Background: Improvements in implant design and surgical technique of unicondylar knee arthroplasty have led to reduced revision rates, but patient selection seems to be crucial for success of such arthroplasties. The purpose of the present study was to analyze the 5-year implant survival rate of unicondylar knee replacements in Germany and to identify patient factors associated with an increased risk of revision, including >30 comorbid conditions.
Methods: Using nationwide billing data of the largest German health-care insurance for inpatient hospital treatment, we identified patients who underwent unicondylar knee arthroplasty between 2006 and 2012. Kaplan-Meier survival curves with revision as the end point and log-rank tests were used to evaluate 5-year implant survival. A multivariable Cox regression model was used to determine factors associated with revision. The risk factors of age, sex, diagnosis, comorbidities, type of implant fixation, and hospital volume were analyzed. Hazard ratios (HRs) and 95% confidence intervals (95% CIs) were calculated.
Results: During the study period, a total of 20,946 unicondylar knee arthroplasties were included. The number of unicondylar knee arthroplasties per year increased during the study period from 2,527 in 2006 to 4,036 in 2012. The median patient age was 64 years (interquartile range, 56 to 72 years), and 60.4% of patients were female. During the time evaluated, the 1-year revision rate decreased from 14.3% in 2006 to 8.7% in 2011. The 5-year survival rate was 87.8% (95% CI, 87.3% to 88.3%). Significant risk factors (p < 0.05) for unicondylar knee arthroplasty revision were younger age (the HR was 2.93 [95% CI, 2.48 to 3.46] for patient age of <55 years, 1.86 [95% CI, 1.58 to 2.19] for 55 to 64 years, and 1.52 [95% CI, 1.29 to 1.79] for 65 to 74 years; patient age of >74 years was used as the reference); female sex (HR, 1.18 [95% CI, 1.07 to 1.29]); complicated diabetes (HR, 1.47 [95% CI, 1.03 to 2.12]); depression (HR, 1.29 [95% CI, 1.06 to 1.57]); obesity, defined as a body mass index of ≥30 kg/m2 (HR, 1.13 [95% CI, 1.02 to 1.26]); and low-volume hospitals, denoted as an annual hospital volume of ≤10 cases (HR, 1.60 [95% CI, 1.39 to 1.84]), 11 to 20 cases (HR, 1.47 [95% CI, 1.27 to 1.70]), and 21 to 40 cases (HR, 1.31 [95% CI, 1.14 to 1.51]) (>40 cases was used as the reference).
Conclusions: Apart from known risk factors, this study showed a significant negative influence of obesity, depression, and complicated diabetes on the 5-year unicondylar knee replacement survival rate. Surgical indications and preoperative patient counseling should consider these findings.
Level of Evidence: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.
1Research Institute of the Local Health Care Funds (WIdO), Berlin, Germany
2HELIOS ENDO-Clinic, Berlin, Germany
3University Medical Center Schleswig-Holstein (UKSH), Kiel, Germany
4Federal Association of the Local Health Care Funds (AOK), Berlin, Germany
5German Society of Orthopaedics and Orthopaedic Surgery (DGOOC), Berlin, Germany
6Kreisklinik Jugenheim, Jugenheim, Germany
7HELIOS Kliniken, Berlin, Germany
8Sana Kliniken Sommerfeld, Sommerfeld, Germany
E-mail address for E. Jeschke: email@example.com
E-mail address for T. Gehrke: firstname.lastname@example.org
E-mail address for C. Günster: christian.günster@wido.bv.aok.de
E-mail address for J. Hassenpflug: email@example.com
E-mail address for J. Malzahn: jürgen.firstname.lastname@example.org
E-mail address for F.U. Niethard: email@example.com
E-mail address for P. Schräder: firstname.lastname@example.org
E-mail address for J. Zacher: email@example.com
E-mail address for A. Halder: firstname.lastname@example.org
Since its introduction in the 1970s, the role of unicondylar knee arthroplasty in the management of knee osteoarthritis remains a subject of continuing discussion1. The advantages of medial unicondylar knee arthroplasty are the preservation of the lateral and patellofemoral compartments as well as the cruciate ligaments and near-normal knee kinematics2. Patients experience less pain and recover faster and the hospital stay is shorter3. Although short-term results have been promising1,4, long-term survival is inferior to that of total knee replacement5-7. Improvements in implant design and surgical technique have led to reduced revision rates, but patient selection seems to be crucial for success in unicondylar knee arthroplasty8.
