Abstract
Background
There is inequity in total knee arthroplasty (TKA) use between racial groups. The social vulnerability index (SVI) may characterize mechanisms underlying disparity, which is essential for identifying targeted interventions. We examined the relationship of the overall SVI and its themes with TKA use rates for Black and White Medicare beneficiaries and the difference in TKA utilization between these groups.
Methods
We used 2013-2017 Medicare data to calculate annual TKA use rates per 1000 beneficiaries in each hospital referral region (HRR). Multivariable mixed-effects linear regressions estimated the relationship between social vulnerability and TKA use.
Results
The mean (standard deviation) TKA rates were 9.4 (1.7) and 5.9 (2.2) per 1000 White and Black beneficiaries, respectively. HRRs with highest social vulnerability (quartile 4) had lower TKA rates for White (estimate for quartile 4: −1.7, 95% confidence interval [CI]: −2.2 to −1.1, P < .001) and Black beneficiaries (quartile 4: −1.7, 95% CI: −2.3 to −1.0, P < .001) compared to HRRs with the least vulnerability. The SVI’s minority status and language theme was linked with wider rate differences between White and Black beneficiaries (quartile 4: 0.6, 95% CI: 0.0-1.2, P = .046).
Conclusions
Worse overall SVI was related to lower TKA use rates for both White and Black Medicare beneficiaries. Areas with the highest minority status and language vulnerability had disproportionately lower TKA rates for Black compared to White beneficiaries. Our findings suggest regional mechanisms that health-care professionals and policymakers will need to address to ensure equitable TKA access.
Keywords: Total knee arthroplasty, Health-care disparities, Social vulnerability
Introduction
Total knee arthroplasties (TKAs) are important surgeries for patients with end-stage knee osteoarthritis and are associated with improved quality of life [1]. For over 30 years, there has been clear evidence demonstrating racial differences in utilization rates of TKAs such that Black patients and those belonging to other racial and ethnic minority groups less frequently receive TKAs compared to White patients [2,3]. Furthermore, these differences persist after accounting for clinical severity, medical comorbidities, and social factors such as socioeconomic status (SES), thus highlighting a racial disparity in the receipt of orthopaedic care [4,5]. Nationwide analyses of Medicare beneficiaries investigating these disparities have consistently shown geographic variation, with specific regions having worse differences in utilization rates compared to others [6,7]. Characterizing mechanisms that make one area more susceptible to disparate TKA utilization compared to another may help identify targets for intervention.
Markers of relative geographic vulnerability, including the social vulnerability index (SVI), social disadvantage index, and area deprivation index (ADI), have been increasingly utilized across fields to understand differences in health-care utilization and outcomes that may be impacted by the social circumstances within which one lives [[8], [9], [10], [11]]. Among these, the SVI has recently been used to further understand differences in total joint arthroplasty utilization and outcomes, in part because it defines 4 themes that provide further details regarding the social determinants of health [[12], [13], [14], [15], [16]]. These themes include SES (theme 1), household composition and disability (HCD) (theme 2), minority status and language (MSL) (theme 3), and housing type and transportation (HTT) (theme 4). When examining the role of these themes in the use of total hip arthroplasties, our team found that the MSL theme was linked with the worsening of utilization differences between White and Black beneficiaries in a given HRR [16]. Determining whether specific SVI themes contribute to differences in TKA utilization may provide insight into underlying factors that influence surgical rates.
Our study estimated the relationship of the (i) overall SVI score and (ii) SVI theme scores with (a) TKA use rates for White Medicare beneficiaries, (b) TKA use rates for Black beneficiaries, and (c) the utilization rate difference between these groups, which we defined as the disparity gap in a particular area. This work is important for key stakeholders including health-care providers and policymakers working to address these issues and improve accessibility.
Material and methods
Dataset aggregation and study cohort
The study was exempted from review by our institution’s Research Subject Review Board. Data were obtained from Medicare enrollment and claims from January 1, 2013, to December 31, 2017. Age- and sex-standardized rates of inpatient TKAs per 1000 individuals were calculated annually for non-Hispanic White and non-Hispanic Black beneficiaries in 306 hospital referral regions (HRRs), which is described further in prior work [7]. HRRs were used as the geographic unit of analysis for their utility in understanding variation across tertiary health-care markets [17,18]. The dataset was limited to the end of 2017 because TKAs are increasingly performed on an outpatient basis since being removed from the Centers for Medicare & Medicaid Services’ Inpatient-Only list in January 2018 [19]. We excluded HRRs with less than 1% Black beneficiaries to address potentially high variability in TKA use in these regions. The final study cohort included 241 unique HRRs.
The census-tract level SVI measures were obtained from the Centers for Disease Control & Prevention and Agency for Toxic Substances and Disease Registry [8]. Further details regarding SVI methodology are in Appendix Exhibit 1. Census tracts often cross HRR borders, so weighted averages of SVI scores at the HRR level were calculated using crosswalks from the master area block level equivalency datasets, which are available from the Missouri Census Data Center [20]. Covariates were obtained from the Dartmouth Atlas of Healthcare and the Agency for Healthcare Research and Quality’s Social Determinants of Health database. All variables at the zip code level were averaged to the HRR level using crosswalks provided by the Dartmouth Atlas of Health [21,22]. Further covariate specifications are described in Appendix Exhibit 1.
Outcomes
Annual inpatient TKA use rates per 1000 White and Black beneficiaries, separately, were the first 2 key outcomes. We also calculated the difference in TKA utilization between these groups to quantify the disparity in use of TKAs. Positive values represent the extent to which White beneficiaries use TKAs more frequently compared to Black beneficiaries.
Key explanatory variables
The overall SVI score for each HRR, defined as the percentile rank (from 0 to 100) on the sum of the 4 theme scores, was the key explanatory variable for the first set of models. The percentile scores of the SVI themes were then modeled as explanatory variables in subsequent models. Higher scores indicate worse vulnerability. These SVI variables were specified as quartiles.
Descriptive statistics
For continuous variables, we report means with standard deviations (SDs) and test for differences in their distribution across SVI quartiles using Kruskal–Wallis tests. For categorical variables, we report frequencies with percentages and use chi-square tests to determine any statistically significant differences in how HRR characteristics are distributed across the 4 quartiles of SVI.
Multivariable analysis
Multivariable mixed-effects linear regressions estimated the role of the overall SVI and SVI themes in the variation in TKA use rates, separately for White and Black Medicare beneficiaries, and for rate differences between the 2 groups. All models accounted for potential confounders (such as health-care or geographic factors) and the clustering of observations at the level of the HRR. Relative rurality, encapsulated by an area’s Rural–Urban Commuting Area code, and total population per 10,000 were also accounted for in all models.
Sensitivity analysis
The following multivariable linear regressions were created to assess the robustness of the main analysis: (i) re-estimating models using a continuous specification (rather than a quartile specification) of the overall SVI score, (ii) re-estimating models using a continuous specification of SVI theme scores, and (iii) creating models specifying individual SVI variables, which are used to create the SVI themes, as key explanatory variables.
