Abstract
Objectives
Determine if severity of osteoarthritis-related knee pain is associated with willingness to undergo total knee replacement (TKR) and if this association is confounded or modified by components of socioeconomic status and healthcare coverage.
Methods
Cross-sectional analysis was conducted among 3530 Osteoarthritis Initiative study participants. Logistic regression models were used to assess the effect of knee pain severity (0: none, 1–3: mild, 4–7: moderate, 8–10: severe) on willingness to undergo TKR. Stratified analyses were conducted to evaluate whether socioeconomic status and healthcare coverage modify the effect of knee pain severity on willingness.
Results
Participants with severe knee pain, compared to participants without pain, were less willing to undergo TKR (OR 0.73, 95% CI 0.57–0.93). This association was attenuated when adjusted for age, sex, comorbidity, depression, health insurance coverage, prescription medicine coverage, healthcare source, education, income, employment, race, and marital status (adjusted OR 0.92, 95% CI 0.68–1.24).
The odds of willingness to undergo TKR were significantly lower in those with the highest level of pain compared to those without pain, among participants without health insurance (adjusted OR 0.08, 95% CI 0.01–0.56) but not among those with health insurance (adjusted OR 1.03, 95% CI 0.73–1.38) when adjusted for demographic, clinical, healthcare access and socioeconomic factors (p=0.015). However, <5% of participants were without health insurance.
Conclusion
Among participants without health insurance, severe knee pain was paradoxically associated with less willingness to undergo TKR. Policies that improve access to quality healthcare may affect patient preferences and increase utilization of TKR surgery among vulnerable populations.
INTRODUCTION
Multiple studies have demonstrated that total knee replacement (TKR) surgery leads to long term improvements in pain, physical functioning and quality of life in patients with advanced osteoarthritis (OA) (1–5). It is also a cost-effective surgical procedure (6). Therefore, patient interest in TKR would be expected to increase with greater OA disease severity. However, when Llewellyn-Thomas et al (7) conducted interviews with patients awaiting total joint replacement surgery, no predictive pattern emerged. Similar findings were reported by Hawker et al (8) and by Suarez-Almazor et al (9) who both found that willingness to have the procedure was not significantly related to OA disease severity. These findings highlight that patient treatment preference for and eventual use of TKR surgery cannot be easily predicted by OA disease severity alone.
Decision-making is a complicated phenomenon which involves “gathering of information on available choices, needs, risks, and benefits and then the systematic weighing or analyzing of alternatives with the aim of achieving the highest gain at the least cost” (10). For instance, patients with ulcerative colitis are willing to accept serious adverse risks of complications from immunosuppressive medicines in order to avoid a permanent ostomy (11). Similarly, an extensive amount of accommodation to pain and disability is noted in patients with severe OA in order to avoid total joint replacement surgery (10). African-Americans and women may be less willing to consider arthroplasty despite reporting greater disease burden from OA (12–17). Preference for joint replacement surgery may be influenced by socioeconomic factors, access to care, beliefs about procedure efficacy, familiarity with the procedure, extent of social support, trust in healthcare, use of non-surgical coping and pain management strategies, and expectations about post-operative recovery and potential complications (12–14, 16, 18).
Patients with post-secondary education and those with paid employment are more likely to demand knee or hip surgery (16). In the US, health insurance coverage may also affect demand for joint replacement (16). Hawker et al (15) studied the residents of Ontario, Canada, where everyone has universal health insurance coverage, and reported underuse of arthroplasty. This underutilization is expected to be worse in countries such as the US where health insurance status poses a further barrier to access to health care (15). Yet, utilization of total joint replacement surgery in the US increases after age 65 partly due to Medicare eligibility and supplemental health insurance coverage as well as the increased prevalence of OA among older age groups (19). How socioeconomic status and access to health insurance may confound or change the relationship between OA disease severity and willingness to undergo TKR surgery is unknown.
The primary objective of the study is to estimate the effect of OA-related knee pain severity on willingness to undergo knee replacement surgery. The secondary objective is to assess the extent to which socioeconomic status and healthcare coverage may confound or modify the impact of knee pain severity on willingness. We hypothesized that participants with greater knee pain severity would be more willing to undergo knee replacement surgery; however, poor socioeconomic status and healthcare access might obscure the expected causal impact of pain on inclination towards treatment-seeking behavior.
METHODS
Participants
The Osteoarthritis Initiative (OAI) is a multi-center longitudinal cohort study that primarily recruited patients with or at high risk of developing symptomatic knee OA based on risk factors (e.g., knee symptoms, overweight, knee injury, knee surgery, and family history of TKR) and radiologic findings. A full description of the OAI study protocol, design, and datasets are available for access at http://www.oai.ucsf.edu/datarelease/. Briefly, 4796 men and women ages 45–79 were initially enrolled between February 2004 and May 2006 from the following study sites: University of Maryland School of Medicine and the Johns Hopkins University (Baltimore, MD), Ohio State University (Columbus, OH), University of Pittsburgh (Pittsburgh, PA) and Memorial Hospital of Rhode Island (Pawtucket, RI). Participants were first asked about willingness to undergo TKR at the 72 month clinic visit; thus the current study is a cross-sectional analysis of knee pain severity and willingness to undergo TKR at the 72 month visit. Participants who responded to both questions about willingness to undergo knee replacement surgery and knee pain severity were included. Those who had previously undergone TKR based on self-report prior to this follow-up visit were excluded.
