Arthritis & Rheumatism (Arthritis Care & Research) Vol. 55, No. 2, April 15, 2006, pp 314 –321 DOI 10.1002/art.21856 © 2006, American College of Rheumatology ORIGINAL ARTICLE Expectations of Recovery From Revision Knee Replacement VIJI VENKATARAMANAN,1 MONIQUE A. GIGNAC,2 NIZAR N. MAHOMED,3 AND AILEEN M. DAVIS4 Objective. To evaluate outcome expectations of patients undergoing revision total knee replacement (TKR) and to examine personal factors, patient functioning, previous experiences with knee replacement surgery, concerns about surgery, and general health as predictors of expectations. Methods. Revision TKR patients (n ⴝ 184, 54% women; mean age 69 years) completed a questionnaire up to 2 weeks before surgery. This included demographics, experience with previous knee surgery, concerns about surgery, the Life Orientation Test (LOT), the Arthritis Helplessness Scale, the Western Ontario and McMaster Universities Osteoarthritis Index, and a rating of overall health. Outcome expectations were evaluated as 5 questions assessing global beneﬁt; relief of pain; ease of disability; expectations of having complications; and whether the person expected to be fully recovered from surgery in <6 months, 6 –12 months, >12 months, or did not expect to recover. Predictors of each of the 5 outcome expectations were evaluated using univariable and multivariable regression analyses. Results. Expectations are a multidimensional construct (Cronbach’s ␣ ⴝ 0.63). Expectation of global beneﬁt of surgery was high, but was lower for beneﬁts related to ease of pain and improved function. Concerns about surgery were a consistent predictor of all expectation outcomes in multivariable modeling. When concerns about surgery and general health were entered into the model as an interaction with expectation of recovery time as the outcome, past experience (P ⴝ 0.05), pain (P ⴝ 0.03), LOT (P ⴝ 0.03), and interaction between concerns about surgery and general health were signiﬁcant predictors. Conclusion. Clinicians need to understand and help patients shape appropriate expectations for recovery from revision TKR. KEY WORDS. Revision total knee replacement; Recovery; Predictors. INTRODUCTION Osteoarthritis (OA) is a major cause of disability (1), and the knee is the second most commonly affected joint (2). In the US population, 16.3% of 24 –75-year-olds have some form of arthritis or rheumatism and 75% (12.3% of the US Supported by an operating grant from the Canadian Institutes of Health Research. Dr. Davis’s work was supported by a health career award from the Canadian Institutes of Health Research. 1 Viji Venkataramanan, MA: Toronto Rehabilitation Institute, Toronto, Ontario, Canada; 2Monique A. Gignac, PhD: Toronto Western Hospital, University Health Network, and the University of Toronto, Toronto, Ontario, Canada; 3Nizar N. Mahomed, MD, FRCSC, Dphil: Toronto Western Hospital, and the University of Toronto, Toronto, Ontario, Canada; 4 Aileen M. Davis, PhD: Toronto Rehabilitation Institute, and the University of Toronto, Toronto, Ontario, Canada. Address correspondence to Aileen M. Davis, PhD, Room 1119, Toronto Rehabilitation Institute, 550 University Avenue, Toronto, Ontario, Canada M5G 2A2. E-mail: davis. email@example.com. Submitted for publication March 8, 2005; accepted in revised form September 29, 2005. 314 population) have OA (3,4), whereas in Canada 4 million individuals are living with OA. It is estimated that in Canada there will be a 45% increase in the prevalence and an 88% increase in the number of persons affected by arthritis-associated disability between 1985 and 2020 (5). The World Health Organization suggests similar increasing rates in Europe and developing countries as longevity increases (1). For individuals with moderate to severe OA of the knee, total knee replacement (TKR) is a cost-effective treatment because it reduces pain and disability (6). However, as primary joint implants loosen and fail, revision surgery is required. Patients’ expectations that a treatment or behavior will achieve its desired effect are a key factor in understanding health treatments and outcomes (7,8). For example, patient expectations have been linked to requests for elective treatments, compliance with medical regimens, prediction of functional outcomes, and satisfaction with outcomes (9 –14). However, only a few studies (largely retrospective) have reported on the expectations of patients undergoing primary total joint replacement (15–17). Moreover, although Haddad et al (18) and Eisler et al (10) evaluated Revision Knee Replacement expectations in persons undergoing revision total hip replacement, no reports exist of expectations of patients undergoing revision TKR. Previous research has focused on a single dimension of expectations, most often expectations of satisfaction or expectations of returning to a previous level of functioning (13,17,18). We do not know to what extent patients’ expectations form a global evaluation of TKR or are