Results
HRR characteristics
In the baseline year of 2013, the mean overall SVI score in the 237 HRRs was 51.9 (SD: 10.6) (Table 1). The average theme scores were 52.6 (SD: 11.6) for theme 1 (SES vulnerability), 55.1 (SD: 12.9) for theme 2 (HCD), 46.8 (SD: 17.1) for theme 3 (MSL), and 50.8 (SD: 8.4) for theme 4 (HTT). The average TKA utilization rate for White and Black Medicare beneficiaries were 9.4 (SD: 1.7) and 5.9 (SD: 2.2) per 1000 beneficiaries, respectively, with a utilization difference of 3.5 (SD: 2.0) per 1000 beneficiaries. Between 2013 and 2017, there were 130,973,173 Medicare beneficiary-year observations. Among these, there were 1,110,518 beneficiaries who received a total of 1,189,389 TKAs.
Table 1.
Descriptive statistics for the 237 HRRs included in the baseline year 2013, separated by overall SVI quartiles.
| Variable | Quartile 1 (least vulnerable) (19.8-44.2) | Quartile 2 (44.4-51.3) | Quartile 3 (51.4-58.9) | Quartile 4 (most vulnerable) (59.0-85.1) | Total (19.8-85.1) | P valuea |
|---|---|---|---|---|---|---|
| N | 60 | 59 | 59 | 59 | 237 | |
| SVI: Mean (SD) | ||||||
| Overall | 38.7 (5.2) | 48.1 (1.8) | 55.2 (2.5) | 65.7 (5.1) | 51.9 (10.6) | <.001 |
| Theme 1–SES | 39.0 (7.6) | 49.5 (5.9) | 57.3 (5.0) | 65.1 (6.8) | 52.6 (11.6) | <.001 |
| Theme 2–HCD | 43.2 (9.6) | 52.2 (10.4) | 59.3 (8.4) | 66.0 (10.7) | 55.1 (12.9) | <.001 |
| Theme 3–MSL | 43.4 (13.9) | 42.8 (17.4) | 45.1 (16.4) | 55.9 (17.4) | 46.8 (17.1) | <.001 |
| Theme 4–HTT | 42.8 (6.8) | 49.2 (6.2) | 52.1 (4.5) | 59.5 (6.0) | 50.8 (8.4) | <.001 |
| TKA Rates per 1000: Mean (SD) | ||||||
| Non-Hispanic White | 9.7 (1.7) | 9.5 (1.7) | 9.6 (1.4) | 8.9 (1.8) | 9.4 (1.7) | .049 |
| Non-Hispanic Black | 6.4 (2.7) | 5.8 (1.7) | 6.1 (2.4) | 5.3 (1.7) | 5.9 (2.2) | .06 |
| White–Black Difference | 3.4 (2.2) | 3.7 (1.6) | 3.5 (2.5) | 3.5 (1.6) | 3.5 (2.0) | .96 |
| Health-care factors: Mean (SD) | ||||||
| Primary care physicians (per 100,000) | 77.3 (10.3) | 73.7 (11.5) | 66.4 (9.4) | 65.1 (12.0) | 70.7 (11.9) | <.001 |
| Orthopaedic surgeons (per 100,000) | 6.7 (0.8) | 6.4 (1.0) | 6.0 (1.1) | 5.6 (1.1) | 6.2 (1.1) | <.001 |
| Acute care hospital beds (per 1000) | 1.9 (0.4) | 2.1 (0.4) | 2.3 (0.5) | 2.6 (0.6) | 2.2 (0.5) | <.001 |
| Geographic factors: Mean (SD) | ||||||
| RUCA Code | 2.5 (1.5) | 3.1 (1.7) | 3.1 (1.4) | 3.7 (1.7) | 3.1 (1.6) | .002 |
| Total population (per 10,000) | 216.2 (187.2) | 210.9 (192.6) | 177.4 (202.0) | 166.1 (259.7) | 192.7 (212.0) | .001 |
| Location: N (%) | <.001 | |||||
| Northeast | 19 (31.7) | 10 (16.9) | 1 (1.7) | 3 (5.1) | 33 (13.9) | |
| South | 12 (20.0) | 20 (33.9) | 40 (67.8) | 43 (72.9) | 115 (48.5) | |
| Midwest | 22 (36.7) | 21 (35.6) | 11 (18.6) | 3 (5.1) | 57 (24.1) | |
| West | 7 (11.7) | 8 (13.6) | 7 (11.9) | 10 (16.9) | 32 (13.5) |
N, Number of Hospital Referral Regions; RUCA, Rural–Urban Commuting Area.
P values for Kruskal–Wallis and chi-square tests for continuous and categorical variables, respectively, comparing the distribution of Hospital Referral Region characteristics across quartiles of the overall SVI score in 2013.
Multivariable linear regressions
Overall SVI
The HRRs with higher overall vulnerability had lower TKA utilization among both White and Black Medicare beneficiaries. Rates of TKAs among White beneficiaries were 0.7 per 1000 lower for quartile 2 (95% confidence interval [CI]: −1.1 to −0.3, P < .001), 1.0 per 1000 lower for quartile 3 (95% CI: −1.5 to −0.5, P < .001), and 1.7 lower for quartile 4 (95% CI: −2.2 to −1.1, P < .001) compared to the least vulnerable HRRs (Table 2). Among Black beneficiaries, TKA rates were 0.7 per 1000 lower (95% CI: −1.2 to −0.2, P = .005), 0.9 per 1000 lower (95% CI: −1.4 to −0.3, P = .003), and 1.7 per 1000 lower (95% CI: −2.3 to −1.0, P < .001) in HRRs within SVI quartiles 2 through 4, respectively. No significant relationship between the overall SVI and the difference in TKA rates between the 2 groups was found.
Table 2.
Multivariable linear regression models estimating the association of the overall SVI with TKA utilization per 1000 Medicare beneficiaries.