Therefore, studies have been initiated to identify patient factors predicting implant survival. According to registry data, younger patients have a substantially higher risk of revision than older patients7,8. Many studies found an influence of sex on unicondylar knee replacement survival8-11, but others did not12,13. However, little is known regarding comorbidities as risk factors for failure. For instance, a few studies analyzed the influence of obesity, with conflicting results and often small study sample size12,14-19. Therefore, the purpose of the present study was to analyze the 5-year implant survival rates of unicondylar knee replacements in Germany and to identify patient factors associated with an increased risk of revision, including >30 patient comorbidities, based on a large number of unicondylar knee arthroplasties.
Materials and Methods
In Germany, a national arthroplasty register, the German Arthroplasty Register, has been initiated but does not provide intermediate-term or long-term data yet. For this study, anonymized data of the German health-care insurance Allgemeine Ortskrankenkasse (AOK) were used. The AOK provides nationwide health-care insurance for approximately 30% of the German population20 and is the largest provider of statutory health-care insurance in Germany. All patients are allowed to enroll in the AOK regardless of factors such as age, comorbidity, income, or type of employment. The data are derived from billing data for inpatient hospital treatment. They comprise a unique identification number, age, sex, main diagnosis and comorbidities, procedures, involved side, length of stay, patient survival, and insurance status. Diagnoses were coded according to the International Classification of Diseases, Tenth Revision (ICD-10)21. Procedures were documented using the German version of the International Classification of Procedures in Medicine (ICPM)22, the “Operationen- und Prozedurenschlüssel” (OPS). Health-care providers and health-care insurers jointly issue binding guidelines for the coding of diagnoses and procedures in hospital claims23. Hospital claims data in Germany are thoroughly checked against these guidelines and for plausibility by the Medical Review Board of the Statutory Health Insurance and are returned to hospitals for correction if necessary. Corrections are included in the claims data used in this analysis.
Data were used for this study if patients underwent unicondylar knee arthroplasty between 2006 and 2012 based on the date of admission. Unicondylar knee arthroplasties were identified by OPS code 5-822.0X, including medial and lateral unicondylar knee arthroplasties. If a patient underwent bilateral unicondylar knee arthroplasty in both knees during the study period, these unicondylar knee arthroplasties were included as two separate cases.
The end point of the study was a revision surgical procedure, defined as removal or exchange of at least one implant component in the involved knee for any reason within 5 years. Revision of unicondylar knee arthroplasty was identified by OPS code 5-823.1X and was linked to the primary surgical procedure by the unique identification number of the patient and the side.
We analyzed age, sex, primary diagnosis, comorbidities, type of implant fixation, and hospital volume as risk factors for implant survival. Comorbidities were defined using the Elixhauser measure developed in 1998 to predict mortality from administrative data24. The definition includes 31 acute and chronic comorbidities. Comorbidities were identified using the coding algorithm by Quan et al. based on the ICD-10 coding25, for example, uncomplicated diabetes (E10-E14, fourth digit: 0, 1, or 9), complicated diabetes (i.e., with coma, ketoacidosis, vascular disease; E10-E14, fourth digit: 2-8), depression (F20.4, F31.3-F31.5, F32.X, F33.X, F34.1, F41.2, F43.2), and obesity, defined as a body mass index (BMI) of ≥30 kg/m2 (E66.X). Osteoporosis, which is not included in the Elixhauser measure, was also analyzed as a risk factor for implant survival.