varied and multidimensional in nature. If patients have different types of expectations regarding TKR, this has implications for targeting areas that may need to be addressed to better educate patients and to assist them in decision making. Theories of social cognition have posited that factors such as perceived vulnerability, previous experience, and individual differences may shape individuals’ expectations (7,8). Therefore, examining expectations in revision TKR may be particularly important because patients have previous (unsuccessful) experience with the procedure and may report greater vulnerability or concerns. This, ultimately, may affect the outcome of surgery. As a ﬁrst step in uncovering the role of expectations in revision TKR, we examined personal factors, patient functioning, previous experiences with knee replacement surgery, concerns about surgery and general health, and the relationship of these factors to patients’ expectations prior to receiving revision TKR. PATIENTS AND METHODS Patients. This study was a cross-sectional analysis of patients undergoing revision TKR surgery nationwide (in teaching hospitals in Halifax, London, Toronto, Winnipeg, and Vancouver). Surgeons used their usual criteria (which included a combination of pain, functional deﬁcit, instability, and radiographic loosening or bone loss) for recommending revision TKR surgery. Inclusion criteria for the study were as follows: patient was considered a candidate for revision TKR by the surgeon, had a previous TKR for arthritis, was able to speak and read English, and gave consent to participate. Exclusion criteria included revision TKR for infection, patellar revision only, or polyethylene liner exchange only. Ethics approval was obtained from each participating institution, and all participants consented to participate in the study. Measures. This study focuses on the baseline data of a prospective longitudinal study collecting data over 5 years that will evaluate predictors of various outcomes (i.e., pain, disability, satisfaction) in patients undergoing TKR. Participants completed a set of questionnaires (detailed below) 2 weeks prior to their surgery at their preadmission clinic visit. Demographic data including age, sex, marital status, and education level were collected. We also asked how many prior knee replacement surgeries the individual had on either of their knees. Outcome variables were outcome expectations, and predictor variables were dispositional optimism as measured by the Life Orientation Test (LOT) (19), preoperative pain and physical function as measured by the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) (20 –22), experience 315 with previous TKR, Arthritis Helplessness Scale (AHS) scores (23), perception of control over decision making about having the surgery, concerns about the surgery, and overall general health and general health in relation to the knee. Outcome expectations were evaluated as 5 separate questions. Three questions assessed expectations concerning global beneﬁt, relief of pain, and ease of difﬁculties with daily activities based on a 5-point scale (where 1 ⫽ extremely beneﬁcial and 5 ⫽ not at all beneﬁcial). In question 4, respondents indicated if they expected to have complications (no, yes, or not sure). In question 5, patients were asked whether they expected to be fully recovered from their surgery in ⬍6 months, 6 –12 months, ⬎12 months, or did not expect to ever recover. Our intent was to scale these 5 items into a summative score; however, Cronbach’s alpha was 0.63. (Cronbach’s alpha is a measure of internal consistency. It measures how well a set of items or variables measures a single unidimensional latent construct.) The interitem correlations were low (ranging from 0.14 to 0.41), with item-total correlations ranging from 0.36 to 0.43, explaining the low Cronbach’s alpha value. This low alpha indicates that expectations are a multidimensional construct, and therefore we evaluated each of the 5 expectation questions as separate outcomes. The LOT is a reliable and valid measure of optimism/ pessimism (19) and consists of 12 items: 4 positively worded items, 4 reverse-coded items, and 4 ﬁller items. The score ranges from 8 to 40, indicating extreme optimism and a negative orientation to life, respectively. The AHS is a shortened version of the Arthritis Helplessness Index with similar predictive properties and acceptable reliability (23–27). This 5-item scale, with a maximum score of 25 and minimum score of 5, measures individuals’ perceptions of their helplessness in managing their arthritis. A high score indicates helplessness in managing arthritis. The WOMAC is a reliable and valid measure of symptoms and function for persons with OA of the hip and/or knee, and has been used extensively as an outcome for total joint replacement (20 –22). The WOMAC consists of 3 subscales: pain, stiffness, and physical function. Only the pain and physical function subscales were used in this