Abstract
Background:
Osteoarthritis is associated with a strong biomechanical component. Persistent pain in the index knee after total knee arthroplasty could lead to pain in the contralateral knee. The purpose of the present study was to examine whether a change in the natural history of pain in the contralateral knee was related to postoperative pain in the index knee.
Methods:
Seven hundred and seventy-two patients undergoing primary unilateral total knee arthroplasty with use of the Kinemax prosthesis for the treatment of osteoarthritis comprised the cohort (Kinemax Outcomes Study cohort). Patients were assessed preoperatively and were followed for twenty-four months after surgery with use of the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC). We collected separate WOMAC pain scores for the index knee and the contralateral knee. Our primary outcome measure was the WOMAC pain score (rescaled to 100, with 100 being the best score) for the contralateral knee at three, twelve, and twenty-four months. We examined whether within-subject changes in the WOMAC pain score for the contralateral knee were predicted by the WOMAC pain score for the index knee at three months with use of linear regression and multilevel models after adjustment for sex, age, country, body mass index, income, and mental well-being.
Results:
Improvement in terms of pain was observed in both the index and contralateral knees between baseline and three months. Subsequently, there was a modest deterioration of 3.5 units per year (standard deviation, 9.8 units per year) in the contralateral knee (p < 0.001), which was not predicted by pain in the index knee shortly after surgery (p > 0.6).
Conclusions:
Pain in the index knee at three months after total knee arthroplasty did not appear to predict a symptomatic increase in pain in the contralateral knee over two years of follow-up in our cohort. The contralateral knee did not require any additional clinical surveillance over and above the patients’ reports on their symptoms.
Level of Evidence:
Prognostic Level IV. See Instructions for Authors for a complete description of levels of evidence.
The success of total knee arthroplasty as a treatment for osteoarthritis of the knee is well established1-4. Relatively little is known about the impact of primary total knee arthroplasty on the contralateral, nonreplaced knee, which will often also have a lesser degree of degenerative changes. Previous studies5,6 have shown that radiographically determined and symptomatic severity in the contralateral knee at the time of surgery in the index knee are strongly associated with subsequent progression to surgery in the contralateral knee. Ritter et al.7 reported the probability of normal contralateral knees and contralateral knees with osteoarthritis (at the time of index knee surgery) undergoing subsequent arthroplasty to be 5% and 37%, respectively. McMahon and Block6 described an overall ten-year risk of contralateral total knee arthroplasty of 37%, with the baseline Kellgren-Lawrence grade8 of the contralateral knee at the time of index surgery being a strong predictor of eventual contralateral total knee arthroplasty.
Knee osteoarthritis is a multifactorial disease with a strong biomechanical component9-11. It is known that subjects with medial compartment knee osteoarthritis walk with greater knee adduction moments than normal controls, leading to increased compressive load in the medial compartment. However, it is unclear whether these loading abnormalities cause or occur as a result of osteoarthritis. We hypothesized that persistent pain following an index knee arthroplasty could lead to the development or worsening of pain in the contralateral knee through gait alteration. To our knowledge, no previous studies have examined whether postoperative outcome predicts subsequent increase in symptomatic pain in the contralateral knee and whether this could be a useful clinical indicator of patients with a higher risk for subsequent contralateral surgery.
The purpose of the present study was to assess the early natural history of osteoarthritis in the contralateral knee following primary total knee arthroplasty. We aimed to define whether postoperative pain in the index knee led to the development of pain in the contralateral knee.
Design
Patient data were obtained from the Kinemax Outcomes Study12-17, a prospective observational study of primary total knee arthroplasty for the treatment of osteoarthritis in thirteen centers: six in the United Kingdom (U.K.), four in the United States (U.S.), two in Australia, and one in Canada. The appropriate institutional review board or ethics committee approved the study at each of the participating centers. Independent research assistants recruited patients from September 1997 to December 1998 in the U.K., U.S., and Australia and until December 1999 in Canada.
