3.6%, the hospital is performing better than would be expected, given its patient mix. The detailed statistical model is presented in the technical report22.
Hospital Performance Categories
We identified hospitals with twenty-five or more patients admitted for THA or TKA from 2008 to 2010 and classified these hospitals into three performance categories according to their RSCR, using the approach used for other CMS outcome measures posted on the Hospital Compare web site15. We excluded hospitals with less than twenty-five eligible admissions because the number of cases would generally be insufficient to detect a quality signal for these hospitals.
We used bootstrapping to empirically construct a 95% interval estimate (similar to a confidence interval [CI]) for each RSCR. Whereas the RSCR represents the best estimate of a hospital’s true performance, the 95% interval estimate represents the range of rates that encompasses the hospital’s true rate with a very high degree of probability. We classified a hospital as performing better than the U.S. national rate if the interval estimate for that hospital was entirely below the national complication rate, worse if the estimate for that hospital was entirely above the national rate, and no different if the estimate included the national rate.
To assess potential disparities in the quality of care provided by hospitals caring for vulnerable populations, we categorized hospitals with at least twenty-five admissions for THA or TKA in the three-year sample according to the proportion of patients with low socioeconomic status, defined as the hospital’s proportion of patients enrolled in Medicaid in the 2009 American Hospital Association data38. We compared the RSCR distribution for hospitals in the lowest decile (≤7% Medicaid patients) with the distribution for hospitals in the highest decile (≥28% Medicaid patients). To examine potential disparities in RSCRs among hospitals on the basis of the racial makeup of the patients they serve, we also categorized hospitals according to the proportion of black patients and compared hospitals in the lowest decile (0% black patients) with those in the top decile (≥19% black patients).
All analyses were conducted with use of SAS software (version 9.2; SAS Institute, Cary, North Carolina). The hierarchical models were estimated with use of the SAS GLIMMIX procedure.
Source of Funding
The analyses on which this report is based were funded by the CMS, a part of the Department of Health and Human Services (contract HHSM-500-2008-0025I/HHSM-500-T0001). The content of this report does not necessarily reflect the views or policies of the Department of Health and Human Services. The authors assume full responsibility for the accuracy and completeness of the ideas presented.
The study cohort included 878,098 Medicare Fee-for-Service beneficiaries, sixty-five years of age or older, who underwent elective primary THA or TKA from 2008 to 2010 at 3479 hospitals (Table I; see also Fig. 1 of the technical report22). The median number of patients per hospital was 130 (interquartile range, thirty-nine to 332). The procedure involved unilateral THA in 28.5% of the patients, simultaneous bilateral THA in 0.1%, unilateral TKA in 68.4%, simultaneous bilateral TKA in 3.0%, and simultaneous THA and TKA in 0.01%. The median patient age was seventy-four years, 64% were female, and 7.5% were non-white. The most common reason for exclusion was incomplete administrative data (n = 115,632).
The median RSCR was 3.6% (range, 1.8% to 9.0%), and the interquartile range was 3.2% to 3.9% (Fig. 1). The most common complications were pneumonia (0.86%), pulmonary embolism (0.75%), and periprosthetic joint infection or wound infection (0.67%) (Table I). The odds of a complication were 1.93 times higher if treatment was at a hospital with a complication rate one standard deviation above the national average compared with one standard deviation below the national average.
Hospital Performance Categories
Of 2832 hospitals performing at least twenty-five THA and TKA procedures from 2008 to 2010, seventy-five (2.6% of the hospitals) had an RSCR that was statistically worse than the U.S. national rate (i.e., a 95% interval estimate completely above the national rate) and 103 (3.6%) had an RSCR statistically better than the U.S. national rate (i.e., a 95% interval estimate completely below the national rate).
Risk Factors for Complications and Model Performance
The three risk factors associated with the greatest risk of a complication were protein-calorie malnutrition (odds ratio [OR], 2.7; 95% CI, 2.5 to 2.9), end-stage renal disease or dialysis (OR, 1.7; 95% CI, 1.4 to 2.0), and the number of procedures (OR for two compared with one, 1.7; 95% CI, 1.6 to 1.8). The c statistic for the model was 0.63, and model predictive ability ranged from 2% in the lowest decile of patient risk to 8% in the highest decile.
