Geographic Variability in Medicare Hospital Reimbursement for Primary Total Hip and Knee Arthroplasty Episodes of Care From 2013 to 2023

Abstract

Background

Medicare is the largest payer of total hip arthroplasty (THA) and total knee arthroplasty (TKA) in the United States, yet reimbursement has shifted substantially in the past decade. We described and compared state- and region-level trends in Medicare inpatient reimbursement for primary THA/TKA from 2013 to 2023.

Methods

This study retrospectively analyzed the Medicare Inpatient Hospitals by Provider and Service database for diagnostic-related group 470 (major hip/knee replacement without major complication or comorbidity) from 2013 to 2023. Extracted fields were hospital-submitted charges and Medicare facility payments, inflation adjusted to 2023 US dollars. State means were discharge-volume–weighted to reflect the average beneficiary; states were aggregated to US Census regions (Northeast, Midwest, South, and West). Regional differences were compared, and longitudinal trends were estimated using linear mixed-effects models, including year, region, and their interaction.

Results

In total, 3,724,353 primary THA/TKA discharges were billed to Medicare. Annual inpatient volume fell 84.1% (451,603 to 71,939). Inflation-adjusted reimbursement per discharge declined from $15,808 to $13,696 (−$2113; −13.4%), whereas charges rose from $71,469 to $85,675 (+$14,206; +19.9%). No state experienced an inflation-adjusted increase. Regional declines ranged from −19.1% (Midwest) to −10.1% (West). Mixed-effects modeling showed an overall decline of approximately −$285 per discharge per year (P < .001); the South declined less steeply than the Midwest (+$36/year relative slope difference, P < .05).

Conclusions

Inflation-adjusted Medicare inpatient reimbursement for THA/TKA decreased nationwide with consistent but heterogeneous declines across states and regions. The steepest reductions occurred in the Midwest and the smallest in the West, underscoring persistent geographic variability with implications for access to arthroplasty care and sustainable payment models.

Introduction

Total hip arthroplasty (THA) and total knee arthroplasty (TKA) represent some of the most common and costly procedures in the United States [[1], [2], [3]]. These procedures account for the largest procedural expenditure among Medicare beneficiaries [4], and Medicare serves as the primary payer for total joint arthroplasty (TJA) nationwide [5]. As the population ages and demand for arthroplasty rises, improving the quality and cost-effectiveness of these procedures has become an increasing priority.

In response to rising health care expenditures, the United States health care system has increasingly emphasized value-based care models aimed at improving quality while reducing cost [6,7]. Medicare has taken an active role in this transformation by implementing alternative payment models for high-volume, high-cost procedures, such as TJA [6].

In April 2016, the Centers for Medicare and Medicaid Services (CMS) introduced the Comprehensive Care for Joint Replacement (CJR) model, a bundled payment initiative designed to reduce arthroplasty expenditures and encourage value-based care [8]. Under this bundled payment model, hospitals assumed responsibility for both the cost and quality of the entire episode of arthroplasty care, spanning the index admission through 90 days postdischarge. Although participation was limited to select metropolitan areas, CJR reflects broader Medicare efforts to standardize and reduce variation in arthroplasty reimbursement.

Despite these reforms, substantial geographic variation in Medicare reimbursement for arthroplasty persists. Prior studies evaluating Medicare reimbursements for TJA patients have found significant differences in reimbursement by geographic region within the United States, with some states having more than double the reimbursement of others [6,9]. To our knowledge, few studies have examined state- and region-level Medicare inpatient reimbursement for TKA and THA over the decade spanning 2013 to 2023, a period that includes the coronavirus disease 2019 pandemic and TKA and THA removal from the Medicare inpatient–only list [10,11]. Therefore, the purpose of this study was to describe and compare trends in Medicare inpatient reimbursement for all episodes of care of primary THA and TKA across US states and regions from 2013 to 2023.

Material and methods

Data source

We used the publicly available “Medicare Inpatient Hospitals by Provider and Service” database from the US CMS. This dataset includes all Medicare inpatient Part A claims from more than 3000 US hospitals billing Medicare for at least 10 annual episodes of care [12]. Search criteria focused on the diagnostic-related group (DRG) 470 (major hip and knee joint replacement without major complication or comorbidity) for each year from 2013 to 2023. Cases billed under DRG 469 (major joint replacement with major complication or comorbidity) and outpatient arthroplasty procedures were not captured in the present study.