Patients
Patients undergoing primary total knee arthroplasty with the Kinemax prosthesis (Stryker Howmedica, Mahwah, New Jersey) for the treatment of primary osteoarthritis were recruited prospectively. The inclusion criterion was primary arthroplasty surgery with use of the Kinemax prosthesis for the treatment of knee osteoarthritis. A diagnosis of knee osteoarthritis was made by the operating surgeon in the usual fashion after clinical and radiographic examination. Patients were excluded if they had a history of knee joint infection or previous implant surgery in the index knee or if they were unable to complete the questionnaire because of cognitive or language difficulties.
Data Collection
Independent research assistants at each center obtained informed, written consent and collected data with use of a standardized protocol. Preoperative data were collected within six weeks prior to the total knee arthroplasty, and follow-up data were collected at three, twelve, and twenty-four months postoperatively for both the operatively treated knee (the index knee) and the nonreplaced knee (the contralateral knee). One author (E.A.L.) trained all of the research assistants to standardize the collection of data, which were entered into a single database at the coordinating center (Brigham and Women’s Hospital, Boston, Massachusetts). The data included information on comorbidities (collected with use of a validated self-administered comorbidity questionnaire that included questions related to the history of comorbid conditions18), demographic and socioeconomic information (education, income, employment status, living arrangement), height, and weight. At each evaluation, the patients completed two health questionnaires: the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC)19 (a disease-specific measure of pain, stiffness, and function for patients with hip and knee osteoarthritis) and the Short Form-36 (SF-36) (a generic health-status measure used for quality of life20-23). These scales were selected for the purposes of analysis because previous work by the group has shown them to be responsive measures of the outcomes of total knee arthroplasty and they remove observer bias13. WOMAC pain questions were asked independently for the right and left knees, and a separate pain score was calculated for each knee. WOMAC pain scores were transformed to a scale of 0 to 100 (with 100 being the best) for ease of analysis. WOMAC function scores examined an individual’s ability to carry out a range of activities of daily living and were similarly transformed. The SF-36 score (range, 0 to 100, with 100 being the best) was calculated with use of the standard method, which includes eight subscale scores (four related to physical health and four related to mental health).
Statistical Analysis
Our main outcome of interest was the WOMAC pain score for the contralateral knee at twenty-four months. If a subject had an operation on the contralateral knee over this period, he or she was excluded from the main analysis. Our main exposure variable was the WOMAC pain score in the index knee at three months postoperatively. A poor score indicates that the patient reported persistent pain after surgery. WOMAC pain scores were analyzed either as a continuous variable or as an ordinal variable. To examine predictors of severity, we categorized anyone with a baseline score of ≤30 (bottom tertile) in the index knee as having a poor score. Univariate regression analyses were performed to analyze baseline characteristics between countries, age and sex, and predictors of poor baseline outcome in the study population. We examined the natural history of the contralateral WOMAC pain scores by using box-and-whisker plots with all four time points. Each box spans the interquartile range, with the median indicated as the middle horizontal line. The whiskers stretch 1.5 times the interquartile range above and below the box. Any outlier observations lying beyond the whiskers are indicated by a dot24.
Because we observed a large improvement in the immediate postoperative period for the contralateral WOMAC pain score, which reflects the benefit of operating on the index knee, we modeled the change in WOMAC pain scores with use of only the data between three and twenty-four months.
We used two analytical approaches. First, we generated within-subject gradients (with negative values indicating worsening and positive values indicating improvement) for the repeat measures and used these gradients as our outcome measure in analyses based on the standard linear regression. We checked the assumption of a linear decline by also fitting models with a quadratic term. In addition, we repeated our analysis with use of a random effects multilevel model with the repeat measures (level 1) being modeled as a function of fixed characteristics, for example, age group and sex (level 2). We generally report results with use of the first method for the sake of simplicity, and the results with both methods were qualitatively similar. We ran two multivariable models: (1) a simple model adjusting for age (continuous), sex, and units (dummy variable) and (2) a fully adjusted model that included the above variables but also adjusted for body mass index (BMI), income group (ordinal variable), and the SF-36 mental health domain. These last three variables were conceptualized as potential confounders as they could be risk factors for disease progression as well as operative success. The SF-36 mental health domain is particularly relevant as subjects who have minor degrees of psychological morbidity may be more likely to report negative outcomes25. We also undertook a sensitivity analysis with use of the twelve-month data as the latest follow-up so we could examine the impact of excluding subjects who had an operation on the contralateral knee between twelve to twenty-four months. This subgroup includes patients with either worse baseline scores or more rapid progression, and we were concerned that by censoring them we may have introduced bias in our analyses.