Figure 2 demonstrates considerable overlap in the RSCR distributions of hospitals with the highest decile of Medicaid patients (RSCR range, 2.0% to 7.1%) and hospitals with the lowest decile (1.7% to 6.2%) (see also Table III). The absolute difference between the median RSCRs of the highest decile (3.6%) and lowest decile (3.4%) was 0.2%. Figure 3 demonstrates similar results when patient race rather than Medicaid enrollment was examined; the median RSCR of hospitals with the highest proportion of black patients was 0.3% higher than that of hospitals with the lowest proportion, and the ranges (2.0% to 7.4% and 2.0% to 6.2%, respectively) largely overlapped (Table III).
The hospital-level RSCRs for Medicare Fee-for-Service patients who underwent THA or TKA ranged from 1.8 to 9.0%22, suggesting that there are differences in quality of care across U.S. hospitals that perform these procedures. More than 175 hospitals were better or worse than expected when a statistically conservative approach (the 95% interval estimate) was used to identify outliers. Although the complication rates are low, the differences are clinically meaningful, and small reductions in the complication rate will have a meaningful effect population-wide. Our RSCR measure for THA and TKA procedures, developed with input from clinical leaders in orthopaedic surgery, could serve to stimulate efforts to improve quality of care and reduce variation in outcomes.
Previous authors have reported hospital-level complication rates of 7% to 8%39,40 for patients undergoing THA and TKA. The lower median hospital-level complication rate of 3.6% in our cohort may be explained by several factors. First, by including only patients who underwent elective primary THA or TKA and by excluding patients at higher risk (e.g., those undergoing revision or partial THA or resurfacing and those with a hip or pelvic fracture), we selected a cohort of patients at lower risk for complications23,41-47. Second, our outcome measure includes a limited number of well-defined complications occurring during the index admission or requiring a readmission, and it may therefore have underestimated the rate of less severe complications treated in the outpatient setting. Moreover, the measure does not include complications (e.g., deep venous thrombosis and urinary tract infection) with variable screening and diagnosis rates, as these are inconsistent markers of quality. Although such complications are still of concern to patients and providers, our measure identifies the most clinically important and costly complications that can be reasonably attributed to the index surgical procedure, and it therefore reduces the potential for false positive findings.
In contrast to prior studies7-9, we also found similar RSCR ranges among hospitals serving the highest proportions of Medicaid or black patients. Although the mean complication rates were slightly higher than those for hospitals with the lowest proportion of Medicaid or black patients, suggesting that a measureable disparity exists, there was considerable overlap between the distributions. Many hospitals with the highest proportion of vulnerable patients performed well on the measure, indicating that hospitals caring for vulnerable populations can achieve low complication rates. As this measure does not adjust for socioeconomic status or race and therefore should illuminate, rather than mask, quality differences across populations, it can be used to monitor for changes in disparities of care over time.
Our study has several limitations. First, our approach to classifying hospitals as outliers identified only a small proportion of hospitals as better or as worse than the U.S. national rate (3.6% and 2.6%, respectively). We chose a classification approach designed to limit the potential for misclassifying hospitals as outliers, and we cannot say with certainty that those classified as no different are truly typical in their performance. Furthermore, the measure is designed to not only classify performance but drive quality improvement. To promote quality improvement at all hospitals, not only those identified as worse than average, CMS provides detailed patient-level data for all patients included in the measure score calculation. Second, the model was developed with use of administrative data, which may overestimate or underestimate comorbidities and complications because of the limitations of administrative data and variations in coding practices across institutions. However, we selected complications that are less likely to be influenced by variations in coding practices, and we counted wound infection, periprosthetic joint infection, and surgical site bleeding as complications only if they were associated with an accompanying surgical intervention. Additionally, we conducted a medical record validation of the claims codes used to identify complications, and we found 99% agreement (635 of 644) after refining the measure specifications22. In addition, we used an approach to selecting and defining claims-based risk adjustment variables that has been validated for similar NQF-endorsed hospital measures of mortality and readmission17-21,48. Furthermore, we limited our assessment of complications to those within ninety days of the THA or TKA procedure, as the incidence of complications following THA or TKA declines over time and accurate performance measurement and attribution are increasingly difficult at more distant time points following surgery. Finally, we were unable to evaluate the impact of these complications on the overall functional status of patients following THA or TKA because administrative claims do not include measures of functional status.