| Variable | Non-Hispanic White beneficiaries |
Non-Hispanic black beneficiaries |
White–Black rate difference |
|---|---|---|---|
| Estimate [95% CI] | Estimate [95% CI] | Estimate [95% CI] | |
| SVI Quartile | |||
| Q1 (Least Vulnerable) | Ref | Ref | Ref |
| Q2 | −0.7c [−1.1 to −0.3] | −0.7b [−1.2 to −0.2] | 0.0 [−0.4 to 0.5] |
| Q3 | −1.0c [−1.5 to −0.5] | −0.9b [−1.4 to −0.3] | −0.2 [−0.7 to 0.4] |
| Q4 (Most Vulnerable) | −1.7c [−2.2 to −1.1] | −1.7c [−2.3 to −1.0] | 0.0 [−0.6 to 0.6] |
| Health-care factors | |||
| Primary care physicians (per 100,000) | −0.0c [−0.0 to −0.0] | −0.0c [−0.0 to −0.0] | −0.0 [−0.0 to 0.0] |
| Orthopaedic surgeons (per 100,000) | 0.1 [−0.0 to 0.3] | 0.1 [−0.1 to 0.3] | 0.0 [−0.2 to 0.2] |
| Acute care hospital beds (per 1000) | 0.3 [−0.1 to 0.6] | 0.1 [−0.4 to 0.5] | 0.2 [−0.2 to 0.6] |
| Geographic factors | |||
| RUCA Code | 0.1a [0.0-0.2] | 0.1 [−0.0 to 0.2] | −0.0 [−0.1 to 0.1] |
| Total Population (per 10,000) | −0.0b [−0.0 to −0.0] | −0.0b [−0.0 to −0.0] | 0.0 [−0.0 to 0.0] |
| Location | |||
| Northeast | Ref | Ref | Ref |
| South | 1.2c [0.7-1.7] | 0.8b [0.2-1.4] | 0.5 [−0.1 to 1.0] |
| Midwest | 1.9c [1.4-2.4] | 1.1c [0.5-1.7] | 0.8b [0.3-1.4] |
| West | 0.8a [0.2-1.3] | 1.1b [0.4-1.8] | −0.3 [−1.0 to 0.3] |
| Year | |||
| 2013 | Ref | Ref | Ref |
| 2014 | −0.3c [−0.4 to −0.2] | 0.1 [−0.2 to 0.4] | −0.4a [−0.7 to −0.1] |
| 2015 | −0.2b [−0.3 to −0.1] | 0.2 [−0.1 to 0.5] | −0.4a [−0.7 to −0.1] |
| 2016 | 0.2c [0.1-0.4] | 0.5b [0.2-0.8] | −0.2 [−0.6 to 0.1] |
| 2017 | 0.4c [0.3-0.5] | 0.2 [−0.1 to 0.5] | 0.2 [−0.1 to 0.5] |
| N | 1194 | 1194 | 1194 |
Q, Quartile; N, Number of Hospital Referral Regions; RUCA, Rural–Urban Commuting Area; Ref, Reference.
P < .05.
P < .01.
P < .001.
SVI themes analysis
When compared to HRRs with the least SES vulnerability (quartile 1), those with higher vulnerability had lower TKA utilization among White beneficiaries (Estimate for quartile 2: −0.7, 95% CI: −1.1 to −0.2, P = .003; quartile 3: −0.9, 95% CI: −1.4 to −0.3, P = .001; quartile 4: −1.8, 95% CI: −2.5 to −1.2, P < .001), and Black beneficiaries (quartile 3: −0.7, 95% CI: −1.4 to −0.0, P = .04; quartile 4: −1.5, 95% CI: −2.3 to −0.7, P < .001) (Table 3). The SVI theme for HCD vulnerability was linked with higher TKA rates for White (quartile 2: 0.5, 95% CI: 0.1-1.0, P = .02; quartile 3: 1.1, 95% CI: 0.6-1.7, P < .001; quartile 4: 1.2, 95% CI: 0.5-2.0, P < .001) and Black (quartile 3: 0.8, 95% CI: 0.1-1.5, P = .02) beneficiaries. Notably, high vulnerability in the MSL theme was linked with higher rates for White beneficiaries (quartile 2: 0.5, 95% CI: 0.1-0.9, P = .02; quartile 3: 0.6, 95% CI: 0.1-1.0, P = .009) and lower rates for Black beneficiaries (quartile 4: −0.9, 95% CI: −1.6 to −0.4, P = .002). Greater HTT theme vulnerability was related to lower TKA rates for both White (quartile 2: −0.4, 95% CI: −0.8 to −0.0, P = .049; quartile 3: −0.8, 95% CI: −1.2 to −0.4, P < .001; quartile 4: −0.9, 95% CI: −1.4 to −0.5, P < .001) and Black (quartile 4: −0.9, 95% CI: −1.5 to −0.3, P = .003) beneficiaries.
Table 3.
Multivariable linear regression models estimating the association of the 4 SVI themes with TKA utilization per 1000 Medicare beneficiaries.
| Variable | Non-Hispanic White beneficiaries |
Non-Hispanic black beneficiaries |
White–Black difference |
|---|---|---|---|
| Estimate [95% CI] | Estimate [95% CI] | Estimate [95% CI] | |
| SVI Themes | |||
| Theme 1 (SES) | |||
| Q1 (Least Vulnerable) | Ref | Ref | Ref |
| Q2 | −0.7b [−1.1 to −0.2] | −0.4 [−0.9 to 0.2] | −0.3 [−0.8 to 0.3] |
| Q3 | −0.9b [−1.4 to −0.3] | −0.7a [−1.4 to −0.0] | −0.1 [−0.8 to 0.5] |
| Q4 (Most Vulnerable) | −1.8c [−2.5 to −1.2] | −1.5c [−2.3 to −0.7] | −0.3 [−1.1 to 0.4] |
| Theme 2 (HCD) | |||
| Q1 (Least Vulnerable) | Ref | Ref | Ref |
| Q2 | 0.5a [0.1-1.0] | 0.2 [−0.4 to 0.7] | 0.4 [−0.2 to 1.0] |
| Q3 | 1.1c [0.6-1.7] | 0.8a [0.1-1.5] | 0.3 [−0.4 to 1.0] |
| Q4 (Most Vulnerable) | 1.2c [0.5-2.0] | 0.7 [−0.2 to 1.6] | 0.5 [−0.3 to 1.4] |
| Theme 3 (MSL) | |||
| Q1 (Least Vulnerable) | Ref | Ref | Ref |
| Q2 | 0.5a [0.1-0.9] | −0.1 [−0.6 to 0.4] | 0.5a [0.1 to 1.0] |
| Q3 | 0.6b [0.1-1.0] | −0.1 [−0.7 to 0.4] | 0.8b [0.2 to 1.3] |
| Q4 (Most Vulnerable) | −0.4 [−0.9 to 0.1] | −1.0b [−1.6 to −0.4] | 0.6a [0.0 to 1.2] |
| Theme 4 (HTT) | |||
| Q1 (Least Vulnerable) | Ref | Ref | Ref |
| Q2 | −0.4a [−0.8 to −0.0] | 0.1 [−0.4 to 0.5] | −0.4 [−0.9 to 0.0] |
| Q3 | −0.8c [−1.2 to −0.4] | −0.2 [−0.7 to 0.3] | −0.6a [−1.1 to −0.1] |
| Q4 (Most Vulnerable) | −0.9c [−1.4 to −0.5] | −0.9b [−1.5 to −0.3] | −0.0 [−0.6 to 0.5] |
| Health-care factors | |||
| Primary care physicians (per 100,000) | −0.0b [−0.0 to −0.0] | −0.0a [−0.0 to −0.0] | 0.0 [−0.0 to 0.0] |
| Orthopaedic surgeons (per 100,000) | −0.0 [−0.2 to 0.1] | −0.0 [−0.2 to 0.2] | −0.0 [−0.2 to 0.2] |
| Acute care hospital beds (per 1000) | 0.1 [−0.2 to 0.5] | 0.1 [−0.4 to 0.5] | 0.0 [−0.4 to 0.5] |
| Geographic factors | |||
| RUCA Code | 0.1 [−0.0 to 0.2] | 0.0 [−0.1 to 0.1] | 0.1 [−0.1 to 0.2] |
| Total population (per 10,000) | −0.0 [−0.0 to 0.0] | −0.0a [−0.0 to −0.0] | 0.0 [−0.0 to 0.0] |
| Location | |||
| Northeast | Ref | Ref | Ref |
| South | 1.2c [0.7-1.6] | 0.9b [0.3-1.4] | 0.3 [−0.2 to 0.8] |
| Midwest | 1.8c [1.4-2.3] | 1.0c [0.4-1.5] | 0.8b [0.3-1.4] |
| West | 1.0c [0.5-1.6] | 1.7c [1.0-2.4] | −0.6 [−1.3 to 0.0] |
| Year | |||
| 2013 | Ref | Ref | Ref |
| 2014 | −0.3c [−0.4 to −0.2] | 0.1 [−0.2 to 0.4] | −0.4a [−0.7 to −0.1] |
| 2015 | −0.2b [−0.3 to −0.1] | 0.2 [−0.1 to 0.5] | −0.4a [−0.7 to −0.1] |
| 2016 | 0.2c [0.1-0.4] | 0.5b [0.1-0.8] | −0.2 [−0.6 to 0.1] |
| 2017 | 0.4c [0.3-0.5] | 0.2 [−0.1 to 0.5] | 0.2 [−0.1 to 0.5] |
| N | 1194 | 1194 | 1194 |
Q, Quartile; N, Number of Hospital Referral Regions; RUCA, Rural–Urban Commuting Area; Ref, Reference.