study. The pain score ranges from 0 to 20 and the physical function scale ranges from 0 to 68. Higher scores indicate more pain or disability. Five questions relating to past experience in terms of recovery time, complications, pain, difﬁculty with daily activities, and beneﬁt experienced from the previous surgery were combined to form the Past Experience Scale (Cronbach’s ␣ ⫽ 0.80). Three questions assessed previous experience with knee replacement surgery concerning global beneﬁt, relief of pain, and ease of difﬁculties with daily activities based on a 5-point scale (where 1 ⫽ extremely beneﬁcial and 5 ⫽ not at all beneﬁcial). Question 4 asked if patients experienced or did not experience complications, and question 5 asked patients whether they recovered from their previous surgery in ⬍6 months, 6 months to 1 year, ⬎1 year, or never recovered. The summative score ranges from 5 (a very positive or good expe- 316 rience with the previous knee surgery) to 21 (a negative or bad experience). Five questions relating to a patient’s concerns in terms of the surgery (overall concern, pain, complications, recovery, and difﬁculty with everyday activities) were graded on a 5-point scale (where 1 ⫽ not at all and 5 ⫽ extremely). These questions were combined to form the Concerns Scale, as internal consistency as measured by Cronbach’s alpha was 0.83. A maximum score of 25 indicates extreme concern whereas the minimum score of 5 indicates no concerns with the surgery. Perception of control was evaluated by asking patients to indicate the extent to which they thought the decision to have the knee revision surgery was outside of their control. Responses were indicated on a 5-point scale ranging from “not at all” to “a great deal.” Patients rated their general health on a visual analog scale with gradation from 0 to 100, where 0 represents “worst imaginable health state” and 100 represents the “best imaginable health state” (28). The knee health rating scale is the same as the general health rating scale, but it records patients’ perceptions of their health with respect to their knee. Statistical analyses. Descriptive statistics were calculated for all the variables. As noted above, Cronbach’s alpha was calculated to determine if there was sufﬁcient homogeneity to create summated scales for the past experiences with surgery, concerns, and expectations questions. Univariate regression analyses were performed to determine the factors inﬂuencing individual expectations, namely expectation of beneﬁt from surgery, expected recovery time from surgery, expectation of complications during recovery from surgery, expectation of pain, and expectation of difﬁculties in performing daily activities. Variables from the univariate regression analysis that were signiﬁcant at P ⬍ 0.10 were entered into multiple linear regression models for each of the 5 expectation questions. Before embarking on the regression analyses, we computed correlations (Kendall’s tau and Spearman’s correlation) for the different domains of expectations and also between the predictor variables to check for high correlation values indicating colinearity. As a ﬁnal component to testing the multivariable models, we tested an interaction between the Concerns Scale and perceived general health to address whether patients with lower levels of health status had more concerns and whether the interaction of the 2 variables was, in turn, related to expectations. We explored this interaction based on the hypothesis that patients who had poorer general health may be more concerned about having surgery and their recovery. P values ⬍0.05 were considered statistically signiﬁcant. RESULTS The sample consisted of 184 patients who were undergoing revision TKR. A total of 99 (53.8%) patients were women. The mean age was 69 years, the youngest patient being 32 years of age and the oldest patient being 89 years of age. A total of 120 (65.6%) were married or with a Venkataramanan et al Table 1. Description of the sample (n ⴝ 184)* Characteristic Value Women Married or with a partner Living alone Bachelor’s degree or higher Other medical problems Help around the house Using walking aid Taking pain medication Age, mean ⫾ SD years (range) Number of previous replacement surgeries, mean ⫾ SD (range) Arthritis Helplessness Score, mean ⫾ SD (range) Life Orientation Test, mean ⫾ SD (range) Concerns About Surgery scale, mean ⫾ SD (range) Past Experience scale, mean ⫾ SD (range) WOMAC pain subscale, mean ⫾ SD (range) WOMAC physical subscale, mean ⫾ SD (range) General health rating by patient, mean ⫾ SD (range) General health with respect to knee, mean ⫾ SD (range) 99 (53.8) 120 (65.6) 48 (26.2) 16 (8.7) 136 (75.1) 140 (76.1) 144 (79.1) 28 (15.4) 69.1 ⫾ 11.3 (31.9–89.4) 1.9 ⫾ 1.1 (1–8) 15.0 ⫾ 4.0 (5.0–25.0) 17.7 ⫾ 4.9 (8.0–34.0) 15.0 ⫾ 4.0 (5.0–25.0) 12.8 ⫾ 4.1 (5.0–21.0) 11.1 ⫾ 3.9 (3.0–20.0) 38.6 ⫾ 12.8 (14.3–52.5) 73.5 ⫾ 18.7 (10–100) 39.2 ⫾ 22.1 (0–90) * Values are the number (percentage) unless