Part A hospital payments for DRG 470 encompass all expenses related to the surgical procedure, implants, and postoperative inpatient hospital care [8]. For each state and year, we extracted the total number of discharges, the mean and total Medicare reimbursement (mean payment multiplied by discharges), and the charges submitted by hospitals. Hospital geographic information was also recorded, and regional classification (Northeast, Midwest, South, and West) was defined by the US Census [13].

Data preparation

All monetary values were adjusted to 2023 US dollars using the US Consumer Price Index for all urban consumers (US city average, all items) published by the Bureau of Labor Statistics [14]. State-level estimates of Medicare reimbursement were calculated as discharge-volume–weighted means, in which hospital-level average payments were weighted by their respective discharge counts so that estimates reflected the experience of the average Medicare beneficiary rather than the average hospital. For each region and year, the weighted mean was defined as the sum of each hospital’s mean payment multiplied by its discharge volume, divided by the total discharges in that region-year.

Statistical analysis

Descriptive statistics were generated at the state, regional, and national levels for each year and included the total number of discharges, mean Medicare reimbursements per discharge, mean submitted charges by the hospital, and the percent change over the study period. Regional comparisons of 10-year changes in reimbursement were evaluated using the nonparametric Kruskal-Wallis test. To examine longitudinal trends across all years, we fit a linear mixed-effects model with state-level random intercepts and discharge volume as analytic weights. The model included fixed effects for year, region, and year-by-region interactions, thereby estimating both the overall annual rate of change and difference in slopes between regions. From this model, we obtained discharge-weighted annual changes in Medicare payment per discharge with 95% confidence intervals. Pairwise comparisons of regional slopes were conducted with Tukey's adjustment to account for multiple testing. For each state-year, we calculated discharge rates per 100,000 population. Regional rates were then derived by summing discharges and populations across states within each region and year. To compare regions within each year, we performed 1-way analysis of variance (ANOVA) on state-level rates, with region as the factor. When the omnibus test was significant (α = 0.05), we conducted Tukey’s honest significant difference tests to identify specific regional differences. Results are reported as population-weighted regional rates, ANOVA P-values, and significant pairwise comparisons. All statistical analyses were performed using R (version 4.4.3), and figures were created in Microsoft Excel (version 16.100). A P-value less than 0.05 indicated significance.

This study was exempt from institutional review board approval as no patient information was used. No outside funding was received for this study.

Results

Descriptive statistics

A total of 3,724,353 primary THA and TKA procedures were billed to Medicare by US hospitals from 2013 to 2023 under DRG 470 (Table 1). The annual number of THA and TKA procedures billed to Medicare inpatient services decreased by 379,664 procedures (−84.1%) from 451,603 in 2013 to 71,939 in 2023. Prior to adjusting for inflation to 2023 US dollars, the mean Medicare reimbursement increased by $1607.29 per episode (+13.3%), from $12,088.27 in 2013 to $13,695.56 in 2023. Meanwhile, the mean charge submitted by hospitals increased by $31,023.89 (+56.8%) per episode, from $54,651.54 in 2013 to $85,675.43 in 2023.

Table 1.

National Medicare trends in DRG 470 surgical volume, charges, and payments, 2013-2023.