Study Outcome Variable: Willingness
Participants were asked the question: “If your knee pain were ever to get severe enough, would you be willing to have knee replacement surgery if your doctor recommended it?” (15). Responses were dichotomized as willing (“definitely willing” and “probably willing”) or unwilling (“unsure,” “probably not willing” and “definitely not willing”), as in past publications (12–14, 18). The measure has good face and construct validity and is associated with receipt of physician recommendation for TKR and actual TKR (20, 21).
Primary Exposure Variable: Knee Pain Severity
Knee pain severity was ascertained based on response to the question: “Please rate the pain that you’ve had in your right (or left) knee during the past 30 days by picking a number from 0 to 10 that best describes the pain at its worst.” 0 corresponds to having “no pain. If a participant reported pain in both knees, the score in the knee with more pain was used in the analysis. Positive responses were categorized based on roughly equally sized intervals along the pain scale: Mild (1–3), Moderate (4–7), Severe (8–10). The measure has high test-retest reliability (r=0.95) and has good construct validity, highly correlating to the visual analogue pain scale (0.86–0.95)(22, 23).
The WOMAC – Pain subscale score was used as an alternative measure of knee pain severity (24). 0 corresponds to having “no OA-related pain.” Positive responses were categorized based on approximate tertiles of the remaining non-zero distribution: Low (1–2), Medium (3–5), Moderate to High (6–20).
Covariates and Potential Effect Modifiers/Confounders
Clinical Characteristics
Participant sex, age and race were noted. Comorbidity burden was assessed using the Charlson comorbidity index (25) score (range: 0–12) and depression was measured using the Center for Epidemiologic Studies Depression (CES-D) scale (26) score (range: 0–60). Use of any medication for OA symptoms of either knee in the past 12 months was ascertained based on self-report. Central readings of radiologic OA disease severity of the knee was assessed using the Kellgren-Lawrence (K-L) scoring system (27) using x-rays from the 48 month visit, as x-rays were not scored for K-L grade at the 72-month clinic visit for the entire cohort. K-L grade class is defined based on the K-L score of both knees (Class I: 0, 0; Class II: 0, 1; Class III: 1, 1; Class IV: 0,≥2; Class V: 1,≥2; Class VI:≥2,≥2).
Healthcare Access
Healthcare access at the 72-month clinic visit was assessed based on the following components: (1) presence of any health care/insurance coverage (“Do you currently have any kind of health care coverage? This would include private health insurance, prepaid plans, Preferred Provider Organizations, or any government-sponsored plans, such as Medicare, Medicaid, or VA coverage.”), (2) prescription medicine coverage (“Do you have any health insurance plan that pays for all or part of the cost of prescription medicines?”), and (3) usual location for healthcare or advice about healthcare (private doctor, health maintenance organization [HMO] physician, public clinic, hospital clinic, emergency room [ER], other).
Socioeconomic Status
Self-reported race, education attainment, and annual household income were available from the OAI baseline visit. Employment (works for pay, not working in part due to health/other reasons) and marital status (married/living with partner or not) were ascertained at the 72 month clinic visit.
Statistical Analyses
Participants’ demographic, clinical, healthcare access and socioeconomic characteristics were compared by knee pain severity score category, with the purpose of describing potential confounders of the association between knee pain severity and willingness to undergo knee replacement surgery.
Odds ratios (ORs) and 95% confidence intervals (CI) were estimated using logistic regression models to assess the effect of knee pain severity on willingness to undergo knee replacement surgery. Covariates/potential confounders were selected a priori for inclusion in the adjusted models. The initial model only included knee pain severity (categorized) as the independent variable, with adjustment for age and sex. The following groups of covariates/potential confounding variables were then added in a hierarchical fashion: clinical characteristics, healthcare access variables, and socioeconomic status. A greater than 10% change in OR after the addition of a group of variables suggests that they are confounders in the multivariable model (28).
Whether or not race would modify the association between willingness and knee pain severity was determined by evaluating the interaction between race and pain severity in the full model and stratifying the logistic regression analyses by race.
Stratified analyses were also conducted to determine if the effect of knee pain severity (categorized) on willingness was modified by individual components of socioeconomic status and healthcare access variables. Models were appropriately adjusted for demographic, clinical, socioeconomic, and healthcare access variables. Interactions between knee pain severity (categorized) and each of the potential modifiers were evaluated by comparing models with and without interaction terms using likelihood ratio tests. A likelihood ratio test p-value of <0.05 suggests that the variable is a modifier.
As a sensitivity analysis, WOMAC Pain score, instead of knee pain severity score, was used as the independent variable. Sensitivity analysis of the impact of missing covariate data was also assessed using multiple imputation with chained equations (29).
Data management and analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC), and Figures were generated using Stata version 14.0 (StataCorp, College Station, TX).
RESULTS
A total of 3530 OAI participants responded to the willingness question and reported knee pain severity at the 72-month clinic visit, with no prior knee replacement. 81.1% were White, and 63.1% had at least a college degree. Half were employed, and 63.1% reported being married. More than 95% had health insurance. Participants not included in the analysis were predominantly lost to follow-up before the willingness question was administered; they were more likely to be African-American, less educated, and less likely to be employed (data not shown).”