P < .05.
P < .01.
P < .001.
With regard to the difference in TKA utilization between the 2 groups, high MSL vulnerability was linked to worsening of the use difference (quartile 2: 0.5, 95% CI: 0.1-1.0, P = .02; quartile 3: 0.8, 95% CI: 0.2-1.3, P = .004; quartile 4: 0.6, 95% CI: 0.0-1.2, P = .046). Therefore, the most vulnerable HRRs with 100,000 Medicare beneficiaries will have a TKA utilization gap widened by 60 surgeries compared to HRRs with the least vulnerability in this theme. Additionally, high vulnerability in the HTT theme was linked with a lower utilization difference (quartile 3: −0.6, 95% CI: −1.1 to −0.1, P = .02).
Sensitivity analysis
The results of the sensitivity analysis overwhelmingly support the findings of our main analysis. In models specifying overall SVI as a continuous variable, higher overall vulnerability reduced TKA use for both White and Black beneficiaries but there was no statistically significant relationship with TKA use differences (Appendix Exhibit 2). In models specifying SVI themes as continuous variables, the MSL vulnerability increased the White–Black disparity in utilization rates (Appendix Exhibit 3). In the models specifying all SVI variables as explanatory variables, minority status and residence in mobile homes increased disparity in utilization rates (Appendix Exhibit 4).
Discussion
Racial and ethnic disparities in the utilization of TKAs remain a critical issue that demands our attention to effectively address. In this study, we used the SVI to further characterize HRR-level geographic variation in TKA use rates among White and Black Medicare beneficiaries as well as the difference in utilization between these groups. Higher vulnerability in the overall SVI lowered TKA use for both groups of beneficiaries, but was not linked with the disparity in utilization. In models using SVI themes, high vulnerability in SES and HTT (themes 1 and 4) lowered TKA use rates, while HCD (theme 2) was linked with higher TKA use for both groups. Most notably, high vulnerability in MSL (theme 3) increased TKA use for White beneficiaries and lowered TKA use for Black beneficiaries, which further widened the gap in TKA use rates. Because Black beneficiaries have lower rates of utilization to begin with, this highlights a factor that further exacerbates this difference.
Our study supports the growing body of evidence demonstrating the impact of one’s lived environment on limiting access to these important surgeries [23,24]. Our work fills an important gap by demonstrating that higher vulnerability in the overall SVI lowered utilization rates across racial groups. Interestingly, it has also been shown that higher ADI scores are significantly associated with lower TKA procedural volume [25]. Regional differences in social disadvantage are therefore associated with both patient-level and provider-level decreases in TKA volume. Because both the SVI and ADI encompass several variables to create a single score, it is difficult to ascertain the specific mechanisms driving these interactions. It is probable that such metrics include multiple contributing variables, such as lower accessibility to care or poorer health literacy, highlighting the multidimensionality of this issue. Despite this, such metrics are useful to identify areas that are likely to have lower TKA utilization.
The use of SVI themes in our analysis provides insight into possible mechanisms that impact rates of TKAs. Both SES vulnerability and HTT vulnerability lowered TKA use rates for Black and White beneficiaries, a finding we also noted with total hip arthroplasty utilization [16]. Although patients receiving TKAs over the last decade are significantly older and have more comorbidity burden, Black patients and those with lower SES continue to make up a minority of those receiving TKAs [26]. Interestingly, patients who are White and have higher SES are more likely to travel further to receive a TKA [27]. It follows that transportation barriers, which are also impacted by one’s SES, may be important considerations. The association of HTT vulnerability with lower use rates further supports this notion. Finally, HCD vulnerability increased utilization for both Black and White beneficiaries. Arthritis is the most common cause of disability in the United States [28], so HRRs with higher disability secondary to knee osteoarthritis may consequently receive more TKAs, but more work is required to clarify this.
The association of high vulnerability in MSL with exacerbation of the White-Black difference in TKA utilization is consistent with our prior work on racial disparity in total hip arthroplasty utilization [16]. A plausible explanation for this is differential access to high-volume surgical centers and arthroplasty specialists within these regions. There is abundant literature demonstrating that Black patients are more likely to receive a TKA from a lower volume center compared to White patients [[29], [30], [31]], which is also concerning given that use of low-volume hospitals has been associated with worse outcomes [32]. Policies to improve accessibility to high-volume hospitals for vulnerable groups may be an appropriate intervention, especially since travel burden seemingly does not increase for most patients [33], but such initiatives must be carefully considered to prevent inadvertently limiting access to care [34,35]. Future studies examining within HRR differences in accessibility to high-volume hospitals and rates of TKAs may be helpful to further explore this possible mechanism. Addressing possible language barriers is also necessary to ensure adequate counseling on these surgeries. Limited English proficiency has been suggested to significantly restrict postoperative care utilization following arthroplasty [[36], [37], [38]], so providing effective communication with the assistance of interpreters may help with this challenge.
There are several possible limitations of this work to consider. The geographic unit of analysis, the HRR, represents larger tertiary markets of health-care delivery and cannot account for contributing factors at the level of patient and physician. Furthermore, HRRs are relatively broad in scope, and policy initiatives may require smaller areas for effective change to be implemented. Despite this, our study highlights geographic differences in surgical use that may partially be explained by underlying vulnerabilities in these regions. Additionally, this study represents Medicare beneficiaries and includes inpatient TKAs only, so our findings have limited generalizability to those covered by other insurance types or those undergoing outpatient arthroplasties. Finally, because this study was limited to years prior to the removal of TKAs off of the Centers for Medicare & Medicaid Services’ Inpatient-Only list in 2018, our findings may not reflect the current state of disparities that persist in the receipt of orthopaedic care. Nevertheless, our work is significant because it provides the foundation to measure the association of the SVI with TKA disparities and how this relationship may be driven by specific SVI themes.