Given the distribution of the gradients, we also used bootstrap methods (5000 samples) to generate the 95% confidence interval (CI) without the need to make assumptions about the sampling distributions of the regression coefficients26. As the results were very similar, we presented conventional 95% CIs in the tables.
Source of Funding
This work was funded by Stryker Howmedica, Mahwah, New Jersey. The funders had no role in the design, analysis, or interpretation of the data for this publication.
Preoperative Characteristics
One thousand, one hundred and ninety-eight patients (1325 knees) were recruited from September 1997 to December 1999 (Fig. 1). Four hundred and twenty-six patients were excluded as they had either a simultaneous bilateral procedure (fifty-six patients), had already had previous surgery on the contralateral knee (199 patients), had missing data (100 patients), or had an operation on the contralateral knee within twelve months after the start of the study (seventy-one patients). The study cohort consisted of 772 patients (772 knees). Fifty-one patients had a contralateral total knee arthroplasty more than twelve months after the arthroplasty on the index knee. These subjects were excluded from the twenty-four-month analysis but were included in a twelve-month sensitivity analysis. Twenty-six patients died during the follow-up period. Data on the contralateral knee were available for 608 patients at twelve months (response rate, 79%). An additional 111 patients were lost to follow-up for various reasons from twelve to twenty-four months, but an additional eleven patients who had not been included at twelve months were included at twenty-four months, resulting in 508 of 772 patients being evaluated at twenty-four months (response rate, 66%) (Fig. 1). Other reasons for the loss of patients are described in the flowchart.
Table I shows the demographic and preoperative characteristics of the patients. The greatest numbers of the patients were white, had one to two comorbidities, and came from the U.K. centers. The mean age (and standard deviation) was 69 ± 9.5 years. There were more women (59.7%) than men (40.3%) in the study, and women had a higher BMI (p < 0.0001). Table II presents the predictors of worse index WOMAC pain scores at baseline. Women had worse preoperative index WOMAC pain scores compared with men (mean, 44.6 compared with 35.6; p < 0.0001). Preoperative index WOMAC pain scores were less severe for patients at the Australian center than at the other centers (odds ratio, 0.4; 95% CI, 0.3 to 0.7; p < 0.0001 compared with the U.K. center). Other predictors of poor baseline index WOMAC pain score were a younger age, a higher BMI, and a worse contralateral WOMAC pain score at baseline (p < 0.0001 for all).
Natural History of the Contralateral Knee
There was evidence of improvement in WOMAC pain scores for both the index knee (mean improvement, 33.7 ± 24.1 units; p < 0.001) and the contralateral knee (mean improvement, 7.2 ± 17.8 units; p < 0.001) from baseline to three months postoperatively, followed by a modest deterioration in the contralateral knee (mean deterioration, −3.5 ± 9.8 units per year; p < 0.001) until the twenty-four-month assessment. There was no trend toward deterioration or improvement in pain scores between three and twenty-four months (Figs. 2-A and 2-B).
There was no evidence that a quadratic term had any better fit than a linear term in describing the change in WOMAC pain scores over time (p = 0.57). The predictors of within-subject gradients in the contralateral WOMAC pain score are shown in Table III. In the simple model, we found no evidence that the postoperative WOMAC pain score for the index knee at three months (see Appendix) or older age predicted the rate of decline. However, women did show a greater rate of decline compared with men (p = 0.04). Further adjustment for BMI, income, and SF-36 mental health score as potential confounders resulted in attenuation of this sex difference so that it was now consistent with chance. These results were seen in our multilevel analysis; that is, there was no evidence of an effect of the WOMAC pain score for the index knee at three months and modest evidence of an effect of sex on the change in contralateral knee pain between three and twenty-four months postoperatively.
We repeated the analysis with change between three to twelve months as the outcome, so that subjects who had an operation in the twelve to twenty-four-month period were included. The covariates were index knee pain at three months, age, sex, and study center. This analysis included data from 550 participants. We found no convincing evidence that the WOMAC pain score for the index knee predicted the rate of decline during the three to twelve-month period (p = 0.33). The sex-related differences were even larger, with females experiencing an additional 2.86 units of worsening of pain per year in comparison with males (95% CI, −6.83 to 1.10), but, because of greater variability in the outcome measure, these results were consistent with chance (p = 0.17).