Table 1 shows the demographic, clinical, healthcare access, and socioeconomic status of OAI participants by knee pain severity categories. Participants who reported no pain were older on average (mean 68.2 years, SD 9.3) compared to those who reported severe pain (mean 66.0 years, SD 8.6). Charlson comorbidity index score (mean 0.94 ± 1.43) and CES-D score (mean 10.53 ± 9.62) were highest among those the highest level of pain severity. Increasing OA severity with higher levels of reported knee pain severity was reflected on radiography based on K-L grade class.
Table 1.
Participants’ characteristics by knee pain severity
| Characteristic | 0: No pain (n=989) | 1–3 (n=1011) | 4–7 (n=1139) | 8–10 (n=391) |
|---|---|---|---|---|
| Age, n (%) | ||||
| 50–59 | 216 (21.8) | 296 (29.3) | 327 (28.7) | 102 (26.1) |
| 60–64 | 172 (17.4) | 187 (18.5) | 217 (19.1) | 88 (22.5) |
| 65–74 | 295 (29.8) | 321 (31.8) | 349 (30.6) | 121 (31.0) |
| 74–85 | 306 (30.9) | 207 (20.5) | 246 (21.6) | 80 (20.5) |
| Age, mean (SD) years | 68.19 (9.26) | 65.94 (8.61) | 66.05 (9.03) | 65.97 (8.63) |
| Sex, n (%) Male | 428 (43.3) | 440 (43.5) | 646 (40.7) | 129 (33.0) |
| Race, n (%) | ||||
| Black or African-American | 147 (14.9) | 93 (9.2) | 208 (18.3) | 140 (35.9) |
| White or Caucasian | 826 (83.5) | 899 (88.9) | 901 (79.2) | 238 (61.0) |
| Other Non-white | 9 (0.9) | 10 (1.0) | 21 (1.9) | 10 (2.6) |
| Asian | 7 (0.7) | 9 (0.9) | 8 (0.7) | 2 (0.5) |
| Education, n (%) | ||||
| Less than high school graduate | 22 (2.2) | 14 (1.4) | 29 (2.6) | 24 (6.2) |
| High school graduate | 122 (12.4) | 82 (8.1) | 134 (11.8) | 65 (16.8) |
| Some college | 195 (19.8) | 197 (19.6) | 288 (25.4) | 116 (30.1) |
| College graduate | 219 (22.2) | 252 (25.0) | 260 (22.9) | 61 (15.8) |
| Some graduate school | 92 (9.3) | 98 (9.7) | 73 (6.4) | 41 (10.6) |
| Graduate degree | 337 (34.1) | 364 (36.2) | 351 (30.9) | 79 (20.5) |
| Income, n (%) | ||||
| Less than $10K | 18 (2.0) | 13 (1.4) | 42 (3.9) | 28 (7.9) |
| $10K to < $25K | 73 (7.9) | 62 (6.5) | 104 (9.8) | 55 (15.5) |
| $25K to < $50K | 220 (23.9) | 208 (21.9) | 270 (25.3) | 99 (27.8) |
| $50K to < $100K | 376 (40.8) | 384 (40.4) | 387 (36.3) | 110 (30.9) |
| $100K or greater | 235 (25.5) | 283 (29.8) | 263 (24.7) | 64 (18.0) |
| Employment status, n (%) | ||||
| Works for pay | 461 (50.6) | 520 (54.7) | 555 (52.8) | 188 (51.4) |
| Unpaid work for family business | 6 (0.7) | 12 (1.3) | 11 (1.1) | 5 (1.4) |
| Not working in part due to health | 27 (3.0) | 27 (2.8) | 66 (6.3) | 42 (11.7) |
| Not working other reason | 418 (45.8) | 391 (41.2) | 420 (39.9) | 128 (35.6) |
| Marital status, n (%) | ||||
| Married | 610 (64.4) | 704 (71.5) | 719 (64.1) | 194 (51.6) |
| Widowed | 119 (12.6) | 83 (8.4) | 121 (10.8) | 53 (14.1) |
| Divorced/Separated | 138 (14.6) | 117 (11.9) | 186 (16.6) | 76 (20.2) |
| Never married | 81 (8.5) | 80 (8.1) | 95 (8.5) | 53 (14.1) |
| With prescription coverage, n (%) | 886 (94.1) | 935 (95.3) | 1048 (94.3) | 338 (91.1) |
| With health insurance, n (%) | 933 (98.6) | 967 (98.3) | 1099 (98.4) | 356 (94.9) |
| Health care provider, n (%) | ||||
| Private doctor | 862 (90.9) | 893 (90.8) | 986 (87.8) | 295 (78.3) |
| HMO | 32 (3.4) | 32 (3.3) | 45 (4.0) | 24 (6.4) |
| Public Clinic | 13 (1.4) | 23 (2.3) | 41 (3.7) | 24 (6.4) |
| Hospital Clinic | 15 (1.6) | 14 (1.4) | 18 (1.6) | 14 (3.7) |
| Emergency Room | 1 (0.1) | 2 (0.2) | 6 (0.5) | 4 (1.1) |
| Other | 25 (2.6) | 20 (2.0) | 27 (2.4) | 13 (3.5) |
| Comorbidity score, mean (SD) | 0.56 (1.07) | 0.50 (1.02) | 0.69 (1.25) | 0.94 (1.43) |
| CES-D, mean (SD) | 5.53 (6.28) | 6.19 (6.49) | 7.54 (7.70) | 10.53 (9.62) |
| OA medication use, past 12 months | 36 (3.6) | 119 (11.8) | 310 (27.3) | 186 (47.7) |
| KL grade class (score of both knees) | ||||
| I (0,0) | 275 (32.7) | 270 (29.9) | 223 (22.0) | 49 (14.8) |
| II (0,1) | 110 (13.1) | 93 (10.3) | 69 (6.8) | 11 (3.3) |
| III (1,1) | 72 (8.6) | 51 (6.7) | 55 (5.4) | 12 (3.6) |
| IV (0,≥2) | 95 (11.3) | 129 (14.3) | 130 (12.9) | 51 (15.4) |
| V (1,≥2) | 95 (11.3) | 122 (13.5) | 135 (13.3) | 34 (10.3) |
| VI (≥2,≥2) | 195 (23.2) | 237 (26.3) | 400 (39.5) | 174 (52.6) |
CES-D: Center for Epidemiologic Studies Depression Scale (CES-D) Score; KL: Kellgren Lawrence
Socioeconomic status and healthcare access
There was a higher proportion of African Americans in the highest pain severity group (35.9%) compared to the no pain group (14.9%). Participants in the highest knee pain severity group were characterized by lower socioeconomic status, as reflected by lower levels of education and income. Not having health insurance or not having a health insurance plan that covers prescription medicines was reported more often by those with moderate/severe pain than others. Among those with moderate level of pain or less, more than 87% reported of having a private physician. In contrast, only ~78% of those with severe pain had a private physician who they relied on for healthcare. The rest relied on a public clinic, a hospital clinic, the emergency room or others for healthcare assistance.