The present study is based on data provided by hospitals for health insurance accounting. The recommendations for good practice in secondary data analysis developed by the German Working Group on the Collection and Use of Secondary Data (AGENS)26 were applied in full. Therefore, no formal ethical committee approval was needed.
Descriptive statistics including medians, interquartile ranges (IQRs), and proportions were used to describe the study sample. Kaplan-Meier survival curves with revision as the end point and log-rank tests were used to evaluate 5-year implant survival. Subgroup analysis was performed for patient age (<55 years, 55 to 64 years, 65 to 74 years, and >74 years), sex, type of implant fixation, and comorbidities.
A multivariable Cox regression model was used to determine factors associated with revision. Hazard ratios (HRs) and 95% confidence intervals (95% CIs) were calculated.
We included patient age groups, sex, type of implant fixation (cemented, uncemented, hybrid), primary diagnosis (osteoarthritis compared with other), 32 comorbidities, and hospital volume as independent variables in the regression model. All comorbidities were entered as separate dichotomous variables. Hospital volume was categorized into four groups (≤10, 11 to 20, 21 to 40, >40 unicondylar knee arthroplasties annually). These groups were based on cut-points reported by Badawy et al.27. Because the study population represents approximately 30% of the patients per hospital, the annual hospital volume was calculated by multiplying the included cases by an inflation factor of 100/30. Patients were entered into the volume groups according to the number of procedures performed at their hospital during the year of the surgical procedure. Thus, hospital volume was examined individually for each year. To correct for overly optimistic estimation of the 95% CI, bootstrapping was applied using 500 bootstrap replicates clustered by unique hospital identifier. Data were censored in the survival analyses in case of death and AOK membership termination. Stata version 11.2 (StataCorp) was used to analyze the data. Significance was set at p < 0.05.
During the study period from 2006 to 2012, a total of 20,946 unicondylar knee arthroplasties in 19,719 patients were evaluated. The median patient age was 64 years (IQR, 56 to 72 years). Female patients accounted for 60.4% of the cases. On average, female patients were 2 years older than male patients. Table I provides a detailed overview of patient characteristics.
In this study, the number of unicondylar knee arthroplasties per year increased after 2008: 2,527 in 2006, 2,520 in 2007, 2,535 in 2008, 2,823 in 2009, 3,058 in 2010, 3,447 in 2011, and 4,036 in 2012. At the same time, the median age of patients undergoing unicondylar knee arthroplasty decreased from 66 years in 2006 to 63 years in 2012. The proportion of female patients also decreased from 65.1% in 2006 to 58.3% in 2012.
Kaplan-Meier Survival Curves
Of a total of 20,946 unicondylar knee arthroplasties, 2,194 (10.5%) were revised within 5 years. During the study period, patients in 761 cases (3.6%) died. Overall, the Kaplan-Meier survivorship of unicondylar knee replacements was 87.8% (95% CI, 87.3% to 88.3%) at 5 years (Fig. 1). An analysis of survivorship by year shows that most revisions were performed within the first 2 years, with a mean survivorship of 95.3% (95% CI, 95.0% to 95.6%) at 1 year, 92.0% (95% CI, 91.7% to 92.4%) at 2 years, 90.3% (95% CI, 89.8% to 90.7%) at 3 years, and 89.1% (95% CI, 88.6% to 89.5%) at 4 years. During the time evaluated, the 1-year revision rate decreased according to the year of implantation: 14.3% in 2006, 14.0% in 2007, 13.1% in 2008, 11.9% in 2009, 10.5% in 2010, and 8.7% in 2011.