Year	Total surgeries	Total charges	Total Medicare payments	Average charge per procedure (N ± standard deviation)	Average Medicare reimbursement per procedure (N ± standard deviation)
2013	451,603	$32,275,705,486.89	$7,139,003,419.85	$71,469.20 ± $35,932.16	$15,808.14 ± $4111.48
2014	455,382	$33,020,798,957.75	$7,133,010,107.70	$72,512.31 ± $36,590.46	$15,663.79 ± $4278.75
2015	471,763	$35,099,250,029.33	$7,258,817,076.73	$74,400.18 ± $38,623.18	$15,386.58 ± $4193.49
2016	504,279	$38,015,365,368.59	$7,557,915,689.74	$75,385.58 ± $40,863.56	$14,987.57 ± $4973.19
2017	513,699	$38,728,629,788.59	$7,518,384,851.46	$75,391.68 ± $42,101.48	$14,635.78 ± $4281.27
2018	443,252	$34,028,892,847.77	$6,348,329,305.82	$76,770.99 ± $43,328.37	$14,322.17 ± $4258.99
2019	413,876	$32,275,222,789.58	$5,839,540,489.84	$77,982.83 ± $44,391.21	$14,109.39 ± $4204.86
2020	203,480	$16,818,698,758.42	$2,918,358,394.39	$82,655.29 ± $46,823.41	$14,342.24 ± $4484.27
2021	110,827	$9,182,158,262.10	$1,559,666,954.54	$82,851.28 ± $49,323.12	$14,072.99 ± $4777.07
2022	84,253	$6,979,646,704.20	$1,127,442,847.06	$82,841.52 ± 48,566.41	$13,381.63 ± 3381.53
2023	71,939	$6,163,404,826.99	$985,245,234.62	$85,675.43 ± 50,817.75	$13,695.56 ± 3782.53
Δ 2013-2023	−379,664 (−84.1%)	$−$26,112,300,659.90 (−80.9%)	$−$6,153,758,185.23 (−86.2%)	$14,206.23 (+19.9%)	−$2112.58 (−13.4%)

All monetary values were adjusted for inflation to 2023 US dollars.

The mean inflation–adjusted (all monetary values adjusted to 2023 US dollars) Medicare reimbursement to hospitals for such episodes spanning 2013 to 2023, decreased from $15,808.14 to $13,695.56 for DRG 470 (−$2112.58; −13.4%). The mean inflation–adjusted charge submitted by hospitals for DRG 470 increased by $14,206.23 (+19.9%) per episode, from $71,469.20 in 2013 to $85,675.43 in 2023.

Geographical variance

In 2023, the 3 states with the highest mean Medicare inpatient reimbursement included Maryland ($23,530.13), Alaska ($19,530.81), and Wyoming ($18,365.77). The 3 states with the lowest mean reimbursement per episode included Alabama ($9451.15), Vermont ($10,613.94), and Arkansas ($10,765.57) (Fig. 1). The change in mean reimbursement from 2013 to 2023 varied greatly, ranging from a minimum change of −$201.04 (−1.4%) in Utah to a maximum change of −$9642.08 (−47.60%) in Vermont (Fig. 2). No state had an inflation-adjusted increase in mean Medicare reimbursement over this time frame.

Figure 1.

2023 Mean Medicare reimbursement to hospitals for hip and knee arthroplasty episodes of care under DRG 470. All monetary values were adjusted for inflation to 2023 US dollars.

Figure 2.

Change in inflation-adjusted Medicare reimbursement to hospitals for hip and knee arthroplasty episodes of care under DRG 470 from 2013 to 2023. All monetary values were adjusted for inflation to 2023 US dollars.

From 2013 to 2023, inflation-adjusted (all monetary values adjusted to 2023 US dollars) Medicare reimbursements per discharge declined in all 4 US regions. Absolute 10-year changes were descriptively largest in the Midwest and smallest in the West (Table 2). Notably, the Midwest had the highest nonweighted mean reimbursement in both 2013 and 2023. A Kruskal-Wallis rank-sum test likewise found no statistically significant difference between regional changes (χ² = 3.0, df = 3, P = .39). When regional reimbursements were weighted by discharge volume, the Midwest had the lowest mean reimbursement ($12,222) and experienced the largest decrease over time (−19.1%), whereas the West remained the region with the smallest change over time (−10.1%) (Table 3). Further, when examining longitudinal reimbursement trends, all 4 regions demonstrated significant yearly declines in Medicare payment per discharge (P < .001). The estimated annual decreases ranged from approximately −$249 in the South to −$307 in the Northeast, with the Midwest and West falling in between (Table 4). As demonstrated in Figure 3, in discharge-weighted linear mixed-effects regression with random intercepts for state, there was a significant overall decline of approximately −$285 per discharge per year across the study period (P < .001). The South had a significantly different slope compared with the Midwest, declining less markedly (+$36 per year relative difference, P < .05). Interaction terms for the Northeast (−$21 per year relative difference, P > .05) and West (−$11 per year relative difference, P > .05) were not statistically significant, indicating similar downward trajectories to the Midwest. From 2013 through 2020, ANOVA testing demonstrated significant differences between regions, most consistently showing higher discharge rates in the Midwest compared with the West, South, and Northeast. Post hoc Tukey testing confirmed that the Midwest differed significantly from the other regions during these years. No statistically significant regional differences were observed after 2020 (Table 5).