Association of pain with willingness to undergo TKR
The majority of OAI participants (2344/3530) were willing to undergo TKR surgery if their knee pain became severe and if recommended by a provider (Table 2). Participants with severe knee pain, compared to participants with no knee pain, were less willing to undergo TKR (crude OR 0.73, 95% CI 0.57–0.93). This association was essentially unchanged when adjusted for age and sex (Model 1; OR 0.70, 95% CI 0.55–0.90). When further adjusted for comorbidity and CES-D score, the magnitude of the association attenuated and was no longer statistically significant (Model 2; OR 0.80, 95% CI 0.61–1.03). When further adjusted for health insurance coverage, prescription medicine coverage and healthcare source, the OR minimally changed (Model 3; OR 0.83, 95% CI 0.63–1.07). The OR migrated towards the null when further adjusted for education, income, employment, race and marital status (Model 4; OR 0.92, 95% CI 0.68–1.24). In the full model, there was also no significant interaction between race and pain severity (p=0.3551; Supplement 1).
Table 2.
Willingness to consider knee replacement surgery and knee pain severity in the last 30 days
| Knee pain severity | Willing/Participants | Adjusted Model 1*
|
Adjusted Model 2†
|
Adjusted Model 3‡
|
Adjusted Model 4§
|
||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | (95% CI) | p-value | OR | (95% CI) | p-value | OR | (95% CI) | p-value | OR | (95% CI) | p-value | ||
|
|
|
|
|
||||||||||
| 0: no pain | 655/989 | 1.00 | ref | 1.00 | ref | 1.00 | ref | 1.00 | ref | ||||
| 1 – 3 | 702/1011 | 1.11 | (0.91, 1.34) | 0.300 | 1.08 | (0.88, 1.31) | 0.460 | 1.07 | (0.88, 1.30) | 0.518 | 1.01 | (0.81, 1.25) | 0.954 |
| 4 – 7 | 757/1139 | 0.97 | (0.81, 1.16) | 0.729 | 0.97 | (0.80, 1.18) | 0.779 | 0.97 | (0.80, 1.18) | 0.784 | 1.02 | (0.82, 1.26) | 0.869 |
| 8 – 10 | 230/391 | 0.70 | (0.55, 0.90) | 0.005 | 0.80 | (0.61, 1.03) | 0.086 | 0.83 | (0.63, 1.07) | 0.153 | 0.92 | (0.68, 1.24) | 0.590 |
Adjusted for age (<65 vs. >=65), sex
Adjusted for age, sex, comorbidity score, CES-D
Adjusted for age, sex, comorbidity score, CES-D, healthcare access (health insurance coverage, prescription coverage and health care provider)
Adjusted for age, sex, comorbidity score, CES-D, healthcare access and socioeconomic status (race, education, income, current employment, marital status)
A similar pattern was found upon assessing the association between WOMAC pain and willingness to undergo TKR surgery with adjustment for potential confounders (Supplement 2). Participants with high WOMAC pain, compared to those with no WOMAC pain, were less willing to undergo TKR (crude OR 0.71, 95% CI 0.58–0.86). After adjustment for demographic, clinical, healthcare access, and socioeconomic variables, the association attenuated and was no longer statistically significant (Model 4; OR 0.85, 95% CI 0.67–1.08).
Potential modifiers: socioeconomic status & healthcare access
Stratified analyses by the components of socioeconomic status, including education, income, and current employment, suggested that the most disadvantaged groups were less willing to undergo TKR with high levels of knee pain severity. However, the interaction terms between knee pain severity and each of the three socioeconomic variables in separate adjusted models were statistically not significant (Figure 1).
Figure 1.
Willingness to undergo knee replacement and knee pain severity, by socieconomic status.