Survival curves and log-rank tests indicated differences in survival depending on sex (Fig. 2), age (Fig. 3), implant fixation (Fig. 4), obesity (Fig. 5), and depression (Fig. 6). The Kaplan-Meier survivorship at 5 years was 88.8% (95% CI, 88.0% to 89.6%) for male patients, 87.2% (95% CI, 86.5% to 87.8%) for female patients, 79.9% (95% CI, 78.4% to 81.3%) for patients who were <55 years of age, 87.5% (95% CI, 86.5% to 88.3%) for patients who were 55 to 64 years of age, 89.8% (95% CI, 89.0% to 90.6%) for patients who were 65 to 74 years of age, 93.6% (95% CI, 92.6% to 94.4%) for patients who were >74 years of age, 85.5% (95% CI, 84.5% to 86.7%) for patients with obesity (BMI ≥ 30 kg/m2), and 83.8% (95% CI, 80.7% to 86.4%) for patients with depression. There was no difference in survival depending on primary diagnosis or other comorbidities (e.g., rheumatic disease and osteoporosis). Reasons for revisions were aseptic and septic failures.
Risk Factors for Unicondylar Knee Arthroplasty Revision
Table II shows the results of the multivariable Cox regression analysis. Patient-related risk factors for unicondylar knee arthroplasty revision included younger age, female sex, complicated diabetes, depression, and obesity (BMI ≥ 30 kg/m2). For example, patients who were <55 years of age had a 2.9-fold increased risk compared with patients who were >74 years of age. Additionally, we found low hospital volume to be an independent risk factor for revision.
Overall Survival Rate
We evaluated the data of 20,946 unicondylar knee arthroplasties performed between 2006 and 2012 in 19,719 patients with a median age of 64 years (IQR, 56 to 72 years). In our study, the 5-year implant survival rate was 87.8%.
The reported survival rate is lower than in most registry data, which often show 5-year survival rates above 90%, such as in the National Joint Registry for England, Wales and Northern Ireland9 (93.1%), the Swedish Knee Arthroplasty Register28 (92.5%), and the Australian Orthopaedic Association National Joint Replacement Registry (91.5%)8. It is similar to the data published by the Norwegian Arthroplasty Register (89.5%), the Finnish National Arthroplasty Register (89.4%), and the Danish Knee Arthroplasty Register (87.5%), although the absolute numbers of unicondylar knee arthroplasties reported by these registers are substantially lower at 3,297 for Norway, 4,713 for Finland, and 2,481 for Denmark29,30. Younger age is not an explanation for lower unicondylar knee replacement survival in our study, because patient age is comparable with that in register studies with higher survival rates9,28. However, there is a relatively high proportion of female patients in our study, which could contribute to the reported lower survival rate31. The mortality in our population was lower than the expected age and sex-specific mortality in Germany32.
During the time evaluated, the number of unicondylar knee replacements implanted in Germany nearly doubled, but the 1-year revision risk decreased by 40%. The decreasing revision rate during the study period could be due to various factors such as improved prosthetic design but also to increased surgeon experience.
Risk Factors for Unicondylar Knee Arthroplasty Revision
Most register data confirm our finding that younger age is associated with a lower survival rate. The National Joint Registry for England, Wales and Northern Ireland shows a 5-year revision rate of 3.1% for male patients and 4.9% for female patients who were >75 years of age, compared with 10.1% for male patients and 10.9% for female patients who were <55 years of age9,10. The Australian Orthopaedic Association National Joint Replacement Registry reports a 5-year revision rate of 5.5% in patients who were >75 years of age, compared with 13.2% in patients <55 years of age8. According to the Swedish Knee Arthroplasty Register, patients under the age of 65 years have twice the risk of revision compared with patients older than 75 years of age28. In comparison, we found that patients who were <55 years of age have a 2.9-fold increased risk compared with patients who were >74 years of age.
In some register data, female sex is associated with a lower survival rate. In the National Joint Registry for England, Wales and Northern Ireland, female patients of all ages have a higher 5-year revision rate than male patients, especially when they are >75 years of age9,10. In the Australian Orthopaedic Association National Joint Replacement Registry, male patients have a 5-year revision rate of 7.7% compared with a rate of 9.0% for female patients, equivalent to a 16.9% increase in risk8. In our study, the 5-year revision risk for female patients was 18% higher than for male patients when taking into account confounding factors such as age and comorbid obesity and depression. However, data from the Swedish Knee Arthroplasty Register, from the Finnish National Arthroplasty Register, and for the Oxford medial unicompartmental knee replacement studies showed no significant differences in revision rate between the sexes12,28,29,33,34.