We found no evidence (p = 0.63) that the association between the index knee pain at three months and the change in contralateral knee pain between three and twenty-four months was modified by the level of pain in the contralateral knee at baseline. We tested this by adding a term to the analysis model that would capture any interaction between the effects of index knee pain at three months and contralateral knee pain at baseline. We further examined whether the WOMAC function score was a predictor of deterioration. While the baseline WOMAC function score was nonpredictive (p = 0.74), a better postoperative WOMAC function score at three months was associated with greater deterioration in the WOMAC pain score for the contralateral knee: for every one unit more favorable the score on the postoperative WOMAC function measure at three months, the average deterioration in the contralateral knee was 0.09 units greater on the WOMAC pain score (95% CI, −0.17 to 0.00; p = 0.06) and this association was strengthened after further adjustment for baseline function score and other covariates, providing weak evidence against the null hypothesis (deterioration, 0.11 units greater [95% CI, −0.20 to −0.01]; p = 0.03).
To our knowledge, this is the first study that has assessed the impact of unilateral total knee arthroplasty as a predictor of worsening contralateral knee pain and represents the first attempt to define the early natural history of contralateral knee pain after index total knee arthroplasty. We failed to find evidence that poor postoperative outcome in terms of pain in the index knee was associated with greater worsening of pain in the contralateral knee over a two-year period. This observation is clinically important as persistent pain after total knee arthroplasty is distressing for patients27. However, seventy-one (6%) of 1198 patients had a total knee arthroplasty on the contralateral knee within twelve months and were not included in the analysis. We cannot say whether this procedure was a planned staged intervention because the contralateral knee was already severely affected or whether it reflected postoperative deterioration. Furthermore, 199 (17%) of 1198 patients were excluded from our analysis because they had had a previous total knee arthroplasty. This finding highlights that it is common to have bilateral knee disease in general. However, we did find that women reported increasing contralateral knee pain at a faster rate as compared with men postoperatively, which has not been previously described but is consistent with other natural history data (as discussed below). Adjustment for BMI, income, and mental well-being explained this sex difference to some degree. Interestingly, better functional ability at three months postoperatively was associated with a greater increase in contralateral knee pain, possibly reflecting greater use and hence weight-bearing on the contralateral joint secondary to the postoperative improvement in terms of pain and function for the index knee.
Previous epidemiological studies assessing the influence of sex on the natural history of knee osteoarthritis have consistently shown an increased prevalence of knee osteoarthritis in women as compared with men28-31. The Framingham Osteoarthritis Study31 showed that women were significantly more likely to have radiographic evidence of osteoarthritis (age-adjusted relative risk [RR] for females compared with males = 1.79) and symptomatic osteoarthritis at baseline (RR = 1.96), to have a progressive disease trajectory (RR = 1.43), and to be at greater risk for developing contralateral knee osteoarthritis than men (RR = 1.54) over a mean duration of follow-up period of 8.1 years. This apparent sex-related disparity may be partially explained by biomechanical gait differences, sex-related differences in joint morphology32, the influence of obesity33,34, and increased levels of pain-reporting behavior in women35, although this complex relationship between sex and osteoarthritis is not well understood.
Our study design and findings have several strengths. First, the results are based on a reasonably large sample size as most previous studies generally have included <100 patients. Second, our cohort was a multicenter cohort, enhancing its generalizability to patients undergoing total knee arthroplasty. However, the procedures were performed by select surgeons from each of these select sites with use of only one implant, so this cohort is hardly population-based. Third, four repeat measures of pain were obtained over a two-year period with use of internationally accepted patient-reported outcome measures. Fourth, we adjusted for a variety of potential confounders and used sophisticated statistical methods that allow for the longitudinal nature of the data.