P-values for interaction between: pain severity and education (0.405), pain severity and income (0.139), pain severity and employment (0.496)
*Adjusted for age(<65 vs. >=65), sex, education, income, current employment, comorbidity, CESD, marital status, health insurance, prescription coverage, health care provider
Among participants with health insurance coverage, there was no evidence of an association between knee pain severity and odds of willingness to undergo TKR after adjustment for demographic, clinical, healthcare access and socioeconomic factors (adjusted OR 1.03, 95% CI 0.73–1.38), as shown in Figure 2. Among participants without coverage, the odds of willingness to undergo TKR were significantly lower in those with severe pain compared to those with no pain after adjustment for the same variables (adjusted OR 0.08, 95% CI 0.01–0.56). A significant interaction between pain severity and health insurance coverage in the adjusted model (p=0.015) suggests that health insurance modifies the effect of knee pain severity on the odds of willingness to undergo TKR.
Figure 2.
Willingness to undergo knee replacement and knee pain severity, by healthcare access.
P-values for interaction between: pain severity and health insurance coverage (0.003), pain severity and prescription medicine coverage (0.015), pain severity and healthcare source (0.651)
*Adjusted for age(<65 vs. >=65), sex, education, income, current employment, comorbidity, CESD, marital status, health insurance, prescription coverage, health care provider
Among participants with prescription medicine coverage, the odds of willingness to undergo TKR were only slightly higher in those with severe pain compared to those with no pain after adjustment for demographic, clinical, healthcare access and socioeconomic factors (adjusted OR 1.10, 95% CI 0.79–1.53). Among participants without prescription medicine coverage, the odds of willingness were significantly lower in those with severe pain compared to those with no pain after adjustment for the same variables (adjusted OR 0.12, 95% CI 0.03–0.41). Again, a significant interaction between prescription medicine coverage and knee pain severity (p=0.003) suggests that prescription medicine coverage modifies the effect of knee pain severity on the odds of willingness to undergo TKR.
Participants who usually rely on a private/HMO physician for healthcare did not differ with respect to knee pain severity in terms of their odds of willingness to undergo TKR (adjusted OR 0.98, 95% CI 0.71–1.36). Among those who usually rely on a public clinic, hospital clinic or the emergency room for care, the odds of willingness were substantially but not significantly lower among those with severe pain compared to those with no pain (adjusted OR 0.42, 95% CI 0.11–1.44). While the difference in the magnitude of effect sizes is suggestive, no significant difference was found based on the test for interaction (p=0.651).
Similar patterns were observed when assessing the role of each socioeconomic and healthcare access variable in potentially modifying the association between WOMAC pain severity and willingness to undergo TKR (Supplement 3, Supplement 4).
Sensitivity analyses of the impact of missing covariate data using multiple imputation with chained equations revealed no substantial differences in effect size or statistical significance (data not shown).
DISCUSSION
In this large group of participants with or at high risk of developing knee OA, we found no evidence of an effect of knee pain severity on willingness to undergo TKR surgery after adjustment for participant clinical information, healthcare access coverage and socioeconomic status. Yet, our study is the first to show that this association between pain and patient treatment preference for joint replacement surgery could be modified by healthcare access and not by patient socioeconomic status. Severe knee pain was associated with decreased willingness to undergo TKR surgery among those without health insurance coverage but not among those with health insurance coverage. In parallel, severe knee pain was associated with decreased willingness to undergo TKR surgery among those without prescription medicine coverage but not among those with prescription medicine coverage.
Patient treatment preference is partly an attitudinal disposition (30). Attitudes arise spontaneously and inevitably as we form beliefs about an object (31). Patient preference towards joint replacement may be guided by better understanding and expectations regarding TKR, having lower levels of religiosity, having a surgical discussion with a physician, and having higher trust in physicians and the healthcare system (13). Patient preferences have been reported to influence the utilization of various medical procedures and treatments (32). They, in fact, have the most significant influence on OA patients receiving a recommendation for joint replacement from an orthopedic surgeon (20). In turn, receiving this recommendation leads to subsequent receipt of joint replacement surgery (20). Based on a prospective study of OA patients in Canada, willingness to consider joint replacement is the strongest determinant of time to first total joint arthroplasty (21).
Finding that knee pain severity was not independently associated with willingness to undergo joint replacement in OAI participants, specifically those with health insurance, is supported by previous research, though contrary to intuition. Based on qualitative studies, pain and disability are highly influential factors in the decision to undergo knee arthroplasty, yet decisions regarding the need for TKR are not explained by patient symptoms alone (33, 34). Among patients from Ontario, Canada with hip/knee arthritis, those who were willing to consider joint replacement had higher WOMAC summary scores than others (18). However, WOMAC summary score was not found to be an independent correlate of willingness to have arthroplasty (8, 18). In a study of knee OA patients from Houston, Texas, those who would consider having TKR had slightly higher WOMAC pain scores than those who would not consider having the procedure (9). Yet, WOMAC pain was not significantly associated with considering TKR surgery in both unadjusted and adjusted models. Similarly, in a sample of symptomatic knee/hip OA patients from Johnston County, North Carolina, neither WOMAC pain nor WOMAC function score was significantly associated with willingness to undergo TJR in all adjusted models (35).
Participants with no health insurance coverage for medical and surgical treatments, however, had significantly lower odds of willingness to undergo TKR at higher levels of knee pain severity than those with health insurance coverage. Unfortunately, this finding is not completely surprising. Persons who lack insurance tend to receive less medical care, including screening and treatment (36, 37). Uninsured adults are far less likely to receive medical or surgical care even when they develop new symptoms that may represent serious medical conditions or that have major adverse effects on quality of life (16, 37). They have fewer physician visits and are less likely to be hospitalized than the insured (36). In the US, having no insurance among those ≤ 64 years of age was independently associated with low TKR utilization (38). Medicare eligibility and additional insurance increased the likelihood of having joint replacement surgery (19). Among Medicare recipients, having supplemental health insurance increased the chance of having TKR (38).