otherwise indicated. WOMAC ⫽ Western Ontario and McMaster Universities Osteoarthritis Index. partner; 48 (26.2%) lived alone; and the remainder lived with family members, friends, or in a residential institution. For education level, 8.7% had a bachelor’s degree or higher. On average, patients had 1.9 previous knee replacement surgeries with a maximum of 8 prior surgeries on index and nonindex knees combined (Table 1). The mean ⫾ SD score on the AHS was 15.0 ⫾ 4.0, suggesting that most patients were in the middle range of the scale in reporting helplessness in dealing with their arthritis. The LOT scores revealed that patients were generally optimistic (mean ⫾ SD score 17.7 ⫾ 4.9). The Concerns Scale had a mean ⫾ SD score of 15.0 ⫾ 4.0, indicating that patients were only somewhat concerned about surgery. Patients’ past experience with knee revision surgery was unremarkable as suggested by the mid-range scores (mean ⫾ SD score 12.8 ⫾ 4.1). The WOMAC pain and physical subscales had mean ⫾ SD scores of 11.1 ⫾ 3.9 and 38.6 ⫾ 12.8, respectively. Patients regarded their overall general health as being good (mean ⫾ SD 73.5 ⫾ 18.7) but rated their health as poor with respect to their knee (mean ⫾ SD 39.2 ⫾ 22.1). Most patients had relatively high expectations (Table 2). A total of 54.9% believed the upcoming surgery would be extremely beneﬁcial in helping them resume their everyday activities (group mean ⫾ SD 1.7 ⫾ 0.8). Half of the patients undergoing revision TKR (51.1%) expected to recover in ⬍6 months after surgery, whereas a slightly lower Revision Knee Replacement 317 Table 2. Percentage of response, number of patients, and mean ⴞ SD for each of the 5 expectations Expectation/response options Global beneﬁt 1 (extremely) 2 3 (somewhat) 4 5 (not at all) Pain 1 (extremely) 2 3 (somewhat) 4 5 (not at all) Difﬁculty 1 (extremely) 2 3 (somewhat) 4 5 (not at all) Recovery time, months ⬍6 6–12 ⬎12 Do not expect to recover Complications No Not sure Yes Mean ⴞ SD No. (%) 1.7 ⫾ 0.8 100 (54.9) 43 (23.6) 37 (20.3) 1 (0.5) 1 (0.5) 3.4 ⫾ 0.9 27 (14.8) 35 (19.2) 108 (59.3) 8 (4.4) 4 (2.2) beneﬁcial effects from surgery, with a trend for age (P ⫽ 0.08). Expectations of recovery time. Expectation of a shorter recovery time was associated with having low concerns about the upcoming surgery (P ⬍ 0.0001), a positive past experience with surgery (P ⬍ 0.0001), experiencing pain (P ⬍ 0.0001), having difﬁculty with physical functioning (P ⬍ 0.0001), low feelings of helplessness (P ⫽ 0.001), optimism (P ⫽ 0.001), and higher education (P ⫽ 0.04), as was evident from univariate regression analyses. Expectation of a faster recovery was predicted by low concerns about surgery (P ⫽ 0.003), with a trend for positive past surgical experiences (P ⫽ 0.06) and higher education (P ⫽ 0.06) when the variables from the above univariate analyses were entered into a multiple linear regression model. 3.2 ⫾ 0.9 22 (12.1) 26 (14.3) 117 (64.3) 9 (4.9) 8 (4.4) 1.5 ⫾ 0.6 93 (51.1) 82 (45.1) 6 (3.3) 1 (0.5) 1.4 ⫾ 0.6 108 (59.0) 68 (37.2) 7 (3.8) number of patients (45.1%) thought they would recover between 6 months and 1 year. The remainder believed that it would take longer than 1 year or that they would never fully recover. In terms of expecting complications, more than half the patients (59%) did not expect to experience any complications. A total of 59.3% expected the recovery to be somewhat painful. Expectation of difﬁculty with respect to everyday activities during recovery was similar to expectation of pain, with 64.3% expecting a somewhat difﬁcult recovery, 14.3% expecting between somewhat and extreme difﬁculty, and 12.1% expecting extreme difﬁculty. The others thought it would be somewhat or not at all difﬁcult. Predictor variables were not highly correlated, with the correlation values ranging from 0.01 to 0.59. WOMAC pain and physical subscales had the highest correlation (r ⫽ 0.59). Expectations of beneﬁt. Univariate regression analyses revealed that low concerns about surgery (P ⬍ 0.0001), optimism (P ⫽ 0.001), positive past experience (P ⫽ 0.005), perceiving overall health as good (P ⫽ 0.01), older age (P ⫽ 0.02), and seeing one’s self as being responsible for the decision to have surgery (P ⫽ 0.037) were associated with perceptions of increased beneﬁts of surgery. However, when the above variables were entered into a multiple linear regression model (Table 3), only low concerns about surgery (P ⫽ 0.02) was predictive of perceived Expectations of complications. Univariate regression analyses revealed that being less concerned about surgery (P ⬍ 0.0001), having a positive past surgery experience (P ⫽ 0.001), optimism (P ⫽ 0.002), and not feeling helpless (P ⫽ 0.04) were all associated with low expectations of complications, whereas older patients (P ⫽ 0.006) and those who did not rate their general health as high (P ⫽ 0.02) expected complications to arise while recovering from surgery. Multiple regression analysis revealed that concerns about surgery (P ⫽ 0.004) and age (P ⫽ 0.05) were the only independent predictors of expectation of complications during recovery from surgery. Expectations of pain. Univariate regression analyses demonstrated that expectations of pain were low for women (P ⫽ 0.08), when concerns about surgery were low (P ⬍ 0.0001), when patients had a positive past experience with surgery (P ⬍ 0.0001), when current pain was low (P ⫽ 0.01), when patients did not feel helpless (P ⫽ 0.03), and when physical functioning was higher (P ⫽ 0.04). Expectations of pain were high when the patients were older (P ⫽ 0.04), had fewer numbers of previous replacement surgeries (P ⫽ 0.10), or perceived themselves to have poorer overall health (P ⫽ 0.04). The only signiﬁcant predictor of expectations of pain was concerns about surgery (P ⬍ 0.0001) when the above variables were entered into the multiple regression analysis. Expectations of difﬁculty with everyday activities. Univariate regression analyses showed that patients expected less difﬁculty with everyday activities after surgery when they were less concerned about surgery (P ⬍ 0.0001), female (P ⫽ 0.08), had a positive past experience with surgery (P ⫽ 0.002), and did not feel helpless (P ⫽ 0.02). Upon entering the above variables into a multiple linear regression model, expectation of experiencing fewer difﬁculties with everyday activities after surgery was predicted by low concerns about surgery (P ⫽ 0.03). There was a trend toward signiﬁcance for past experience with surgery (P ⫽ 0.07). Interaction. An interaction between concerns about surgery and general health as perceived by the patient was 318 Venkataramanan et al Table 3. Predictor variables (P < 0.10) in multivariable regression models* Predictor variables Expectation of beneﬁt (R2 ⫽ 0.17; n ⫽ 162) Concerns about surgery Life Orientation Test Past Experience Scale General health Age Responsible for decision Expected recovery time (R2 ⫽ 0.27; n ⫽ 164) Concerns about surgery Past Experience Scale WOMAC pain subscale WOMAC physical subscale Arthritis Helplessness Scale Life Orientation Test Education Expectation of complications (R2 ⫽ 0.16; n ⫽ 162) Concerns about surgery Past Experience Scale Life Orientation Test Age General health Arthritis Helplessness Scale Expectation of pain (R2 ⫽ 0.16; n ⫽ 160) Concerns about surgery Past Experience Scale WOMAC pain subscale WOMAC physical subscale Arthritis Helplessness Scale General health Age Sex Previous surgeries Expectation of difﬁculty (R2 ⫽ 0.22; n ⫽ 170) Concerns about surgery Past Experience Scale Arthritis Helplessness Scale Sex ␤ SE t P 0.037 0.018 0.015 ⫺0.003 ⫺0.360 0.079 0.015 0.014 0.018 0.004 0.208 0.049 2.435 1.316 0.841 ⫺0.928 ⫺1.726 1.612 0.016 0.190 0.402 0.355 0.086 0.109 0.029 0.022 0.030 0.002 ⫺0.010 0.015 ⫺0.028 0.010 0.011 0.019 0.006 0.014 0.009 0.015 2.970 1.910 1.568 0.336 ⫺0.743 1.646 ⫺1.896 0.003 0.058 0.119 0.737 0.459 0.102 0.06 0.029 0.008 0.012 ⫺0.273 0.000 0.000 0.010 0.012 0.009 0.136 0.002 0.012 2.952 0.733 1.291 ⫺2.011 0.020 0.019 0.004 0.465 0.199 0.046 0.984 0.985 0.064 0.020 0.039 ⫺0.008 ⫺0.0013 ⫺0.001 ⫺0.222 ⫺0.109 ⫺0.038 0.016 0.018 0.030 0.009 0.022 0.004 0.216 0.134 0.061 4.085 1.100 1.281 ⫺0.926 ⫺0.598 ⫺0.301 ⫺1.030 ⫺0.814 ⫺0.613 ⬍ 0.0001 0.273 0.202 0.356 0.551 0.764 0.305 0.417 0.541 0.032 0.032 0.007 ⫺0.207 0.015 0.018 0.018 0.131 2.144 1.781 0.389 ⫺1.581 0.034 0.077 0.698 0.116 * WOMAC ⫽ Western Ontario and McMaster Universities Osteoarthritis Index. entered into the above multivariate models to understand if patients with lower levels of health status had more concerns and if the interaction of the 2 variables was, in turn, related to expectations. This interaction was signiﬁcant in the model with expected recovery time as the outcome variable. Including this interaction term in the multivariable model resulted in optimism (P ⫽ 0.03), experiencing pain (P ⫽ 0.03), and the interaction between concerns about surgery and perceived general health (P ⫽ 0.03) becoming signiﬁcant predictors, along with having a good past experience with surgery (P ⫽ 0.05). However, concerns about surgery were no longer a signiﬁcant predictor, indicating that concerns about surgery were moderated by patients’ perceptions of their general overall health in predicting expectations about recovery time from surgery (Figure 1). Speciﬁcally, the interaction revealed that at low levels of general health, patients’ concerns about surgery did not inﬂuence their expectations of recovery. Most of these individuals believed they would recover within a year. Among those with higher ratings of Figure 1. Interaction between general health rating and concerns about surgery in relation to expected recovery time. 1 ⫽ recover in ⬍6 months; 2 ⫽ recover in 6 months to a year; 3 ⫽ recover in more than a year; 4 ⫽ do not expect to ever recover. Revision Knee Replacement general health, a different pattern emerged. Despite having better reported health, patients with high ratings of general health who reported