In our study, obesity (BMI ≥ 30 kg/m2) significantly increased the risk of revision by 13%. To our knowledge, few studies have analyzed the influence of obesity on unicondylar knee arthroplasty outcome. In a large study of 15,700 patients, Kandil et al. reported recently that obesity was associated with higher complication rates and a short-term revision rate almost twice as high as that in non-obese patients19. Reports of smaller case series have contradicting results. Kuipers et al.12, Berend et al.14, Pandit et al.35, and Sébilo et al.36 did not find a negative influence of body weight on the 10-year survival of the Oxford UKA. Bonutti et al. found a decreased survival rate associated with obesity18. For total knee arthroplasty, Bozic et al. reported that obesity is a significant risk factor for periprosthetic joint infection and a revision surgical procedure within 12 months37,38.
Liddle et al. found that patients who underwent unicondylar knee arthroplasty and had extreme anxiety or depression had significantly worse subjective outcome scores than patients who did not have anxiety or depression10. In our study, the percentage of women was relatively high, and there was a higher prevalence of depression in women than in men. Therefore, the rate of comorbid depression is relatively high at 3.8%. However, our analysis found depression to be an independent risk factor for unicondylar knee arthroplasty revision. This finding corresponds with that of total knee arthroplasty studies. Bozic et al. found that depression was an independent risk factor for periprosthetic joint infection and for revision surgical procedures within 12 months37,38. Buller et al. reported that depression was associated with increased odds of adverse events39. In a meta-analysis, Lewis et al. described depression as a strong predictor of persistent pain40, and Singh and Lewallen found that depression is associated with suboptimal improvement in knee function after total knee arthroplasty41, both of which could ultimately lead to a revision surgical procedure.
In our study population, 14.3% of the patients were diagnosed with uncomplicated diabetes and 1.2% were diagnosed with complicated diabetes. Diabetes alone was no risk factor, but complicated diabetes was significantly associated with a revision surgical procedure (HR, 1.47 [95% CI, 1.03 to 2.12]). Again, this finding corresponds to total knee arthroplasty studies. Bozic et al. identified diabetes as a risk factor for periprosthetic joint infection37. Singh and Lewallen found that diabetes was associated with poorer functional outcome42.
In our study, 729 cementless unicondylar knee arthroplasties were included among the total of 20,946 unicondylar knee arthroplasties. Uncemented fixation of unicondylar knee arthroplasty reduced the 5-year survival rate significantly by about 2%. Most register data confirm this finding for total knee arthroplasty8,9,28, but not for unicondylar knee arthroplasty, probably because the number of uncemented unicondylar knee replacements is below the number required for significance. However, after adjusting for other patient characteristics, the type of fixation is not an independent risk factor for unicondylar knee arthroplasty revision in our study. Some authors of smaller case series also reported no difference in survival for cementless unicondylar knee replacements43,44. Kendrick et al. did not find increased radiographic evidence of loosening in cementless Oxford UKAs at 2 years postoperatively45.
Recent studies found that the risk for unicondylar knee arthroplasty revision is higher in low-volume hospitals27,46. We also found consistently higher revision rates in hospitals with lower volumes. For example, the revision risk in hospitals with up to 10 cases annually was 60% higher than in hospitals with >40 cases.