In conclusion, we found a fourfold variation in the risk-standardized rate of complications following elective THA and TKA procedures across U.S. hospitals, which suggests that opportunities exist to improve the quality of THA and TKA care at the hospital level. The range of performance among hospitals with high proportions of patients with low socioeconomic status or black patients illuminates the ability of certain hospitals that serve vulnerable populations to achieve a low RSCR; however, there is a need to focus quality improvement efforts within those hospitals and communities that are not achieving the same level of performance. Our findings underscore the need to measure and reduce hospital-level variation in outcomes for patients undergoing THA or TKA in order to optimize care for all patients who undergo these potentially life-altering procedures.
Investigation performed at Yale New Haven Health Services Corporation/Center for Outcomes Research & Evaluation, New Haven, Connecticut
2. Kurtz SM Lau E Ong K Zhao K Kelly M Bozic KJ. Future young patient demand for primary and revision joint replacement: national projections from 2010 to 2030. Clin Orthop Relat Res. 2009 Oct;467(10):2606–12. Epub 2009 Apr 10.
4. Cram P Vaughan-Sarrazin MS Wolf B Katz JN Rosenthal GE. A comparison of total hip and knee replacement in specialty and general hospitals. J Bone Joint Surg Am. 2007 Aug;89(8):1675–84.
5. Bozic KJ Chiu V. Quality measurement and public reporting in total joint arthroplasty. J Arthroplasty. 2008 Sep;23(6)(Suppl 1):146–9. Epub 2008 Jun 13.
6. Cram P Cai X Lu X Vaughan-Sarrazin MS Miller BJ. Total knee arthroplasty outcomes in top-ranked and non-top-ranked orthopedic hospitals: an analysis of Medicare administrative data. Mayo Clin Proc. 2012 Apr;87(4):341–8.
7. Cai X Cram P Vaughan-Sarrazin M. Are African American patients more likely to receive a total knee arthroplasty in a low-quality hospital? Clin Orthop Relat Res. 2012 Apr;470(4):1185–93. Epub 2011 Aug 31.
8. Hinman A Bozic KJ. Impact of payer type on resource utilization, outcomes and access to care in total hip arthroplasty. J Arthroplasty. 2008 Sep;23(6)(Suppl 1):9–14.
9. SooHoo NF Farng E Zingmond DS. Disparities in the utilization of high-volume hospitals for total hip replacement. J Natl Med Assoc. 2011 Jan;103(1):31–5.
10. Fry DE. Surgical site infections and the Surgical Care Improvement Project (SCIP): evolution of national quality measures. Surg Infect (Larchmt). 2008 Dec;9(6):579–84.
11. Nguyen N Yegiyants S Kaloostian C Abbas MA Difronzo LA. The Surgical Care Improvement Project (SCIP) initiative to reduce infection in elective colorectal surgery: which performance measures affect outcome? Am Surg. 2008 Oct;74(10):1012–6.
13. Krumholz HM Brindis RG Brush JE Cohen DJ Epstein AJ Furie K Howard G Peterson ED Rathore SS Smith SC Jr Spertus JA Wang Y Normand SL; American Heart Association. Quality of Care and Outcomes Research Interdisciplinary Writing Group; Council on Epidemiology and Prevention; Stroke Council; American College of Cardiology Foundation; endorsed by the American College of Cardiology Foundation. Standards for statistical models used for public reporting of health outcomes: an American Heart Association Scientific Statement from the Quality of Care and Outcomes Research Interdisciplinary Writing Group: cosponsored by the Council on Epidemiology and Prevention and the Stroke Council. Circulation. 2006 Jan 24;113(3):456–62. Epub 2005 Dec 19.