It is also important to consider potential weaknesses. First, our follow-up period was only two years. We cannot exclude the possibility that any decline in the contralateral knee in subjects with a poor postoperative outcome in terms of pain only becomes important after a longer time period. Second, we assumed that the three-month index WOMAC pain score was a valid measure of one aspect of patient-related outcomes. It is likely that some patients report a poor outcome because of psychosocial factors such as depression rather than any technical problems related to the arthroplasty. We were able to adjust for mental well-being with use of a general measure derived from the SF-36 mental health index, and this adjustment did not alter the results, so we suspect that this is not a major limitation. It is possible that a few patients were using an assistive device at three months, in which case their WOMAC score may have been better than expected. Third, some patients were lost to follow-up, as is inevitable in any longitudinal follow-up study, and some subjects had to be excluded because they had an arthroplasty on the contralateral knee over the two-year period. This may have biased our results in underestimating the true decline over the two-year period, although our sensitivity analysis demonstrated similar null effects for the one-year follow-up. Fourth, we did not have imaging data so we cannot say whether changes in pain scores were also reflected in structural changes in the joint.
The clinical implications of our results are twofold. Patients presenting for primary total knee arthroplasty may be concerned about the implications of surgery for the contralateral knee. In this cohort, seventy-one (6%) of the originally identified 1198 patients had a total knee arthroplasty on the contralateral knee within twelve months. However, for those who did not, it is encouraging to find that there was a major improvement in both the index and contralateral knees in the immediate postoperative period. Patient-reported pain in the index knee after total knee arthroplasty does not appear to predict the rate of decline of the contralateral knee. Second, our findings suggest that clinicians do not need to routinely monitor the contralateral knee more aggressively for patients with persistent postoperative pain in the index knee.
A scatterplot showing the change in contralateral knee pain against index knee pain (as assessed with the WOMAC score) at three months and a table showing the mean change per year in the WOMAC score for the contralateral knee according to WOMAC score category for the index knee at three months are available with the online version of this article as a data supplement at jbjs.org.
Note: The authors acknowledge the work of the Kinemax Outcomes Group in collecting the data presented in this article. The Kinemax Outcomes Group consists of the following surgeons and their research assistants: William Gillespie, Colin Howie, Ian Annan, Alastair Gibson, and Judith Lane (Princess Margaret Rose Hospital, Edinburgh, Scotland); Ian Pinder, David Weir, Nigel Brewster, and Karen Bettinson (Freeman Hospital, Newcastle upon Tyne, England); Maurice Needhoff and Roz Jackson (King’s Mill Centre, Mansfield, England); Tim Wilton and Peter Howard (Derbyshire Royal Infirmary, Derby, England); Ian Forster, Paul Szyprt, Chris Moran, David Whitaker, Mike Bullock, and Zena Hinchcliffe (Queen’s Medical Centre, Nottingham, England); Ian Learmonth, John Newman, Chris Ackroyd, George Langkamer, Robert Spencer, Mark Shannon, Evert Smith, John Dixon, and Sarah Whitehouse (Avon Orthopedic Centre, Bristol, England); Clement Sledge, Frederick Ewald, Robert Poss, John Wright, Scott Martin, John Kwon, and Yvette Valderamma (Brigham and Women’s Hospital, Boston, Massachusetts); Steven Harwin and Michael Lichardi (Beth Israel Medical Center, New York, NY); Mark Mehlhoff, Linda Weiler, and Tom Cahalan (Iowa Medical Clinic, Cedar Rapids, Iowa); Richard Cronk and Allyson Sandago (Neuromuscular and Joint Center, Corvallis, Oregon); Stephen Rackemann and Emma McLaughlin (The Knee Centre, Gold Coast, Queensland, Australia); Peter Lewis, Robert Bauze, Gordon Morrison, Tom Stevenson, and Jane Stirling (Queen Elizabeth Hospital, Adelaide, South Australia, Australia); and James Waddell, Emil Schemitsch, and Jane Moreton (Saint Michael’s Hospital, Toronto, Ontario, Canada).
Callahan
CM;
Drake
BG;
Heck
DA;
Dittus
RS. Patient outcomes following tricompartmental total knee replacement. A meta-analysis. JAMA.
1994;271(
17):1349-57.[CrossRef][PubMed]
Pradhan
NR;
Gambhir
A;
Porter
ML. Survivorship analysis of 3234 primary knee arthroplasties implanted over a 26-year period: a study of eight different implant designs. Knee.