Although our study was based on data of the largest health-care insurance in Germany, there may have been variations in terms of age, sex, social status, and morbidity among patients insured by different German health-care providers31. Comparing AOK cases with those of all German patients with knee replacement in 2012, there were slight differences to our study population with regard to female sex (67.2% for AOK compared with 65.2% for Germany) and age of ≥70 years (57.0% for AOK compared with 56.1% for Germany)47. Although a number of relevant patient factors were analyzed in this study, other factors such as laterality of unicondylar knee arthroplasty, previous surgical procedures, range of motion, lower-limb axis, concomitant patellofemoral arthritis, and integrity of cruciate ligaments are not part of the insurance data. Although obesity (BMI ≥ 30 kg/m2) was analyzed in our study, detailed BMI information was not available for the whole study period. Finally, the ICD-10 coding system does not include information about the detailed reasons for revision surgical procedures.
In conclusion, in this nationwide study using German routine insurance data of 20,946 unicondylar knee arthroplasties performed between 2006 and 2012, we found a 5-year survival rate of 87.8%. During this period, the number of unicondylar knee arthroplasties nearly doubled, but the 1-year revision rate decreased. As in most register studies, we found younger age and female sex to be associated with a lower survival rate of unicondylar knee replacements. In our study, obesity (BMI ≥ 30 kg/m2) significantly increased the risk of revision. Furthermore, depression and complicated diabetes had a significant negative influence on the 5-year survival rate of unicondylar knee replacements. These findings should be taken into account for determining surgical indications and for preoperative patient counseling.
Investigation performed at the Research Institute of the Local Health Care Funds (WIdO), Berlin, Germany
Disclosure: There was no external funding for this study. On the Disclosure of Potential Conflicts of Interest forms, which are provided with the online version of the article, one or more of the authors checked “yes” to indicate that the author had a relevant financial relationship in the biomedical arena outside the submitted work.
1. Marmor L. Results of single compartment arthroplasty with acrylic cement fixation. A minimum follow-up of two years. Clin Orthop Relat Res. 1977 ;122:181–8.
2. Heyse TJ, El-Zayat BF, De Corte R, Chevalier Y, Scheys L, Innocenti B, Fuchs-Winkelmann S, Labey L. UKA closely preserves natural knee kinematics in vitro. Knee Surg Sports Traumatol Arthrosc. 2014 ;22(8):1902–10. Epub 2013 Nov 9.
3. Macaulay W, Yoon RS. Fixed-bearing, medial unicondylar knee arthroplasty rapidly improves function and decreases pain: a prospective, single-surgeon outcomes study. J Knee Surg. 2008 ;21(4):279–84.
4. Jain RK, Neville LT, Ezzet KA, Sterling RS, Horwood RL, Colwell CW Jr. Two year follow-up of the preservation unicompartmental knee implant. HSS J. 2011;7(2):125–9.
5. O’Donnell T, Neil MJ. The Repicci II® unicondylar knee arthroplasty: 9-year survivorship and function. Clin Orthop Relat Res. 2010 ;468(11):3094–102. Epub 2010 Aug 13.
6. Tabor OB Jr, Tabor OB. Unicompartmental arthroplasty: a long-term follow-up study. J Arthroplasty. 1998 ;13(4):373–9.
7. Koskinen E, Paavolainen P, Eskelinen A, Pulkkinen P, Remes V. Unicondylar knee replacement for primary osteoarthritis: a prospective follow-up study of 1,819 patients from the Finnish Arthroplasty Register. Acta Orthop. 2007 ;78(1):128–35.
10. Liddle AD, Judge A, Pandit H, Murray DW. Determinants of revision and functional outcome following unicompartmental knee replacement. Osteoarthritis Cartilage. 2014 ;22(9):1241–50. Epub 2014 Jul 17.
11. Thompson SA, Liabaud B, Nellans KW, Geller JA. Factors associated with poor outcomes following unicompartmental knee arthroplasty: redefining the “classic” indications for surgery. J Arthroplasty. 2013 ;28(9):1561–4. Epub 2013 Mar 21.
12. Kuipers BM, Kollen BJ, Bots PC, Burger BJ, van Raay JJ, Tulp NJ, Verheyen CC. Factors associated with reduced early survival in the Oxford Phase III medial unicompartment knee replacement. Knee. 2010 ;17(1):48–52. Epub 2009 Aug 28.