The implications of this work suggest that there are measurable vulnerabilities that offer an explanation to the geographic variation in utilization of TKAs, and specifically to the variation in the disparity in the receipt of these surgeries for Black Medicare beneficiaries compared to their White counterparts. The MSL SVI theme was implicated as a factor exacerbating the disparity in utilization, which may represent differential access to high-volume hospitals in these regions which disproportionately affects Black beneficiaries. Policymakers working to improve quality and equity in arthroplasty care should consider such mechanisms and geographic differences to improve accessibility. It remains important for physicians caring for patients with knee osteoarthritis to be mindful of possible barriers that limit access to these surgeries and, when possible, work to address them.
Conclusions
The SVI is a marker of relative geographic disadvantage that provides insight into the variation in TKA utilization. In our study of Medicare beneficiaries, higher overall SVI was linked with lower use rates for non-Hispanic White and Black beneficiaries while high MSL vulnerability was linked with worsening of the difference in utilization between these groups. These findings suggest that there are regional differences, captured in part by the SVI, that should be elucidated further to promote equity in care for patients with knee osteoarthritis.
Funding
The National Institute on Minority Health and Health Disparities provided support for this work [grant number R01MD018573].
CRediT authorship contribution statement
Jordan J. Cruse: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Derek T. Schloemann: Writing – review & editing, Validation, Methodology, Investigation, Conceptualization. Elizabeth C. Danielson: Writing – review & editing, Validation, Methodology, Investigation, Conceptualization. Benjamin F. Ricciardi: Writing – review & editing, Validation, Investigation, Conceptualization. Rishi Balkissoon: Writing – review & editing, Validation, Investigation, Conceptualization. Patricia D. Franklin: Writing – review & editing, Validation, Investigation, Conceptualization. Caroline P. Thirukumaran: Writing – review & editing, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Data curation, Conceptualization.
Conflicts of interest
Derek T. Schloemann reports research support from the American Foundation for Surgery of the Hand and the University of Rochester.
Benjamin F. Ricciardi reports editorial or governing board positions with CORR, The Knee, Arthroplasty Today, and HSS Journal. He has appointments in the Practice Management Committee (AAOS) and the Knee Program Committee (AAOS).
Rishi Balkissoon reports consulting payments from the Zimmer-Biomet C3 Advisory Council, Solenic Medical, Inc., and RevelAI (Formerly AzraCare); stock or stock options in Solenic Medical, Inc., and RevelAI (Formerly AzraCare). He serves as an ad hoc reviewer for JBJS, Arthroplasty Today, and CORR. He has appointments in JRGOS and Nth Dimensions.
Patricia D. Franklin reports research support from AAHKS and previous research support from Depuy. She is a board member on the Medical Science Advisory Board for the National Arthritis Foundation.
Caroline P. Thirukumaran reports grant funding from the National Institutes of Health; speaker honoraria from the National Institutes of Health, Brown University, Boston University, and Brigham and Women’s Hospital; reviewer honoraria from the National Institutes of Health and the Veterans Affairs. She serves as the Disparities Section Editor for Current Osteoporosis Reports and Anesthesia and Analgesia, and the Research Chair for the Strategies in Clinical Research Section of the Orthopedic Research Society.
The other authors declare no potential conflicts of interest.
For full disclosure statements refer to https://doi.org/10.1016/j.artd.2025.101801.
Footnotes
This article is part of a supplement entitled ‘Achieving the Vision — Striving for Health Equity in Arthroplasty’ supported by Zimmer Biomet.
Appendix Exhibit 1 Details of SVI Methodology and Covariate Specifications
SVI Methodology
The SVI aggregates data from the American Community Survey over the previous 5 years and includes 15 variables which measure social risk factors. These variables are then categorized into the 4 SVI themes. Theme 1 (SES) includes the percentage of people living below the poverty line, the percentage of individuals who are unemployed, average income, and percentage of people without a high school diploma. Theme 2 (HCD) includes the percentage of people aged 65 or older, the percentage of people aged 17 or younger, the percentage of civilians with a disability, and the number of single-parent households. Theme 3 (MSL) includes minority status and percentage of people who speak English “less than well.” Theme 4 (HTT) includes the number of multiunit structures, mobile homes, crowded living conditions, number of people without a vehicle, and the number of people living in group quarters.
We specified the following variables, obtained from the SVI dataset, in our analyses: RPL_THEME1, RPL_THEME2, RPL_THEME3, and RPL_THEME4. These scores represent percentile rankings on each theme. The individual SVI theme scores are calculated by finding the percentile rank on each variable listed above that comprise of the theme. The sum of these percentiles is then obtained, and a new percentile ranking is calculated from this sum, which is defined as the SVI theme score. For the overall SVI, the RPL_THEMES variable was utilized, which represents the percentile score on the sum of the 4 theme scores.
Our study analyzes trends using Medicare data between 2013 and 2017. Because SVI data are released every 2 years, we populated the 2013 SVI data with the 2014 SVI values, 2015 SVI data with 2016 SVI values, and 2017 SVI data with 2018 SVI values.
Covariates
All multivariable linear regression models controlled for possible HRR-level confounding variables. The Dartmouth Atlas of Health was utilized to obtain the number of primary care physicians per 100,000 residents, the number of orthopaedic surgeons per 100,000 residents, and the number of acute care hospital beds per 1000 residents in each HRR to control for variability in the supply health-care providers and facilities. A geographic area’s relative rurality, measured by the Rural–Urban Commuting Area codes, and the total population were both obtained from Agency for Healthcare Research and Quality Social Determinants of Health database. The year and the United States Census geographic region (Northeast, Midwest, South, and West) were specified as categorical indicators to control for variability due to possible larger regional differences and trends in utilization over time.
Appendix Exhibit 2.
Multivariable mixed-effects linear regression models using a continuous specification of the overall SVI to estimate its association with TKA utilization per 1000 Medicare beneficiaries.