Table 2.

Absolute change in inflation-adjusted Medicare payments per discharge, 2013-2023 (unweighted).

Region	2013 Medicare reimbursement ($)	2023 Medicare reimbursement ($)	Absolute change ($)	Percent change (%)
Midwest	17,741	14,300	−3440	−19.4
Northeast	15,067	12,184	−2883	−19.1
South	14,993	12,731	−2262	−15.1
West	16,474	14,061	−2413	−14.7

All monetary values were adjusted for inflation to 2023 US dollars.

Table 3.

Change in inflation-adjusted Medicare payment per discharge, 2013-2023 (weighted by discharges).

Region	2013 Medicare reimbursement ($)	2023 Medicare reimbursement ($)	Absolute change ($)	Percent change (%)
Midwest	15,106	12,222	−2884	−19.1
Northeast	17,538	15,216	−2322	−13.2
South	14,660	12,235	−2424	−16.5
West	17,810	16,015	−1795	−10.1

All monetary values were adjusted for inflation to 2023 US dollars.

Table 4.

Estimated yearly change in payment per discharge by region.

Region	Estimated yearly change ($/y)	95% CI	P-value
Midwest	−285.3	[−311.0, −259.6]	<.001
Northeast	−306.5	[−331.8, −281.2]	<.001
South	−248.7	[−272.9, −224.5]	<.001
West	−296.2	[−321.5, −270.9]	<.001

Figure 3.

Regional Medicare reimbursement trends (weighted by case volume) for hip and knee arthroplasty episodes of care under DRG 470 from 2013 to 2023. All monetary values were adjusted for inflation to 2023 US dollars.

Table 5.

Regional discharge rates per 100,000 population.

Year	West	Midwest	South	Northeast	ANOVA P-value	Significant pairs
2013	109	180.8	148.6	130.2	0.000073	West-Midwest (Δ = −86.7 per 100 k; P = 4.02e-05); Northeast-Midwest (Δ = −67.2 per 100 k; P = .0046); South-Midwest (Δ = −50.5 per 100 k; P = .015)
2014	109.6	177.4	147.8	136.2	0.00022	West-Midwest (Δ = −86.2 per 100 k; P = .0001); Northeast-Midwest (Δ = −62.5 per 100 k; P = .015); South-Midwest (Δ = −49.6 per 100 k; P = .027)
2015	111.1	183.3	151.7	142.2	0.00022	West-Midwest (Δ = −94.7 per 100 k; P = 9.93e-05); Northeast-Midwest (Δ = −67.6 per 100 k; P = .017); South-Midwest (Δ = −54.9 per 100 k; P = .025)
2016	119	195.5	159.8	151.1	0.00054	West-Midwest (Δ = −93.9 per 100 k; P = .00027); Northeast-Midwest (Δ = −68.6 per 100 k; P = .023); South-Midwest (Δ = −56.1 per 100 k; P = .032)
2017	121.9	198.9	159.6	155	0.00053	West-Midwest (Δ = −95.3 per 100 k; P = .00028); Northeast-Midwest (Δ = −70.9 per 100 k; P = .0205); South-Midwest (Δ = −60.2 per 100 k; P = .022)
2018	105.1	169.7	133.7	141.2	0.0016	West-Midwest (Δ = −79.8 per 100 k; P = .0012); South-Midwest (Δ = −60 per 100 k; P = .012)
2019	94.9	154	127.2	133.2	0.0025	West-Midwest (Δ = −75.9 per 100 k; P = .0012); South-Midwest (Δ = −47.8 per 100 k; P = .048)
2020	42.5	70.3	62.8	73.3	0.014	West-Midwest (Δ = −43.9 per 100 k; P = .0067)
2021	23.5	37.6	29.8	49.8	0.084	—
2022	18.6	26.4	23.8	36.1	0.21	—
2023	17	23.5	18.3	31.8	0.13	—

In this table, Δ represents the pairwise difference in discharge rates (per 100,000 population) between regions in a given year.