The most common reason cited by the uninsured for not receiving necessary care is inability to pay (37). Using data collected on 10,155 unilateral TKR procedures from sixty-one hospitals in the US in 2008, the estimated average knee implant cost per case ranged from $1,797 to $12,093 (39). Besides implant cost, hospitalization fee, surgeon’s fee, post-surgical rehabilitation, and potential for lost wages must also be taken into account. Even for those with Medicare coverage, out-of-pocket expenses could reach several thousand dollars (38). Among the uninsured, favoring TKR surgery likely also implies agreeing to incur all direct and indirect costs associated with the procedure. Indeed, conjoint analysis studies found that treatment costs highly influence patient preferences for knee OA therapies (40, 41).
While the impact of socioeconomic status on the association between knee pain severity and willingness was suggestive and consistent with prior research, evidence to support an impact of access to healthcare was notably stronger. African-American race had been consistently linked to decreased willingness to undergo TKR surgery (12–14). Having at least a high school education had been associated with higher likelihood of undergoing TKR (21, 38). Income was also independently associated with TKR utilization in the US, but not in Canada, where universal health insurance exists (21, 38). Using data from the United States, receipt of TKR was higher among those currently employed relative to the unemployed (38). Yet, paid employment had no independent effect on time to first TKR in Canada (21). Neither income nor education had been independently associated with willingness to consider arthroplasty in various studies (8, 9, 13, 35, 42).
There are limitations to this study. First, we evaluated willingness to consider TKR using a standardized survey rather than through conversations with an orthopedic surgeon. Patient preference towards TKR may change during a comprehensive consultation with an orthopedic surgeon. Second, a relatively small proportion of OAI participants were uninsured and without prescription coverage. While confidence intervals were wide in some stratified analyses (Figure 2), we were able to detect statistically significant interactions between pain severity and both health insurance coverage and prescription coverage, suggesting that both change the effect of knee pain severity on willingness. A larger population-based study may be necessary to more accurately estimate the effect of healthcare access on the relationship between pain severity and willingness to undergo TKR. Third, we do not have information on the adequacy of coverage for TKR surgery. As already mentioned, out-of-pockets costs may be substantial, even among those with Medicare coverage. Our analysis may underestimate the adverse effects of being uninsured. Fourth, unmeasured confounders such as patient knowledge and attitudes about joint replacement surgery could influence patient treatment preferences for joint replacement and affect the observed relationships that we uncovered (13). Income, a covariate in our analysis, could have also potentially changed between ascertainment at baseline and assessment of willingness at 72 months. OAI data, however, does not contain such information.
Among participants without health insurance coverage, those reporting severe pain were less willing to undergo TKR than those reporting no pain, while no evidence for an effect of knee pain on willingness was observed among the insured. This raises concern that TKR is not being considered by patients at a time when it might result in greatest benefit. Advocacy for health policies that improve access to quality medical care should be encouraged. Public support for these policies can motivate local and national leaders to support legislations that may improve patient preference for surgical procedures such as TKR and increase the use of joint replacement surgery among appropriate candidates.
Supplementary Material
SIGNIFICANCE AND INNOVATION.
Patient preferences significantly influence actual receipt of joint replacement surgery, a highly effective treatment for advanced knee osteoarthritis (OA).
This is the largest study (n=3530) in the United States that examined how socioeconomic status and healthcare coverage could affect the relationship between knee pain severity and patient willingness to undergo knee replacement surgery.
It was found that severe knee pain was associated with decreased preference for knee replacement surgery among those without health insurance but not among those with health insurance.
Our findings suggest that policy changes that improve healthcare access could potentially increase preference for and receipt of knee replacement surgery of patients with severe knee OA.
Acknowledgments
The OAI is a public-private partnership comprised of five contracts (N01-AR-2-2258; N01-AR-2-2259; N01-AR-2-2260; N01-AR-2-2261; N01-AR-2-2262) funded by the National Institutes of Health, a branch of the Department of Health and Human Services, and conducted by the OAI Study Investigators. Private funding partners include Merck Research Laboratories; Novartis Pharmaceuticals Corporation, GlaxoSmithKline; and Pfizer, Inc. Private sector funding for the OAI is managed by the Foundation for the National Institutes of Health. This manuscript was prepared using an OAI public use data set and does not necessarily reflect the opinions or views of the OAI investigators, the NIH, or the private funding partners.
Footnotes
None of the authors declare any potential conflicts of interest in regard to this manuscript.