greater concerns about surgery reported the longest expected recovery times compared with all other groups. When the interaction between concerns for surgery and general health as rated by the patient was replaced with the interaction between concerns about surgery and WOMAC physical subscale score, no signiﬁcant change was noted in any model except for the expected recovery time model. Patients’ previous experiences with surgery (P ⫽ 0.08), optimism (P ⫽ 0.09), higher education (P ⫽ 0.06), and the interaction between concerns about surgery and WOMAC physical functioning scores (P ⫽ 0.07) predicted patients’ expectation of a faster recovery. When the interaction variable in the above models for the 5 expectations was replaced with the interaction between concerns about surgery and WOMAC pain subscale score, no signiﬁcant change was found in any of the models. DISCUSSION Revision total joint replacement is more complex than primary replacement, and with increased volume of primary total joint replacement, revision total joint replacement rates are also increasing (29). Currently, there are little data available on the outcomes of revision TKR (30) with which to guide patients, and there are no data on what patients expect after revision TKR. Previous work has evaluated expectations of patients undergoing primary hip or knee replacement or revision hip replacement with the intent of evaluating how expectations relate to functional outcome and satisfaction with outcome (13,15–18). To our knowledge, this is the ﬁrst study that has evaluated expectations of patients undergoing revision TKR and that has attempted to understand factors that shape patients’ expectations about their recovery from surgery. In capturing expectations, we included an overall question of expected beneﬁt as well as several questions related to speciﬁc beneﬁts expected in terms of symptoms and function. When tested, these items were not homogeneous based on a low Cronbach’s alpha, suggesting that expectations about recovery following revision TKR are formed from distinct components. If expectations were a homogeneous concept, we would also expect similar factors to predict expectations. However, our modeling of various factors demonstrated that the predictors varied depending on which expectation question was the dependent variable. In the total joint replacement literature, expectations have been evaluated by individual questions similar to our study and by summing the responses to individual items (13,15–18). It is unclear if these summed scales reﬂect a homogeneous concept. The multidimensional nature of expectations revealed in this study suggests that studies need to evaluate multiple aspects of expectations and ultimately look at how these different expectations relate to outcome. Also, further work needs to be done to determine if individuals have additional expectations that have not been evaluated. Despite the fact that their knee replacement failed, pa- 319 tients expected the overall beneﬁt from their revision TKR to be high (mean 1.68); however, their expectations of speciﬁc beneﬁts in pain relief and disability were less positive (mean 3.40 and 3.25, respectively). (For these questions, 1 indicated a higher level of expectation.) In this study, the indications for revision TKR included pain, functional deﬁcit, instability, and radiographic loosening or bone loss. It is rare that radiographic loosening, bone loss, or instability would occur without the patient experiencing pain and/or functional compromise. It is, therefore, unexpected that patients would have high global expectations with much lower expectations for pain relief and functional recovery. This suggests that global expectations may represent additional speciﬁc expectations that were not evaluated by our questions. In this study, most patients reported relatively high expectations from their surgery. Although there is a fair amount of variability in patients’ expectations of recovery, complications, pain, and potential interference with daily activities, on average patients expect the recovery to be without complications, the surgery to be extremely beneﬁcial, some pain, and some amount of difﬁculty with daily activities. Patients expect to be fully recovered or to be near full recovery in 6 months. Similar results of high expectations have been reported in studies on patients who underwent primary total joint replacement and revision total hip replacement (10,16 –18). Saleh et al concluded that revision TKR is a beneﬁcial procedure (30). However, their meta-analysis was based on a clinician’s global ratings of pain, function, and range of motion rather than the patients’ perceptions of their function, and the data provide no indication of the timeline of recovery. Only through a longitudinal study of recovery will we be able to determine if these expectations match true recovery and predict outcome. Although research shows that past experience may shape individuals’ expectations (7,8), there seems to be no known scale measuring past experience of patients, especially in relation to surgery outcomes. To determine the relevant domains of interest, we drew extensively from research on total joint replacements, consequently suggesting the outcomes, namely pain, recovery, difﬁculty resuming daily activities, complication, and beneﬁt to be the most relevant to surgery (10,12,15–17,30,31). Our preliminary analyses of the scale indicate an acceptable internal reliability (Cronbach’s ␣ ⫽ 0.80). Past experience was only predictive of expectation of recovery time. There may be other aspects of past experience that might be relevant to the other domains of expectation. More research is needed to assess how patients’ previous experiences might be a factor associated with outcomes of surgery, and also to validate the domains of expectation in other samples. Sex and individual differences in optimism and perceptions of helplessness related to arthritis were not related to the expectations that patients have about revision TKR. However, patients’ concerns about surgery were a signiﬁcant factor in shaping their expectations of global beneﬁt, complications, pain relief, and difﬁculty with daily activities. Even considering that we had multiple expectation outcomes, these ﬁndings are consistent, with the exception of expectations of difﬁculty with daily activities, had 320 a Bonferroni correction been used where a signiﬁcant P value would be 0.01. In a study of patients undergoing primary total knee and total hip arthroplasty, Trousdale et al also found patients to be concerned about pain after surgery and the length of recovery (31). These results suggest that individual, relatively stable characteristics such as sex and optimism may not be as important in determining expectations as situational or speciﬁc contextual factors such as concerns or worries about different aspects of surgery. Concerns about surgery interacted with general health status to predict expected time to recovery. When patients rated their overall health as low, there was little variability in the recovery time, irrespective of concerns about surgery. This expectation of a relatively short recovery time regardless of concerns about surgery may mean that, at low levels of general health, patients are focused primarily on their current health and are motivated to do something to alleviate their pain and disability. As a result, they expect that surgery will allow them some degree of recovery in a relatively short time. The implications for this expectation warrant further study, particularly because older patients with more comorbidities are undergoing revision surgery (29) and their expectations of recovery time may not be realized. When general health is higher, the focus may not only be on improving health status but also on the costs and beneﬁts of the surgery itself. Individuals with concerns may be more likely to question their decision to go ahead with surgery and more likely to expect a longer recovery time. It is unclear whether these expectations will be more realistic in terms of actual outcomes than the expectations of individuals in other groups. Future research would beneﬁt from examining the concordance of patients’ expectations of recovery with outcomes. In summary, most of the patients undergoing revision TKR in this study expected their surgery to be extremely beneﬁcial. Most expected to recover within a year of surgery, and more than half expected not to experience complications after surgery but to have a somewhat painful and difﬁcult recovery period with respect to performing everyday activities. Concerns about surgery are the most signiﬁcant factor inﬂuencing expectations about revision TKR surgery. In addition, past experience also inﬂuenced expectation of recovery time, while age seemed to inﬂuence expectation of complications. Understanding these individual expectations has implications for targeting areas that may need to be addressed to better educate patients and to assist them in understanding their recovery and the extent of recovery following revision TKR, particularly as work in primary joint replacement has shown that expectations are a predictor of outcome (17). ACKNOWLEDGMENTS We would like to acknowledge the contributions of the lead surgeons at each of the centers where patients were accrued to this study: Dr. Michael Dunbar, Queen Elizabeth II, Halifax, Nova Scotia; Dr. Steve MacDonald, London Health Sciences Center, London, Ontario; Dr. Nizar Mahomed, Toronto Western Hospital, Toronto, Ontario; Dr. Allan Gross, Mount Sinai Hospital, Toronto, Ontario; Venkataramanan et al Dr. Jeffrey Gollish, Sunnybrook and Women’s College Health Science Center, Toronto, Ontario; Dr. Emil Schemitsch, St. Michael’s Hospital, Toronto, Ontario; Dr. David Hedden, St. Boniface General Hospital, Winnipeg, Manitoba; Dr. Donald Garbuz, Vancouver General Hospital, Vancouver, British Columbia. REFERENCES 1. Kaplan W, Laing R. Priority medicines for Europe and the world. Geneva: World Health Organization, Department of Essential Drugs and Medicines Policy; 2004. 