| Variable | Non-Hispanic White beneficiaries |
Non-Hispanic black beneficiaries |
White–Black difference |
|---|---|---|---|
| Estimate [95% CI] | Estimate [95% CI] | Estimate [95% CI] | |
| SVI Total | −0.0c [−0.1 to −0.0] | −0.1c [−0.1 to −0.0] | 0.0 [−0.0 to 0.0] |
| Health-care factors | |||
| Primary care physicians (per 100,000) | −0.0c [−0.0 to −0.0] | −0.0c [−0.1 to −0.0] | 0.0 [−0.0 to 0.0] |
| Orthopaedic surgeons (per 100,000) | 0.1 [−0.0 to 0.3] | 0.1 [−0.1 to 0.3] | 0.0 [−0.1 to 0.2] |
| Acute care hospital beds (per 1000) | 0.2 [−0.2 to 0.6] | 0.1 [−0.3 to 0.6] | 0.1 [−0.3 to 0.6] |
| Geographic factors | |||
| RUCA Code | 0.1a [0.0 to 0.2] | 0.1 [−0.0 to 0.2] | −0.0 [−0.1 to 0.1] |
| Total population (per 10,000) | −0.0b [−0.0 to −0.0] | −0.0b [−0.0 to −0.0] | 0.0 [−0.0 to 0.0] |
| Location | |||
| Northeast | Ref | Ref | Ref |
| South | 1.1c [0.6 to 1.5] | 0.7a [0.1 to 1.3] | 0.4 [−0.1 to 0.9] |
| Midwest | 1.8c [1.3 to 2.3] | 0.9b [0.3 to 1.5] | 0.8b [0.3 to 1.4] |
| West | 0.6a [0.0 to 1.2] | 1.2b [0.4 to 1.9] | −0.4 [−1.1 to 0.2] |
| Year | |||
| 2013 | Ref | Ref | Ref |
| 2014 | −0.3c [−0.4 to −0.2] | 0.1 [−0.2 to 0.4] | −0.4a [−0.7 to −0.1] |
| 2015 | −0.2b [−0.3 to −0.1] | 0.2 [−0.1 to 0.5] | −0.4a [−0.7 to −0.1] |
| 2016 | 0.2c [0.1 to 0.4] | 0.5b [0.2 to 0.8] | −0.2 [−0.6 to 0.1] |
| 2017 | 0.4c [0.3 to 0.5] | 0.2 [−0.1 to 0.6] | 0.2 [−0.1 to 0.5] |
| N | 1194 | 1194 | 1194 |
N, Number of Hospital Referral Regions; RUCA, Rural–Urban Commuting Area; Ref, Reference.
P < .05.
P < .01.
P < .001.
Appendix Exhibit 3.
Multivariable mixed-effects linear regression models using continuous specifications of the SVI themes to estimate their associations with TKA utilization per 1000 Medicare beneficiaries.
| Variable | Non-Hispanic White beneficiaries |
Non-Hispanic black beneficiaries |
White–Black difference |
|---|---|---|---|
| Estimate [95% CI] | Estimate [95% CI] | Estimate [95% CI] | |
| SVI Themes | |||
| Theme 1–SES | −0.0 [−0.0 to 0.0] | −0.1c [−0.1 to −0.0] | 0.0 [−0.0 to 0.0] |
| Theme 2–HCD | 0.0 [−0.0 to 0.0] | 0.0b [0.0 to 0.1] | 0.0 [−0.0 to 0.0] |
| Theme 3–MSL | −0.0 [−0.0 to 0.0] | −0.0b [−0.0 to −0.0] | 0.0a [0.0 to 0.0] |
| Theme 4–HTT | −0.1c [−0.1 to −0.0] | −0.0 [−0.1 to 0.0] | −0.0 [−0.0 to 0.0] |
| Health-care factors | |||
| Primary care physicians (per 100,000) | −0.0a [−0.0 to −0.0] | −0.0b [−0.0 to −0.0] | 0.0 [−0.0 to 0.0] |
| Orthopaedic surgeons (per 100,000) | 0.1 [−0.0 to 0.2] | 0.0 [−0.1 to 0.2] | 0.0 [−0.2 to 0.2] |
| Acute care hospital beds (per 1000) | 0.0 [−0.3 to 0.4] | −0.1 [−0.6 to 0.3] | 0.2 [−0.3 to 0.6] |
| Geographic factors | |||
| RUCA Code | 0.1 [−0.0 to 0.2] | −0.0 [−0.1 to 0.1] | 0.1 [−0.1 to 0.2] |
| Total population (per 10,000) | −0.0 [−0.0 to 0.0] | −0.0 [−0.0 to 0.0] | 0.0 [−0.0 to 0.0] |
| Location | |||
| Northeast | 0.0 [0.0-0.0] | 0.0 [0.0-0.0] | 0.0 [0.0-0.0] |
| South | 1.0c [0.5-1.4] | 0.7a [0.1-1.2] | 0.4 [−0.2 to 0.9] |
| Midwest | 1.6c [1.1-2.1] | 0.8b [0.2-1.4] | 0.8b [0.2-1.3] |
| West | 0.7a [0.1-1.3] | 1.4c [0.7-2.1] | −0.6 [−1.2 to 0.1] |
| Year | |||
| 2013 | 0.0 [0.0 to 0.0] | 0.0 [0.0 to 0.0] | 0.0 [0.0 to 0.0] |
| 2014 | −0.3c [−0.4 to −0.2] | 0.1 [−0.2 to 0.4] | −0.4a [−0.7 to −0.1] |
| 2015 | −0.2c [−0.4 to −0.1] | 0.2 [−0.1 to 0.5] | −0.4a [−0.7 to −0.1] |
| 2016 | 0.2c [0.1-0.3] | 0.5b [0.1-0.8] | −0.2 [−0.6 to 0.1] |
| 2017 | 0.4c [0.3-0.5] | 0.2 [−0.1 to 0.6] | 0.2 [−0.2 to 0.5] |
| N | 1194 | 1194 | 1194 |
N, Number of Hospital Referral Regions; RUCA, Rural–Urban Commuting Area; Ref, Reference.
P < .05.
P < .01.
P < .001.
Appendix Exhibit 4.
Estimates from multivariable mixed-effects linear regression models using continuous specifications of individual SVI variables to examine their association with TKA utilization per 1000 Medicare beneficiaries.