2006;13(
1):7-11.[CrossRef][PubMed]
Gandhi
R;
Tsvetkov
D;
Davey
JR;
Mahomed
NN. Survival and clinical function of cemented and uncemented prostheses in total knee replacement: a meta-analysis. J Bone Joint Surg Br.
2009;91(
7):889-95.[CrossRef][PubMed]
Koskinen
E;
Eskelinen
A;
Paavolainen
P;
Pulkkinen
P;
Remes
V. Comparison of survival and cost-effectiveness between unicondylar arthroplasty and total knee arthroplasty in patients with primary osteoarthritis: a follow-up study of 50,493 knee replacements from the Finnish Arthroplasty Register. Acta Orthop.
2008;79(
4):499-507.[CrossRef][PubMed]
Mont
MA;
Mitzner
DL;
Jones
LC;
Hungerford
DS. History of the contralateral knee after primary knee arthroplasty for osteoarthritis. Clin Orthop Relat Res.
1995;(
321):145-50.
McMahon
M;
Block
JA. The risk of contralateral total knee arthroplasty after knee replacement for osteoarthritis. J Rheumatol.
2003;30(
8):1822-4.[PubMed]
Ritter
MA;
Carr
KD;
Keating
EM;
Faris
PM. Long-term outcomes of contralateral knees after unilateral total knee arthroplasty for osteoarthritis. J Arthroplasty.
1994;9(
4):347-9.[CrossRef][PubMed]
Schiphof
D;
Boers
M;
Bierma-Zeinstra
SM. Differences in descriptions of Kellgren and Lawrence grades of knee osteoarthritis. Ann Rheum Dis.
2008;67(
7):1034-6.[CrossRef][PubMed]
Teichtahl
A;
Wluka
A;
Cicuttini
FM. Abnormal biomechanics: a precursor or result of knee osteoarthritis?Br J Sports Med.
2003;37(
4):289-90.[CrossRef][PubMed]
Jackson
BD;
Wluka
AE;
Teichtahl
AJ;
Morris
ME;
Cicuttini
FM. Reviewing knee osteoarthritis—a biomechanical perspective. J Sci Med Sport.
2004;7(
3):347-57.[CrossRef][PubMed]
Lingard
EA;
Berven
S;
Katz
JN; Kinemax Outcomes Group. Management and care of patients undergoing total knee arthroplasty: variations across different health care settings. Arthritis Care Res.
2000;13(
3):129-36.[CrossRef][PubMed]
Lingard
EA;
Katz
JN;
Wright
RJ;
Wright
EA;
Sledge
CB; Kinemax Outcomes Group. Validity and responsiveness of the Knee Society Clinical Rating System in comparison with the SF-36 and WOMAC. J Bone Joint Surg Am.
2001;83(
12):1856-64.[PubMed]
Lingard
EA;
Wright
EA;
Sledge
CB; Kinemax Outcomes Group. Pitfalls of using patient recall to derive preoperative status in outcome studies of total knee arthroplasty. J Bone Joint Surg Am.
2001;83(
8):1149-56.[CrossRef][PubMed]
Lingard
EA;
Katz
JN;
Wright
EA;
Sledge
CB; Kinemax Outcomes Group. Predicting the outcome of total knee arthroplasty. J Bone Joint Surg Am.
2004;86(
10):2179-86.[PubMed]
Lingard
EA;
Sledge
CB;
Learmonth
ID; Kinemax Outcomes Group. Patient expectations regarding total knee arthroplasty: differences among the United States, United kingdom, and Australia. J Bone Joint Surg Am.
2006;88(
6):1201-7.[CrossRef][PubMed]
Miner
AL;
Lingard
EA;
Wright
EA;
Sledge
CB;
Katz
JN; Kinemax Outcomes Group. Knee range of motion after total knee arthroplasty: how important is this as an outcome measure?J Arthroplasty.
2003;18(
3):286-94.[CrossRef][PubMed]
Sangha
O;
Stucki
G;
Liang
MH;
Fossel
AH;
Katz
JN. The Self-Administered Comorbidity Questionnaire: a new method to assess comorbidity for clinical and health services research. Arthritis Rheum.
2003;49(
2):156-63.[CrossRef][PubMed]
Bellamy
N;
Buchanan
WW;
Goldsmith
CH;
Campbell
J;
Stitt
LW. Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. J Rheumatol.