16. Centers for Medicare & Medicaid Services. Medicare program; hospital inpatient prospective payment systems for acute care hospitals and the long-term care hospital prospective payment system and fiscal year 2013 rates; hospitals’ resident caps for graduate medical education payment purposes; quality reporting requirements for specific providers and for ambulatory surgical centers. final rule. Fed Regist. 2012 Aug 31;77(170):53257–750.
17. Bratzler DW Normand SL Wang Y O’Donnell WJ Metersky M Han LF Rapp MT Krumholz HM. An administrative claims model for profiling hospital 30-day mortality rates for pneumonia patients. PLoS ONE. 2011;6(4):e17401. Epub 2011 Apr 12.
18. Krumholz HM Wang Y Mattera JA Wang Y Han LF Ingber MJ Roman S Normand SL. An administrative claims model suitable for profiling hospital performance based on 30-day mortality rates among patients with an acute myocardial infarction. Circulation. 2006 Apr 4;113(13):1683–92. Epub 2006 Mar 20.
19. Keenan PS Normand SL Lin Z Drye EE Bhat KR Ross JS Schuur JD Stauffer BD Bernheim SM Epstein AJ Wang Y Herrin J Chen J Federer JJ Mattera JA Wang Y Krumholz HM. An administrative claims measure suitable for profiling hospital performance on the basis of 30-day all-cause readmission rates among patients with heart failure. Circ Cardiovasc Qual Outcomes. 2008 Sep;1(1):29–37.
20. Krumholz HM Lin Z Drye EE Desai MM Han LF Rapp MT Mattera JA Normand SL. An administrative claims measure suitable for profiling hospital performance based on 30-day all-cause readmission rates among patients with acute myocardial infarction. Circ Cardiovasc Qual Outcomes. 2011 Mar;4(2):243–52.
21. Lindenauer PK Normand SL Drye EE Lin Z Goodrich K Desai MM Bratzler DW O’Donnell WJ Metersky ML Krumholz HM. Development, validation, and results of a measure of 30-day readmission following hospitalization for pneumonia. J Hosp Med. 2011 Mar;6(3):142–50. Epub 2011 Jan 05.
23. Phillips CB Barrett JA Losina E Mahomed NN Lingard EA Guadagnoli E Baron JA Harris WH Poss R Katz JN. Incidence rates of dislocation, pulmonary embolism, and deep infection during the first six months after elective total hip replacement. J Bone Joint Surg Am. 2003 Jan;85(1):20–6.
24. Mahomed NN Barrett J Katz JN Baron JA Wright J Losina E. Epidemiology of total knee replacement in the United States Medicare population. J Bone Joint Surg Am. 2005 Jun;87(6):1222–8.
25. Feinglass J Amir H Taylor P Lurie I Manheim LM Chang RW. How safe is primary knee replacement surgery? Perioperative complication rates in Northern Illinois, 1993-1999. Arthritis Rheum. 2004 Feb 15;51(1):110–6.
26. Meek RM Allan DB McPhillips G Kerr L Howie CR. Epidemiology of dislocation after total hip arthroplasty. Clin Orthop Relat Res. 2006 Jun;447:9–18.
27. Parvizi J Johnson BG Rowland C Ereth MH Lewallen DG. Thirty-day mortality after elective total hip arthroplasty. J Bone Joint Surg Am. 2001 Oct;83(10):1524–8.
28. Seagroatt V Tan HS Goldacre M Bulstrode C Nugent I Gill L. Elective total hip replacement: incidence, emergency readmission rate, and postoperative mortality. BMJ. 1991 Dec 7;303(6815):1431–5.
29. Woolson ST Rahimtoola ZO. Risk factors for dislocation during the first 3 months after primary total hip replacement. J Arthroplasty. 1999 Sep;14(6):662–8.
30. Zhan C Kaczmarek R Loyo-Berrios N Sangl J Bright RA. Incidence and short-term outcomes of primary and revision hip replacement in the United States. J Bone Joint Surg Am. 2007 Mar;89(3):526–33.
31. Memtsoudis SG González Della Valle A Besculides MC Gaber L Sculco TP. In-hospital complications and mortality of unilateral, bilateral, and revision TKA: based on an estimate of 4,159,661 discharges. Clin Orthop Relat Res. 2008 Nov;466(11):2617–27. Epub 2008 Aug 14.