| Variable | Non-Hispanic White beneficiaries |
Non-Hispanic black beneficiaries |
White–Black difference |
|---|---|---|---|
| Estimate [95% CI] | Estimate [95% CI] | Estimate [95% CI] | |
| SVI Variables | |||
| Theme 1 (SES) | |||
| Below poverty | −1.1 [−5.0 to 2.7] | −5.2 [−11.0 to 0.6] | 2.1 [−3.4 to 7.6] |
| Unemployed | −0.5 [−1.9 to 0.8] | −4.1c [−6.3 to −1.8] | 1.6 [−0.6 to 3.8] |
| Per capital income | 1.6 [−2.0 to 5.3] | 9.3c [4.1 to 14.4] | −5.9a [−10.8 to −1.0] |
| No high school diploma | −3.5b [−6.1 to −0.9] | −6.3c [−9.8 to −2.7] | 2.3 [−1.1 to 5.6] |
| Theme 2 (HCD) | |||
| Age 65 and older | −1.3 [−4.0 to 1.3] | 0.5 [−3.2 to 4.1] | −0.5 [−4.0 to 2.9] |
| Age 17 and younger | 1.5 [−1.3 to 4.4] | 2.5 [−1.5 to 6.4] | 0.3 [−3.5 to 4.0] |
| Civilians with a disability | 1.3 [−0.8 to 3.4] | 2.5 [−0.5 to 5.4] | −0.3 [−3.1 to 2.5] |
| Single parent households with children <18 y old | 1.6 [−1.2 to 4.3] | −1.7 [−6.0 to 2.6] | 3.8 [−0.4 to 7.9] |
| Theme 3 (MSL) | |||
| Minority status | −0.1 [−1.9 to 1.6] | −1.9 [−4.1 to 0.4] | 2.7a [0.5 to 4.8] |
| Speaks English less than well | 0.8 [−1.1 to 2.7] | 1.8 [−0.9 to 4.4] | −1.2 [−3.7 to 1.4] |
| Theme 4 (HTT) | |||
| 10+ units/structure | −2.6a [−5.2 to −0.0] | −3.2 [−6.8 to 0.5] | 0.6 [−2.8 to 4.1] |
| Mobile homes | −0.1 [−1.8 to 1.6] | −2.2 [−4.5 to 0.1] | 2.3a [0.2 to 4.5] |
| More people than rooms | −3.6c [−5.6 to −1.5] | −1.9 [−5.2 to 1.3] | −2.0 [−5.2 to 1.1] |
| No vehicle | −3.1a [−6.0 to −0.3] | 3.4 [−0.7 to 7.5] | −5.8b [−9.7 to −1.9] |
| Group quarters | 1.4 [−0.6 to 3.5] | 0.8 [−1.9 to 3.5] | 0.7 [−1.9 to 3.2] |
| Health-care factors | |||
| Primary care physicians (per 100,000) | −0.0 [−0.0 to 0.0] | −0.0a [−0.0 to −0.0] | 0.0 [−0.0 to 0.0] |
| Orthopaedic surgeons (per 100,000) | 0.1 [−0.1 to 0.2] | 0.2a [0.0 to 0.3] | −0.1 [−0.3 to 0.0] |
| Acute care hospital beds (per 1000) | 0.2 [−0.2 to 0.7] | −0.2 [−0.7 to 0.4] | 0.3 [−0.2 to 0.8] |
| Geographic factors | |||
| RUCA Code | −0.0 [−0.2 to 0.1] | −0.1 [−0.2 to 0.1] | −0.0 [−0.2 to 0.1] |
| Total population (per 10,000) | −0.0 [−0.0 to 0.0] | −0.0 [−0.0 to 0.0] | 0.0 [−0.0 to 0.0] |
| Location | |||
| Northeast | Ref | Ref | Ref |
| South | 0.8a [0.0 to 1.5] | 1.6b [0.6 to 2.6] | −0.8 [−1.8 to 0.1] |
| Midwest | 1.3c [0.8 to 1.9] | 1.0b [0.3 to 1.7] | 0.3 [−0.4 to 1.0] |
| West | 0.5 [−0.3 to 1.2] | 2.4c [1.4 to 3.4] | −1.7c [−2.7 to −0.7] |
| Year | |||
| 2013 | Ref | Ref | Ref |
| 2014 | −0.3c [−0.4 to −0.2] | 0.1 [−0.2 to 0.4] | −0.4a [−0.7 to −0.1] |
| 2015 | −0.2a [−0.3 to −0.0] | 0.2 [−0.1 to 0.6] | −0.4a [−0.7 to −0.1] |
| 2016 | 0.3c [0.1 to 0.4] | 0.5b [0.2 to 0.8] | −0.2 [−0.5 to 0.1] |
| 2017 | 0.4c [0.3 to 0.6] | 0.2 [−0.1 to 0.6] | 0.2 [−0.1 to 0.5] |
| N | 1194 | 1194 | 1194 |
N, Number of Hospital Referral Regions; RUCA, Rural–Urban Commuting Area; Ref, Reference.
P < .05.
P < .01.
P < .001.
Appendix A. Supplementary data
Appendix
.
References
- 1.March L.M., Cross M.J., Lapsley H., Brnabic A.J., Tribe K.L., Bachmeier C.J., et al. Outcomes after hip or knee replacement surgery for osteoarthritis. A prospective cohort study comparing patients' quality of life before and after surgery with age-related population norms. Med J Aust. 1999;171:235–238. [PubMed] [Google Scholar]
- 2.Wilson M.G., May D.S., Kelly J.J. Racial differences in the use of total knee arthroplasty for osteoarthritis among older Americans. Ethn Dis. 1994;4:57–67. [PubMed] [Google Scholar]
- 3.Thirukumaran C.P., Cruse J.J., Franklin P.D., Ricciardi B.F., Suleiman L.I., Ibrahim S.A. Two decades since the unequal treatment report: the state of racial, ethnic, and socioeconomic disparities in elective total hip and knee replacement use. J Bone Joint Surg Am. 2024;107:523–538. doi: 10.2106/JBJS.24.00347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Usiskin I., Misra D. Racial disparities in elective total joint arthroplasty for osteoarthritis. ACR Open Rheumatol. 2022;4:306–311. doi: 10.1002/acr2.11399. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Skinner J., Zhou W., Weinstein J. The influence of income and race on total knee arthroplasty in the United States. J Bone Joint Surg Am. 2006;88:2159–2166. doi: 10.2106/JBJS.E.00271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Skinner J., Weinstein J.N., Sporer S.M., Wennberg J.E. Racial, ethnic, and geographic disparities in rates of knee arthroplasty among medicare patients. N Engl J Med. 2003;349:1350–1359. doi: 10.1056/NEJMsa021569. [DOI] [PubMed] [Google Scholar]
- 7.Thirukumaran C.P., Cai X., Glance L.G., Kim Y., Ricciardi B.F., Fiscella K.A., et al. Geographic variation and disparities in total joint replacement use for medicare beneficiaries: 2009 to 2017. J Bone Joint Surg Am. 2020;102:2120–2128. doi: 10.2106/JBJS.20.00246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cdc/atsdr social vulnerability index. Centers for disease control and Prevention/ agency for toxic substances and disease Registry/ geospatial research, analysis, and services program. https://www.atsdr.cdc.gov/place-health/php/svi/index.html
- 9.Social deprivation index (sdi) Robert graham center - Policy studies in family medicine & primary care. https://www.graham-center.org/maps-data-tools/social-deprivation-index.html
- 10.Health UoWSoMaP 2021 area deprivation index. https://www.neighborhoodatlas.medicine.wisc.edu/
- 11.Kind A.J.H., Buckingham W.R. Making neighborhood-disadvantage metrics accessible - the neighborhood atlas. N Engl J Med. 2018;378:2456–2458. doi: 10.1056/NEJMp1802313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Dubin J.A., Bains S.S., Hameed D., Monarrez R., Moore M.C., Mont M.A., et al. The utility of the social vulnerability index as a proxy for social disparities following total knee arthroplasty. J Arthroplasty. 2024;39:S33–S38. doi: 10.1016/j.arth.2024.01.049. [DOI] [PubMed] [Google Scholar]
- 13.Yi S.H., Calanan R.M., Reid M.J.A., Kazakova S.V., Baggs J., McLees A.W. Community-level social vulnerability and hip and knee joint replacement surgery receipt among medicare enrollees with arthritis. Med Care. 2024;62:830–839. doi: 10.1097/MLR.0000000000002068. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Danielson E.C., Li W., Suleiman L., Franklin P.D. Social risk and patient-reported outcomes after total knee replacement: implications for medicare policy. Health Serv Res. 2024;59 doi: 10.1111/1475-6773.14215. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Rahman T.M., Shaw J.H., Mehaidli A., Hennekes M., Hansen L., Castle J.P., et al. The impact of social determinants of health on outcomes and complications after total knee arthroplasty: an analysis of neighborhood deprivation indices. J Bone Joint Surg Am. 2024;106:288–303. doi: 10.2106/JBJS.23.00044. [DOI] [PubMed] [Google Scholar]