References
- 1.Fitzgerald JD, Orav EJ, Lee TH, Marcantonio ER, Poss R, Goldman L, et al. Patient quality of life during the 12 months following joint replacement surgery. Arthritis Rheum. 2004;51(1):100–9. doi: 10.1002/art.20090. [DOI] [PubMed] [Google Scholar]
- 2.Meding JB, Meding LK, Ritter MA, Keating EM. Pain relief and functional improvement remain 20 years after knee arthroplasty. Clin Orthop Relat Res. 2012;470(1):144–9. doi: 10.1007/s11999-011-2123-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Vina ER, Hannon MJ, Kwoh CK. Improvement following total knee replacement surgery: Exploring preoperative symptoms and change in preoperative symptoms. Semin Arthritis Rheum. 2015 doi: 10.1016/j.semarthrit.2015.10.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Zhang W, Moskowitz RW, Nuki G, Abramson S, Altman RD, Arden N, et al. OARSI recommendations for the management of hip and knee osteoarthritis, Part II: OARSI evidence-based, expert consensus guidelines. Osteoarthritis Cartilage. 2008;16(2):137–62. doi: 10.1016/j.joca.2007.12.013. [DOI] [PubMed] [Google Scholar]
- 5.Skou ST, Roos EM, Laursen MB, Rathleff MS, Arendt-Nielsen L, Simonsen O, et al. A Randomized, Controlled Trial of Total Knee Replacement. N Engl J Med. 2015;373(17):1597–606. doi: 10.1056/NEJMoa1505467. [DOI] [PubMed] [Google Scholar]
- 6.Liang MH, Cullen KE, Larson MG, Thompson MS, Schwartz JA, Fossel AH, et al. Cost-effectiveness of total joint arthroplasty in osteoarthritis. Arthritis Rheum. 1986;29(8):937–43. doi: 10.1002/art.1780290801. [DOI] [PubMed] [Google Scholar]
- 7.Llewellyn-Thomas HA, Arshinoff R, Bell M, Williams JI, Naylor CD. In the queue for total joint replacement: patients’ perspectives on waiting times. Ontario Hip and Knee Replacement Project Team. J Eval Clin Pract. 1998;4(1):63–74. doi: 10.1046/j.1365-2753.1998.t01-1-00006.x. [DOI] [PubMed] [Google Scholar]
- 8.Hawker GA, Wright JG, Coyte PC, Williams JI, Harvey B, Glazier R, et al. Determining the need for hip and knee arthroplasty: the role of clinical severity and patients’ preferences. Med Care. 2001;39(3):206–16. doi: 10.1097/00005650-200103000-00002. [DOI] [PubMed] [Google Scholar]
- 9.Suarez-Almazor ME, Souchek J, Kelly PA, O’Malley K, Byrne M, Richardson M, et al. Ethnic variation in knee replacement: patient preferences or uninformed disparity? Arch Intern Med. 2005;165(10):1117–24. doi: 10.1001/archinte.165.10.1117. [DOI] [PubMed] [Google Scholar]
- 10.Clark JP, Hudak PL, Hawker GA, Coyte PC, Mahomed NN, Kreder HJ, et al. The moving target: a qualitative study of elderly patients’ decision-making regarding total joint replacement surgery. J Bone Joint Surg Am. 2004;86-A(7):1366–74. [PubMed] [Google Scholar]
- 11.Bewtra M, Kilambi V, Fairchild AO, Siegel CA, Lewis JD, Johnson FR. Patient preferences for surgical versus medical therapy for ulcerative colitis. Inflamm Bowel Dis. 2014;20(1):103–14. doi: 10.1097/01.MIB.0000437498.14804.50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.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(9):2429–35. doi: 10.1002/art.10494. [DOI] [PubMed] [Google Scholar]
- 13.Kwoh CK, Vina ER, Cloonan YK, Hannon MJ, Boudreau RM, Ibrahim SA. Determinants of patient preferences for total knee replacement: African-Americans and whites. Arthritis Res Ther. 2015;17:348. doi: 10.1186/s13075-015-0864-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Vina ER, Cloonan YK, Ibrahim SA, Hannon MJ, Boudreau RM, Kwoh CK. Race, sex, and total knee replacement consideration: role of social support. Arthritis Care Res (Hoboken) 2013;65(7):1103–11. doi: 10.1002/acr.21925. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Hawker GA, Wright JG, Coyte PC, Williams JI, Harvey B, Glazier R, et al. Differences between men and women in the rate of use of hip and knee arthroplasty. N Engl J Med. 2000;342(14):1016–22. doi: 10.1056/NEJM200004063421405. [DOI] [PubMed] [Google Scholar]
- 16.Mota RE, Tarricone R, Ciani O, Bridges JF, Drummond M. Determinants of demand for total hip and knee arthroplasty: a systematic literature review. BMC Health Serv Res. 2012;12:225. doi: 10.1186/1472-6963-12-225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Karlson EW, Daltroy LH, Liang MH, Eaton HE, Katz JN. Gender differences in patient preferences may underlie differential utilization of elective surgery. Am J Med. 1997;102(6):524–30. doi: 10.1016/s0002-9343(97)00050-8. [DOI] [PubMed] [Google Scholar]
- 18.Hawker GA, Wright JG, Badley EM, Coyte PC. Perceptions of, and willingness to consider, total joint arthroplasty in a population-based cohort of individuals with disabling hip and knee arthritis. Arthritis Rheum. 2004;51(4):635–41. doi: 10.1002/art.20524. [DOI] [PubMed] [Google Scholar]
- 19.Card D, Dobkin C, Maestas N. The Impact of Nearly Universal Insurance Coverage on Health Care Utilization: Evidence from Medicare. Am Econ Rev. 2008;98(5):2242–58. doi: 10.1257/aer.98.5.2242. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Hausmann LR, Mor M, Hanusa BH, Zickmund S, Cohen PZ, Grant R, et al. The effect of patient race on total joint replacement recommendations and utilization in the orthopedic setting. J Gen Intern Med. 2010;25(9):982–8. doi: 10.1007/s11606-010-1399-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Hawker GA, Guan J, Croxford R, Coyte PC, Glazier RH, Harvey BJ, et al. A prospective population-based study of the predictors of undergoing total joint arthroplasty. Arthritis Rheum. 2006;54(10):3212–20. doi: 10.1002/art.22146. [DOI] [PubMed] [Google Scholar]