32. SooHoo NF Lieberman JR Ko CY Zingmond DS. Factors predicting complication rates following total knee replacement. J Bone Joint Surg Am. 2006 Mar;88(3):480–5.
33. Jain NB Guller U Pietrobon R Bond TK Higgins LD. Comorbidities increase complication rates in patients having arthroplasty. Clin Orthop Relat Res. 2005 Jun;435:232–8.
34. Solomon DH Chibnik LB Losina E Huang J Fossel AH Husni E Katz JN. Development of a preliminary index that predicts adverse events after total knee replacement. Arthritis Rheum. 2006 May;54(5):1536–42.
35. Pulido L Ghanem E Joshi A Purtill JJ Parvizi J. Periprosthetic joint infection: the incidence, timing, and predisposing factors. Clin Orthop Relat Res. 2008 Jul;466(7):1710–5. Epub 2008 Apr 18.
36. Calderwood MS Ma A Khan YM Olsen MA Bratzler DW Yokoe DS Hooper DC Stevenson K Fraser VJ Platt R Huang SS; CDC Prevention Epicenters Program. Use of Medicare diagnosis and procedure codes to improve detection of surgical site infections following hip arthroplasty, knee arthroplasty, and vascular surgery. Infect Control Hosp Epidemiol. 2012 Jan;33(1):40–9. Epub 2011 Nov 22.
38. American Hospital Association (AHA). AHA annual survey database fiscal year 2009.
39. Cushner F Agnelli G FitzGerald G Warwick D. Complications and functional outcomes after total hip arthroplasty and total knee arthroplasty: results from the Global Orthopaedic Registry (GLORY). Am J Orthop. 2010 Sep;39(9)(Suppl):22–8.
40. Khatod M Inacio M Paxton EW Bini SA Namba RS Burchette RJ Fithian DC. Knee replacement: epidemiology, outcomes, and trends in Southern California: 17,080 replacements from 1995 through 2004. Acta Orthop. 2008 Dec;79(6):812–9.
41. Jaffer AK Barsoum WK Krebs V Hurbanek JG Morra N Brotman DJ. Duration of anesthesia and venous thromboembolism after hip and knee arthroplasty. Mayo Clin Proc. 2005 Jun;80(6):732–8.
42. Khatod M Barber T Paxton E Namba R Fithian D. An analysis of the risk of hip dislocation with a contemporary total joint registry. Clin Orthop Relat Res. 2006 Jun;447:19–23.
43. Ong KL Lau E Suggs J Kurtz SM Manley MT. Risk of subsequent revision after primary and revision total joint arthroplasty. Clin Orthop Relat Res. 2010 Nov;468(11):3070–6. Epub 2010 May 25.
44. Peersman G Laskin R Davis J Peterson M. Infection in total knee replacement: a retrospective review of 6489 total knee replacements. Clin Orthop Relat Res. 2001 Nov;392:15–23.
45. Suarez J Griffin W Springer B Fehring T Mason JB Odum S. Why do revision knee arthroplasties fail? J Arthroplasty. 2008 Sep;23(6)(Suppl 1):99–103. Epub 2008 Jun 05.
46. Springer BD Fehring TK Griffin WL Odum SM Masonis JL. Why revision total hip arthroplasty fails. Clin Orthop Relat Res. 2009 Jan;467(1):166–73. Epub 2008 Oct 31.
47. Sierra RJ Cooney WP 4th Pagnano MW Trousdale RT Rand JA. Reoperations after 3200 revision TKAs: rates, etiology, and lessons learned. Clin Orthop Relat Res. 2004 Aug;425:200–6.
48. Krumholz HM Wang Y Mattera JA Wang Y Han LF Ingber MJ Roman S Normand SL. An administrative claims model suitable for profiling hospital performance based on 30-day mortality rates among patients with heart failure. Circulation. 2006 Apr 4;113(13):1693–701. Epub 2006 Mar 20.
Disclosure: One or more of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of an aspect of this work. In addition, one or more of the authors, or his or her institution, has had a financial relationship, in the thirty-six months prior to submission of this work, with an entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. Also, one or more of the authors has had another relationship, or has engaged in another activity, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.