When overall differences were significant by ANOVA, post hoc pairwise comparisons were performed using Tukey’s honest significant difference test to identify specific regional differences.

Discussion

The current study demonstrates that between 2013 and 2023, inflation-adjusted Medicare inpatient reimbursements for THA and TKA markedly decreased, despite increasing hospital-submitted charges. Importantly, Medicare reimbursement increased in unadjusted dollars over this period but failed to keep pace with inflation, resulting in a 13.4% decline in real reimbursement by 2023. This change in mean reimbursement varied across states, and by geographic region, with the Midwest experiencing the largest decrease in mean reimbursement over the decade and the lowest discharge-weighted mean reimbursement in 2023.

To the best of our understanding, this is one of the first in the literature to comprehensively evaluate geographic disparities of inpatient hospital Medicare reimbursements and case volume for primary THA and TKA episodes of care billed to Medicare in recent years. That said, at a nationwide and regional level, trends of declining inflation-adjusted mean Medicare reimbursements and increasing inflation-adjusted mean charges per procedure are well known. Previous literature evaluating large Medicare datasets has consistently demonstrated declining inflation-adjusted reimbursements for arthroplasty despite rising hospital charges. Haglin et al. [6] reported an 11.4% decrease in Medicare reimbursement to hospitals between 2011 and 2017, alongside a 9.3% increase in submitted charges. Similarly, Lopez et al. and Schilling et al. [15,16] found reductions in Medicare payments for arthroplasty procedures during the early implementation of value-based payment models. More recently, Palmer et al., using the CMS Physician Fee Schedule Look-Up Tool, identified a 53% decrease in inflation-adjusted Medicare reimbursements for all THA and TKA procedures from 2000 to 2024. While these studies looked at discrepancies on a larger scale, variation in Medicare reimbursement and hospital charges also exists at the regional and hospital levels [[17], [18], [19], [20], [21], [22]].

The current study similarly shows variation in Medicare reimbursement statewide and regionwide. Across all regions of the United States, inflation-adjusted mean Medicare reimbursement decreased between 2013 and 2023, regardless of being weighted or unweighted to account for state variability. This coincided with a nationwide 13.4% decrease in inflation-adjusted reimbursement over this time frame. In unweighted analyses, where each state contributed equally, the Midwest had the largest mean Medicare reimbursement in both 2013 and 2023; however, the Midwest also experienced the largest absolute change (−$3440; −19.4%) in reimbursement over the decade, with the Northeast following closely behind (−19.1%). These findings suggest that at the state level, hospitals in the Midwest and Northeast bore the greatest relative financial contraction over the study period.

When regional means were adjusted for discharge volume, reflecting the experience of the average Medicare beneficiary rather than the average hospital, the Midwest no longer held the highest mean Medicare reimbursement in 2023 but rather the lowest. Further, it remained the most adversely affected region with reimbursements declining 19.1% from 2013 to 2023. Despite starting from the lowest inflation–adjusted mean reimbursement in 2013, the South was not far behind, with a 16.5% decrease in mean Medicare reimbursements by 2023. The West, on the other hand, demonstrated the smallest decline (−10.1%) and consistently maintained the highest mean Medicare inpatient reimbursements (weighted by discharges) across the study period. Collectively, these results highlight that hospitals in the Midwest have experienced disproportionately greater reductions in Medicare reimbursement.

Marked variation was also observed at the state level. In 2023, mean reimbursements ranged more than 2-fold, from a high of $23,530 in Maryland to a low of $9451 in Alabama. Alaska ($19,531) and Wyoming ($18,366) joined Maryland as the highest-reimbursed states, whereas Vermont ($10,614) and Arkansas ($10,766) ranked among the lowest. No state experienced an inflation-adjusted increase in Medicare reimbursements during the study period when adjusting all monetary values to 2023 US dollars. The magnitude of decline also varied substantially, with Vermont demonstrating the steepest relative decrease (−47.6%), followed by Alabama (−27.0%) and Massachusetts (−24.9%). In contrast, Utah (−1.4%), Maryland (−2.1%), and Georgia (−7.7%) had the most modest reductions. These findings demonstrate that Medicare reimbursements for TJA remain highly heterogenous across the country, with some states reimbursed at less than half the rate of others for identical DRG-coded procedures. Particularly notable is Vermont, which experienced both a low terminal mean reimbursement rate and the greatest proportional decline. Conversely, Maryland and Alaska maintained the highest reimbursement levels.