2. Doherty M, Dougados M. Evidence-based management of osteoarthritis: practical issues relating to the data. Best Pract Res Clin Rheumatol 2001;15:517–25. 3. Elders MJ. The increasing impact of arthritis on public health. J Rheumatol Suppl 2000;60:6 – 8. 4. Lawrence RC, Helmick CG, Arnett FC, Deyo RA, Felson DT, Giannini EH, et al. Estimates of the prevalence of arthritis and selected musculoskeletal disorders in the United States. Arthritis Rheum 1998;41:778 –99. 5. Badley EM, Wang PP. Arthritis and the aging population: projections of arthritis prevalence in Canada 1991 to 2031. J Rheumatol 1998;25:138 – 44. 6. Laupacis A, Bourne R, Rorabeck C, Feeny D, Wong C, Tugwell P, et al. Costs of elective total hip arthroplasty during the ﬁrst year: cemented versus noncemented. J Arthroplasty 1994;9: 481–7. 7. Bandura A. Self-efﬁcacy: toward a unifying theory of behavioral change. Psychol Rev 1977;84:191–215. 8. Bandura A. On rectifying conceptual ecumenism. New York: Plenum Press; 1995. 9. Ross CK, Sinacore JM, Stiers W, Budiman-Mak E. The role of expectations and preferences in health care satisfaction of patients with arthritis. Arthritis Care Res 1990;3:92– 8. 10. Eisler T, Svensson O, Tengstrom A, Elmstedt E. Patient expectation and satisfaction in revision total hip arthroplasty. J Arthroplasty 2002;17:457– 62. 11. Ibrahim SA, Siminoff LA, Burant CJ, Kwoh CK. Differences in expectations of outcome mediate African American/white patient differences in “willingness” to consider joint replacement. Arthritis Rheum 2002;46:2429 –35. 12. Leedham B, Meyerowitz BE, Muirhead J, Frist WH. Positive expectations predict health after heart transplantation. Health Psychol 1995;14:74 –9. 13. Burton KE, Wright V, Richards J. Patients’ expectations in relation to outcome of total hip replacement surgery. Ann Rheum Dis 1979;38:471– 4. 14. Mondloch MV, Cole DC, Frank JW. Does how you do depend on how you think you’ll do? A systematic review of the evidence for a relation between patients’ recovery expectations and health outcomes [published erratum appears in CMAJ 2001;165:1303]. CMAJ 2001;165:174 –9. 15. Mancuso CA, Salvati EA, Johanson NA, Peterson MG, Charlson ME. Patients’ expectations and satisfaction with total hip arthroplasty. J Arthroplasty 1997;12:387–96. 16. Mancuso CA, Sculco TP, Wickiewicz TL, Jones EC, Robbins L, Warren RF, et al. Patients’ expectations of knee surgery. J Bone Joint Surg Am 2001;83-A:1005–12. 17. Mahomed NN, Liang MH, Cook EF, Daltroy LH, Fortin PR, Fossel AH, et al. The importance of patient expectations in predicting functional outcomes after total joint arthroplasty. J Rheumatol 2002;29:1273–9. 18. Haddad FS, Garbuz DS, Chambers GK, Jagpal TJ, Masri BA, Duncan CP. The expectations of patients undergoing revision hip arthroplasty. J Arthroplasty 2001;16:87–91. 19. Scheier MF, Carver CS. Optimism, coping, and health: assessment and implications of generalized outcome expectancies. Health Psychol 1985;4:219 – 47. Revision Knee Replacement 20. 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;15:1833– 40. 21. Bellamy N. WOMAC osteoarthritis index: a user’s guide. London: University of Western Ontario; 1995. 22. McConnell S, Kolopack P, Davis AM. The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC): a review of its utility and measurement properties. Arthritis Rheum 2001;45:453– 61. 23. DeVellis RF, Blalock SJ. Psychological and educational interventions to reduce arthritis disability. Baillieres Clin Rheumatol 1993;7:397– 417. 24. Devellis RF, Callahan LF. A brief measure of helplessness in rheumatic disease: the helplessness subscale of the Rheumatology Attitude Index. J Rheumatol 1993;20:866 –9. 25. Stein MJ, Wallston KA, Nicassio PM. Factor structure of the Arthritis Helplessness Index. J Rheumatol 1988;15:427–32. 26. Gignac MA, Cott C, Badley EM. Adaptation to disability: applying selective optimization with compensation to the 321 27. 28. 29. 30. 31. behaviors of older adults with osteoarthritis. Psychol Aging 2002;17:520 – 4. Gignac MA. An evaluation of a psychotherapeutic group intervention for persons having difﬁculty coping with musculoskeletal disorders. Soc Work Health Care 2000;32:57– 75. Brooks R, Rabin R, de Charro F. The measurement and valuation of health status using EQ-5D: a European perspective. Dordrecht, The Netherlands: Kluwer Academic Publishers; 2003. Canadian Institute for Health Information. Canadian joint replacement registry report 2004. Ottawa, Ontario, Canada: CIHI; 2004. Saleh KJ, Dykes DC, Tweedie RL, Mohamed K, Ravichandran A, Saleh RM, et al. Functional outcome after total knee arthroplasty revision: a meta-analysis. J Arthroplasty 2002;17: 967–77. Trousdale RT, McGrory BJ, Berry DJ, Becker MW, Harmsen WS. Patients’ concerns prior to undergoing total hip and total knee arthroplasty. Mayo Clin Proc 1999;74:978 – 82.