- 22.Downie WW, Leatham PA, Rhind VM, Wright V, Branco JA, Anderson JA. Studies with pain rating scales. Ann Rheum Dis. 1978;37(4):378–81. doi: 10.1136/ard.37.4.378. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ferraz MB, Quaresma MR, Aquino LR, Atra E, Tugwell P, Goldsmith CH. Reliability of pain scales in the assessment of literate and illiterate patients with rheumatoid arthritis. J Rheumatol. 1990;17(8):1022–4. [PubMed] [Google Scholar]
- 24.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] [Google Scholar]
- 25.Katz JN, Chang LC, Sangha O, Fossel AH, Bates DW. Can comorbidity be measured by questionnaire rather than medical record review? Med Care. 1996;34(1):73–84. doi: 10.1097/00005650-199601000-00006. [DOI] [PubMed] [Google Scholar]
- 26.Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas. 1977;1(3):385–401. [Google Scholar]
- 27.Kellgren JH, Lawrence JS. Radiological assessment of osteo-arthrosis. Ann Rheum Dis. 1957;16(4):494–502. doi: 10.1136/ard.16.4.494. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Maldonado G, Greenland S. Simulation study of confounder-selection strategies. Am J Epidemiol. 1993;138(11):923–36. doi: 10.1093/oxfordjournals.aje.a116813. [DOI] [PubMed] [Google Scholar]
- 29.White IR, Royston P, Wood AM. Multiple imputation using chained equations: Issues and guidance for practice. Stat Med. 2011;30(4):377–99. doi: 10.1002/sim.4067. [DOI] [PubMed] [Google Scholar]
- 30.Ibrahim SA. Racial variations in the utilization of knee and hip joint replacement: an introduction and review of the most recent literature. Curr Orthop Pract. 2010;21(2):126–31. doi: 10.1097/BCO.0b013e3181d08223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Ajzen I, Fishbein M. Attitudes and the attitude-behavior relation: reasoned and automatic processes. In: Stroebe W, Hewstone M, editors. European review of social psychology. Chichester, England: Wiley; 2000. [Google Scholar]
- 32.Smedley B, Stith A, Nelson A Committee on Understanding & Eliminating Racial and Ethnic Disparities in Health Care BoHSP, Institute of Medicine. Unequal treatment: confronting racial and ethnic disparities in health care. Washington, DC: National Academic Press; 2003. [PubMed] [Google Scholar]
- 33.McHugh GA, Luker KA. Influences on individuals with osteoarthritis in deciding to undergo a hip or knee joint replacement: a qualitative study. Disabil Rehabil. 2009;31(15):1257–66. doi: 10.1080/09638280802535129. [DOI] [PubMed] [Google Scholar]
- 34.Toye FM, Barlow J, Wright C, Lamb SE. Personal meanings in the construction of need for total knee replacement surgery. Soc Sci Med. 2006;63(1):43–53. doi: 10.1016/j.socscimed.2005.11.054. [DOI] [PubMed] [Google Scholar]
- 35.Allen KD, Golightly YM, Callahan LF, Helmick CG, Ibrahim SA, Kwoh CK, et al. Race and sex differences in willingness to undergo total joint replacement: the Johnston County Osteoarthritis Project. Arthritis Care Res (Hoboken) 2014;66(8):1193–202. doi: 10.1002/acr.22295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Hafner-Eaton C. Physician utilization disparities between the uninsured and insured. Comparisons of the chronically ill, acutely ill, and well nonelderly populations. JAMA. 1993;269(6):787–92. [PubMed] [Google Scholar]
- 37.Baker DW, Shapiro MF, Schur CL. Health insurance and access to care for symptomatic conditions. Arch Intern Med. 2000;160(9):1269–74. doi: 10.1001/archinte.160.9.1269. [DOI] [PubMed] [Google Scholar]
- 38.Hanchate AD, Zhang Y, Felson DT, Ash AS. Exploring the determinants of racial and ethnic disparities in total knee arthroplasty: health insurance, income, and assets. Med Care. 2008;46(5):481–8. doi: 10.1097/MLR.0b013e3181621e9c. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Robinson JC, Pozen A, Tseng S, Bozic KJ. Variability in costs associated with total hip and knee replacement implants. J Bone Joint Surg Am. 2012;94(18):1693–8. doi: 10.2106/JBJS.K.00355. [DOI] [PubMed] [Google Scholar]
- 40.Fraenkel L, Suter L, Cunningham CE, Hawker G. Understanding preferences for disease-modifying drugs in osteoarthritis. Arthritis Care Res (Hoboken) 2014;66(8):1186–92. doi: 10.1002/acr.22280. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Posnett J, Dixit S, Oppenheimer B, Kili S, Mehin N. Patient preference and willingness to pay for knee osteoarthritis treatments. Patient Prefer Adherence. 2015;9:733–44. doi: 10.2147/PPA.S84251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Hawker GA, Wright JG, Glazier RH, Coyte PC, Harvey B, Williams JI, et al. The effect of education and income on need and willingness to undergo total joint arthroplasty. Arthritis Rheum. 2002;46(12):3331–9. doi: 10.1002/art.10682. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.