1988;15(
12):1833-40.[PubMed]
Kiebzak
GM;
Campbell
M;
Mauerhan
DR. The SF-36 general health status survey documents the burden of osteoarthritis and the benefits of total joint arthroplasty: but why should we use it?Am J Manag Care.
2002;8(
5):463-74.[PubMed]
Ware
JE
Jr;
Sherbourne
CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care.
1992;30(
6):473-83.[CrossRef][PubMed]
McHorney
CA;
Ware
JE
Jr;
Raczek
AE. The MOS 36-Item Short-Form Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care.
1993;31(
3):247-63.[CrossRef][PubMed]
McHorney
CA;
Ware
JE
Jr;
Lu
JF;
Sherbourne
CD. The MOS 36-item Short-Form Health Survey (SF-36): III. Tests of data quality, scaling assumptions, and reliability across diverse patient groups. Med Care.
1994;32(
1):40-66.[CrossRef][PubMed]
Lingard
EA;
Riddle
DL. Impact of psychological distress on pain and function following knee arthroplasty. J Bone Joint Surg Am.
2007;89(
6):1161-9.[CrossRef][PubMed]
Davison
AC;
Hinkley
DV. Bootstrap methods and their application. Cambridge: Cambridge University Press, 1997.
Jeffery
AE;
Wylde
V;
Blom
AW;
Horwood
JP. “It’s there and I’m stuck with it”: patients’ experiences of chronic pain following total knee replacement surgery. Arthritis Care Res (Hoboken).
2011;63(
2):286-92.[CrossRef][PubMed]
Felson
DT;
Naimark
A;
Anderson
J;
Kazis
L;
Castelli
W;
Meenan
RF. The prevalence of knee osteoarthritis in the elderly. The Framingham Osteoarthritis Study. Arthritis Rheum.
1987;30(
8):914-8.[CrossRef][PubMed]
Zhang
Y;
Xu
L;
Nevitt
MC;
Aliabadi
P;
Yu
W;
Qin
M;
Lui
LY;
Felson
DT. Comparison of the prevalence of knee osteoarthritis between the elderly Chinese population in Beijing and whites in the United States: The Beijing Osteoarthritis Study. Arthritis Rheum.
2001;44(
9):2065-71.[CrossRef][PubMed]
Grotle
M;
Hagen
KB;
Natvig
B;
Dahl
FA;
Kvien
TK. Prevalence and burden of osteoarthritis: results from a population survey in Norway. J Rheumatol.
2008;35(
4):677-84.[PubMed]
Felson
DT;
Zhang
Y;
Hannan
MT;
Naimark
A;
Weissman
BN;
Aliabadi
P;
Levy
D. The incidence and natural history of knee osteoarthritis in the elderly. The Framingham Osteoarthritis Study. Arthritis Rheum.
1995;38(
10):1500-5.[CrossRef][PubMed]
Hanna
FS;
Teichtahl
AJ;
Wluka
AE;
Wang
Y;
Urquhart
DM;
English
DR;
Giles
GG;
Cicuttini
FM. Women have increased rates of cartilage loss and progression of cartilage defects at the knee than men: a gender study of adults without clinical knee osteoarthritis. Menopause.
2009;16(
4):666-70.[CrossRef][PubMed]
Coggon
D;
Reading
I;
Croft
P;
McLaren
M;
Barrett
D;
Cooper
C. Knee osteoarthritis and obesity. Int J Obes Relat Metab Disord.
2001;25(
5):622-7.[CrossRef][PubMed]
Spector
TD;
Hart
DJ;
Doyle
DV. Incidence and progression of osteoarthritis in women with unilateral knee disease in the general population: the effect of obesity. Ann Rheum Dis.
1994;53(
9):565-8.[CrossRef][PubMed]
Keefe
FJ;
Lefebvre
JC;
Egert
JR;
Affleck
G;
Sullivan
MJ;
Caldwell
DS. The relationship of gender to pain, pain behavior, and disability in osteoarthritis patients: the role of catastrophizing. Pain.
2000;87(
3):325-34.[CrossRef][PubMed]
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. No author has had any other relationships, or has engaged in any other activities, 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.