13. Seyler TM, Mont MA, Lai LP, Xie J, Marker DR, Zywiel MG, Bonutti PM. Mid-term results and factors affecting outcome of a metal-backed unicompartmental knee design: a case series. J Orthop Surg Res. 2009;4:39. Epub 2009 Oct 26.
14. Berend KR, Lombardi AV Jr, Adams JB. Obesity, young age, patellofemoral disease, and anterior knee pain: identifying the unicondylar arthroplasty patient in the United States. Orthopedics. 2007 ;30(5)(Suppl):19–23.
15. Cepni SK, Arslan A, Polat H, Yalçin A, Parmaksizoğlu AS. Mid-term results of Oxford Phase 3 unicompartmental knee arthroplasty in obese patients. Acta Orthop Traumatol Turc. 2014;48(2):122–6.
16. Naal FD, Neuerburg C, Salzmann GM, Kriner M, von Knoch F, Preiss S, Drobny T, Munzinger U. Association of body mass index and clinical outcome 2 years after unicompartmental knee arthroplasty. Arch Orthop Trauma Surg. 2009 ;129(4):463–8. Epub 2008 Apr 15.
17. Tabor OB Jr, Tabor OB, Bernard M, Wan JY. Unicompartmental knee arthroplasty: long-term success in middle-age and obese patients. J Surg Orthop Adv. 2005 ;14(2):59–63.
18. Bonutti PM, Goddard MS, Zywiel MG, Khanuja HS, Johnson AJ, Mont MA. Outcomes of unicompartmental knee arthroplasty stratified by body mass index. J Arthroplasty. 2011 ;26(8):1149–53. Epub 2011 Jan 21.
19. Kandil A, Werner BC, Gwathmey WF, Browne JA. Obesity, morbid obesity and their related medical comorbidities are associated with increased complications and revision rates after unicompartmental knee arthroplasty. J Arthroplasty. 2015 ;30(3):456–60. Epub 2014 Oct 25.
24. Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998 ;36(1):8–27.
25. Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, Saunders LD, Beck CA, Feasby TE, Ghali WA. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005 ;43(11):1130–9.
26. Arbeitsgruppe Erhebung und Nutzung von Sekundärdaten (AGENS) der Deutschen Gesellschaft für Sozialmedizin und Prävention; Arbeitsgruppe Epidemiologische Methoden der Deutschen Gesellschaft für Epidemiologie; Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie; Deutschen Gesellschaft für Sozialmedizin und Prävention. [Good practice of secondary data analysis, first revision]. Gesundheitswesen. 2008 ;70(1):54–60. German.
27. Badawy M, Espehaug B, Indrekvam K, Havelin LI, Furnes O. Higher revision risk for unicompartmental knee arthroplasty in low-volume hospitals. Acta Orthop. 2014 ;85(4):342–7. Epub 2014 May 21.
29. Niinimäki T, Eskelinen A, Mäkelä K, Ohtonen P, Puhto AP, Remes V. Unicompartmental knee arthroplasty survivorship is lower than TKA survivorship: a 27-year Finnish registry study. Clin Orthop Relat Res. 2014 ;472(5):1496–501. Epub 2013 Nov 19.
30. Robertsson O, Bizjajeva S, Fenstad AM, Furnes O, Lidgren L, Mehnert F, Odgaard A, Pedersen AB, Havelin LI. Knee arthroplasty in Denmark, Norway and Sweden. A pilot study from the Nordic Arthroplasty Register Association. Acta Orthop. 2010 ;81(1):82–9.
31. Hoffmann F, Icks A. [Structural differences between health insurance funds and their impact on health services research: results from the Bertelsmann Health-Care Monitor]. Gesundheitswesen. 2012 ;74(5):291–7. Epub 2011 Jul 13. German.