- 16.Cruse J.J., Schloemann D.T., Danielson E.C., Ricciardi B.F., Franklin P.D., Balkissoon R., et al. Association of geographic social vulnerability with race-based differences in the utilization of total hip arthroplasty surgeries for medicare beneficiaries. J Arthroplasty. 2024;40:1427–1432.e7. doi: 10.1016/j.arth.2024.11.053. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Wennberg J.E., Cooper M.M., editors. The quality of medical care in the United States: a report on the medicare program: the dartmouth atlas of health care 1999. The center for the evaluative clinical sciences; Chicago (IL): 1999. [PubMed] [Google Scholar]
- 18.Wennberg J.E., Cooper M.M. The dartmouth atlas of health care in the United States. The center for the evaluative clinical sciences; Chicago (IL): 1996. [PubMed] [Google Scholar]
- 19.Schloemann D.T., Sajda T., Ricciardi B.F., Thirukumaran C.P. Association of total knee replacement removal from the inpatient-only list with outpatient surgery utilization and outcomes in medicare patients. JAMA Netw Open. 2023;6 doi: 10.1001/jamanetworkopen.2023.16769. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Missouri Census Data Center . Geographic correspondence engine (Geocorr) The University of Missouri Center for Health Policy; Columbia, MO: 2014. [Google Scholar]
- 21.The dartmouth atlas of health care. Dartmouth atlas project. https://www.dartmouthatlas.org/
- 22.Social determinants of health database. Rockville, MD: Agency for Health-care Research and Quality. https://www.ahrq.gov/sdoh/data-analytics/sdoh-data.html
- 23.Karimi A., Burkhart R.J., Hecht C.J., 2nd, Acuna A.J., Kamath A.F. Is social deprivation associated with usage, adverse events, and patient-reported outcome measures in total joint arthroplasty? A systematic review. Clin Orthop Relat Res. 2023;481:239–250. doi: 10.1097/CORR.0000000000002394. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hartnett D.A., Lama C.J., Brodeur P.G., Cruz A.I., Jr., Gil J.A., Cohen E.M. Socioeconomic disparities in the utilization of total knee arthroplasty. J Arthroplasty. 2022;37:1973–1979.e1. doi: 10.1016/j.arth.2022.04.033. [DOI] [PubMed] [Google Scholar]
- 25.Wu V.S., Acuna A.J., Kim A.G., Burkhart R.J., Kamath A.F. Impact of social disadvantage among total knee arthroplasty places of service on procedural volume: a nationwide medicare analysis. Arch Orthop Trauma Surg. 2023;143:4579–4585. doi: 10.1007/s00402-022-04708-7. [DOI] [PubMed] [Google Scholar]
- 26.Ashkenazi I., Lawrence K.W., Kaplan M., Arshi A., Rozell J.C., Schwarzkopf R., et al. Demographic and socioeconomic trends of patients undergoing total knee arthroplasty from 2013 to 2022-an analysis from an urban orthopaedic hospital. J Arthroplasty. 2024;39:2158–2165. doi: 10.1016/j.arth.2024.04.029. [DOI] [PubMed] [Google Scholar]
- 27.Orringer M., Roberts H., Ngan A., Ward D. Influence of demographic and socioeconomic factors on hospital distance for total knee arthroplasty. Arthroplast Today. 2023;21 doi: 10.1016/j.artd.2023.101131. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Centers for Disease C. Prevention. Prevalence and most common causes of disability among adults--united states, 2005. MMWR Morb Mortal Wkly Rep. 2009;58:421–426. [PubMed] [Google Scholar]
- 29.Liu J.H., Zingmond D.S., McGory M.L., SooHoo N.F., Ettner S.L., Brook R.H., et al. Disparities in the utilization of high-volume hospitals for complex surgery. JAMA. 2006;296:1973–1980. doi: 10.1001/jama.296.16.1973. [DOI] [PubMed] [Google Scholar]
- 30.Zhang W., Lyman S., Boutin-Foster C., Parks M.L., Pan T.J., Lan A., et al. Racial and ethnic disparities in utilization rate, hospital volume, and perioperative outcomes after total knee arthroplasty. J Bone Joint Surg Am. 2016;98:1243–1252. doi: 10.2106/JBJS.15.01009. [DOI] [PubMed] [Google Scholar]
- 31.SooHoo N.F., Zingmond D.S., Ko C.Y. Disparities in the utilization of high-volume hospitals for total knee replacement. J Natl Med Assoc. 2008;100:559–564. doi: 10.1016/s0027-9684(15)31303-1. [DOI] [PubMed] [Google Scholar]
- 32.Losina E., Wright E.A., Kessler C.L., Barrett J.A., Fossel A.H., Creel A.H., et al. Neighborhoods matter: use of hospitals with worse outcomes following total knee replacement by patients from vulnerable populations. Arch Intern Med. 2007;167:182–187. doi: 10.1001/archinte.167.2.182. [DOI] [PubMed] [Google Scholar]
- 33.Ramirez G., Myers T.G., Thirukumaran C.P., Ricciardi B.F. Does hypothetical centralization of revision tha and tka exacerbate existing geographic or demographic disparities in access to care by increased patient travel distances or times? A large-database study. Clin Orthop Relat Res. 2022;480:1033–1045. doi: 10.1097/CORR.0000000000002072. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Dy C.J., Marx R.G., Ghomrawi H.M., Pan T.J., Westrich G.H., Lyman S. The potential influence of regionalization strategies on delivery of care for elective total joint arthroplasty. J Arthroplasty. 2015;30:1–6. doi: 10.1016/j.arth.2014.08.017. [DOI] [PubMed] [Google Scholar]
- 35.FitzGerald J.D., Soohoo N.F., Losina E., Katz J.N. Potential impact on patient residence to hospital travel distance and access to care under a policy of preferential referral to high-volume knee replacement hospitals. Arthritis Care Res (Hoboken) 2012;64:890–897. doi: 10.1002/acr.21611. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Joo H., Nguyen K., Kolodzie K., Chen L.L., Kim M.O., Manuel S. Differences in acute postoperative opioid use by English proficiency, race, and ethnicity after total knee and hip arthroplasty. Anesth Analg. 2025;140:155–164. doi: 10.1213/ANE.0000000000007068. [DOI] [PubMed] [Google Scholar]
- 37.Nguyen K.H., Rambachan A., Ward D.T., Manuel S.P. Language barriers and postoperative opioid prescription use after total knee arthroplasty. Explor Res Clin Soc Pharm. 2022;7 doi: 10.1016/j.rcsop.2022.100171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Nguyen K.H., Suarez P., Sales C., Fernandez A., Ward D.T., Manuel S.P. Patients who have limited English proficiency have decreased utilization of revision surgeries after hip and knee arthroplasty. J Arthroplasty. 2023;38:1429–1433. doi: 10.1016/j.arth.2023.02.024. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.