Maryland’s consistently higher reimbursements are likely influenced by its unique all-payer rate-setting model called the global budget revenue. In this model, hospitals still receive payments from private and public payers, but with enforced rates, there is a common rate for hospital-based inpatient, outpatient, and emergency services [23,24]. When assessing the Maryland TKA cohort from 2014 to 2016, prior work demonstrated that both hospital costs and charges decreased, whereas costs and charges increased nationwide [24]. These findings are consistent with the current study, in which Maryland had the highest reimbursements in 2023 and the second smallest reduction in reimbursement over time. This speaks to the success of the state’s global budget revenue model and the potential benefit of more expansive care-based models. Notably, Vermont also implemented an all-payer model (the All-Payer ACO Model) beginning in 2017 [25]; however, unlike Maryland, the state demonstrated the lowest reimbursements in our study. This discrepancy suggests that the structure and implementation of an all-payer model, as well as its ability to adapt to state-specific demographics, case volumes, and market conditions, may be critical to its success.

Variation in hospital charges for TJA is also well documented [9,15,26]. Proposed drivers for increasing prices include hospital market consolidation [27] and the rising cost of joint implants [28,29]. Prior work by Cram et al. [30] demonstrated that patient characteristics account for only a small proportion of cost variability, with hospital factors and unexplained variation accounting for the majority. Similarly, Haas and Kaplan [5] reported wide variation in episode-of-care costs for TKA even among hospitals with comparable patient populations and procedural volumes. Pollock et al. [31] also reported substantial geographic variability in TJA costs with little correlation to cost of living or median income. Collectively, these findings demonstrate the persistent variation within arthroplasty pricing and highlight ongoing opportunities to improve TJA care and payment models.

Similar to the findings of the current study, prior investigations have also demonstrated substantial variability in Medicare reimbursements for arthroplasty [6,9,32]. Thakore et al. [9] reported the high variability in reimbursements that they found was not fully explained by Medicare’s methodology for calculating reimbursements. According to their study findings, there was a significant difference in mean Medicare reimbursements by region, with the West receiving the greatest reimbursement and the South with the lowest reimbursement, nearly $3000 less [9]. Li et al. [32] likewise found that hospitals in the Northeast and West received higher Medicare payments than those in the Midwest, even after adjustment for regional price differences. More recently, Gill et al. [33] evaluated Medicare outpatient reimbursement trends for TKA from 2013 to 2021 and identified the Midwest as the most adversely affected region, with the greatest interval decline and the lowest mean reimbursement. This finding is particularly striking given that the Midwest has been recognized as a region with high TKA utilization [15,[33], [34], [35]]. In a complementary analysis of outpatient THA reimbursements over the same period, Gill et al. [36] again observed the lowest mean reimbursements and the steepest decline in the Midwest, with no region demonstrating an increase in reimbursements over time. Collectively, these studies, in conjunction with our present findings, highlight persistent and enduring variability in Medicare reimbursement across states and regions that has continued into the most recent decade.

In the current study, we observed a substantial contraction in annual inpatient Medicare discharges for TJA, decreasing from 451,603 cases in 2013 to 71,939 in 2023 (−84.1%). Additionally, regional variation was most pronounced between 2013 and 2020, with consistently higher discharge rates in the Midwest compared with other regions, particularly the West. These differences diminished after 2020, and no significant regional disparities were observed from 2021 onward, following a marked respective decrease in case volume in the Midwest. The higher discharge rate in the Midwest is particularly notable given its low regional mean reimbursement in the last decade.