33. Matharu G, Robb C, Baloch K, Pynsent P. The Oxford medial unicompartmental knee replacement: survival and the affect of age and gender. Knee. 2012 ;19(6):913–7. Epub 2012 Apr 4.
34. Kristensen PW, Holm HA, Varnum C. Up to 10-year follow-up of the Oxford medial partial knee arthroplasty—695 cases from a single institution. J Arthroplasty. 2013 ;28(9)(Suppl):195–8. Epub 2013 Jun 21.
35. Pandit H, Jenkins C, Gill HS, Smith G, Price AJ, Dodd CA, Murray DW. Unnecessary contraindications for mobile-bearing unicompartmental knee replacement. J Bone Joint Surg Br. 2011 ;93(5):622–8.
36. Sébilo A, Casin C, Lebel B, Rouvillain JL, Chapuis S, Bonnevialle P; members of the Société d’Orthopédie et de Traumatologie de l’Ouest (SOO). Clinical and technical factors influencing outcomes of unicompartmental knee arthroplasty: retrospective multicentre study of 944 knees. Orthop Traumatol Surg Res. 2013 ;99(4)(Suppl):S227–34. Epub 2013 Apr 25.
37. Bozic KJ, Lau E, Kurtz S, Ong K, Berry DJ. Patient-related risk factors for postoperative mortality and periprosthetic joint infection in Medicare patients undergoing TKA. Clin Orthop Relat Res. 2012 ;470(1):130–7.
38. Bozic KJ, Lau E, Ong K, Chan V, Kurtz S, Vail TP, Rubash HE, Berry DJ. Risk factors for early revision after primary total hip arthroplasty in Medicare patients. Clin Orthop Relat Res. 2014 ;472(2):449–54.
39. Buller LT, Best MJ, Klika AK, Barsoum WK. The influence of psychiatric comorbidity on perioperative outcomes following primary total hip and knee arthroplasty; a 17-year analysis of the National Hospital Discharge Survey database. J Arthroplasty. 2015 ;30(2):165–70. Epub 2014 Sep 18.
40. Lewis GN, Rice DA, McNair PJ, Kluger M. Predictors of persistent pain after total knee arthroplasty: a systematic review and meta-analysis. Br J Anaesth. 2015 ;114(4):551–61. Epub 2014 Dec 26.
41. Singh JA, Lewallen DG. Depression in primary TKA and higher medical comorbidities in revision TKA are associated with suboptimal subjective improvement in knee function. BMC Musculoskelet Disord. 2014;15:127. Epub 2014 Apr 11.
42. Singh JA, Lewallen DG. Diabetes: a risk factor for poor functional outcome after total knee arthroplasty. PLoS One. 2013 ;8(11):e78991.
43. Epinette JA. Long lasting outcome of hydroxyapatite-coated implants in primary knee arthroplasty: a continuous series of two hundred and seventy total knee arthroplasties at fifteen to twenty two years of clinical follow-up. Int Orthop. 2014 ;38(2):305–11. Epub 2014 Jan 3.
44. Schlueter-Brust K, Kugland K, Stein G, Henckel J, Christ H, Eysel P, Bontemps G. Ten year survivorship after cemented and uncemented medial Uniglide® unicompartmental knee arthroplasties. Knee. 2014 ;21(5):964–70. Epub 2014 Apr 15.
45. Kendrick BJ, Kaptein BL, Valstar ER, Gill HS, Jackson WF, Dodd CA, Price AJ, Murray DW. Cemented versus cementless Oxford unicompartmental knee arthroplasty using radiostereometric analysis: a randomised controlled trial. Bone Joint J. 2015 ;97-B(2):185–91.
Copyright 2016 by The Journal of Bone and Joint Surgery, Incorporated
46. Baker P, Jameson S, Critchley R, Reed M, Gregg P, Deehan D. Center and surgeon volume influence the revision rate following unicondylar knee replacement: an analysis of 23,400 medial cemented unicondylar knee replacements. J Bone Joint Surg Am. 2013 ;95(8):702–9.