While the overall national volume of arthroplasty has steadily increased over the past decade [37], the present study spans several major policy changes affecting arthroplasty care, including the introduction of the CJR model in 2016 and the subsequent removal of TKA and THA from the Medicare inpatient–only list in 2018 and 2020, respectively [10,11]. Following these policy changes, arthroplasty procedures increasingly migrated to outpatient settings. By 2020, more than half of all Medicare TKA procedures were reportedly classified as outpatient [10], and Gordon et al. [38] demonstrated an increase in outpatient THA volume from 5.7% to 35.6% between 2019 and 2020. Further, Pasqualini et al. [11] found that outpatient THA volume increased 751% from 2019 to 2021. While these initiatives were intended to improve the quality and cost of arthroplasty care through value-based payment models and more efficient care delivery, the present study focused on inpatient DRG 470 procedures, which increasingly represent a more medically complex subset of arthroplasty patients as low-risk cases move to outpatient settings. These shifts correspond with the pattern observed in the current study, which showed an increase in annual inpatient case volume through 2017, followed by a dramatic decline from 413,876 cases in 2019 to 203,480 cases in 2020. With the coronavirus disease 2019 pandemic causing dramatic declines in elective arthroplasty procedures [[38], [39], [40], [41], [42]] and previous literature reporting outpatient knee and hip surgical volume did not dramatically increase until the pandemic started in 2020, the combination of policy changes and pandemic-related disruptions likely compounded the reduction in inpatient cases.

The findings of this study have several implications for arthroplasty practice, hospital systems, and health care policy. Although a growing proportion of TJA procedures have transitioned to the outpatient setting, inpatient arthroplasty continues to represent a substantial portion of care for medically complex Medicare patients. Continued reductions in Medicare reimbursement for inpatient TJA, particularly in regions already experiencing lower payments, such as the Midwest, may place disproportionate financial pressure on hospitals that maintain high case volumes. Notably, the Midwest historically demonstrated the highest inpatient discharge rates in our analysis while experiencing the greatest decline in reimbursement, highlighting a potential mismatch between procedural demand and financial support for arthroplasty care in certain regions. Future payment reforms may benefit from closer evaluation of geographic variation to ensure equitable and sustainable reimbursement for arthroplasty services across the United States.

This retrospective study contains several important limitations. First, as noted above, outpatient TJA cases were not captured, despite comprising a growing share of the arthroplasty market in recent years. A more comprehensive analysis would require incorporation of DRG 469 (Major Joint Replacement or Reattachment of Lower Extremity with Major Complication or Comorbidity) as well as outpatient procedural data. In addition, CMS administrative datasets are reliant on accurate billing, coding, and data entry. While some degree of error is possible, we believe that such errors are minimal and that the database remains a reliable reflection of national trends. Another limitation is that the dataset is restricted to financial variables, without accompanying clinical outcomes or detailed descriptors of the case or perioperative care, making it difficult to fully assess drivers of cost or reimbursement variability.

Conclusions

Inflation-adjusted Medicare inpatient reimbursement for primary THA and TKA decreased significantly from 2013 to 2023 in all states, despite increasing hospital-submitted charges. Declines were consistent across all regions, with the steepest reductions in the Midwest and the smallest in the West. These findings highlight persistent downward pressure on hospital reimbursements for TJA, and the geographic variability present may have negative implications for access to arthroplasty care and the advancement of a widely agreeable reimbursement model.

CRediT authorship contribution statement

Adam B. Thompson: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Benjamin R. Paul: Writing – original draft, Methodology, Formal analysis, Data curation. Jack M. Haglin: Writing – original draft, Supervision, Methodology, Investigation, Formal analysis, Conceptualization. Paul R. Van Schuyver: Writing – review & editing, Writing – original draft, Methodology, Formal analysis. Jose M. Iturregui: Writing – review & editing, Writing – original draft, Methodology, Formal analysis. David G. Deckey: Writing – review & editing, Methodology, Conceptualization. Joshua S. Bingham: Writing – review & editing, Writing – original draft, Supervision, Methodology, Investigation.

Conflicts of interest

David G. Deckey, MD, holds stock or stock options in Osgenic; serves on the editorial board of the Journal of Arthroplasty; and holds committee and board appointments with the American Association of Hip and Knee Surgeons and the American Academy of Orthopaedic Surgeons. Joshua S. Bingham, MD, reports financial compensation from Sylke, Inc for participation in a speakers bureau. All other authors declare no potential conflicts of interest.

For full disclosure statements refer to https://doi.org/10.1016/j.artd.2026.102016.