Disparities in Limb Preservation and Associated Socioeconomic Burden among Patients with Diabetes and/or Peripheral Artery Disease in the United States

Abstract

Racial, ethnic, socioeconomic, and geographic disparities in limb preservation and non-traumatic lower extremity amputation (LEA) are consistently demonstrated in populations with diabetes and peripheral artery disease (PAD). Higher rates of major LEA in disadvantaged groups are associated with increased healthcare utilization and higher costs of care. Functional decline that often follows major LEA confers substantial risk of disability and premature mortality, and the burden of these outcomes is more prevalent in racial and ethnic minorities, persons with low socioeconomic status, and persons in geographic regions where limited resources or distance from specialty care are barriers to access. Here, we present a narrative review of the existing literature on estimated costs of diabetic foot disease and PAD, inequalities in care that contribute to excess costs, and disparities in outcomes that lead to a disproportionate burden of diabetes- and PAD-related LEA on systematically disadvantaged populations.

Introduction

Major lower extremity amputation (LEA) is a morbid complication of diabetes and peripheral artery disease (PAD) associated with high healthcare costs, significant functional disability, and early mortality [1-4]. Non-traumatic LEA related to diabetes and PAD is a source of preventable morbidity that disproportionately affects persons of racial and ethnic minority groups and low socioeconomic status [5,6]. The current narrative review aims to summarize existing literature on healthcare cost and utilization associated with diabetic lower extremity disease and PAD, inequalities in care that contribute to excess costs, and outcomes disparities that lead to a disproportionate economic burden of LEA on systematically disadvantaged populations in the United States (US).

Existing literature on the healthcare expenditures associated with treating threatened limbs is limited and features highly variable methodology. It is additionally challenging to contextualize existing literature given the changing prevalence and medical complexity of persons with diabetes and PAD, and the differences in healthcare funding, access, and delivery methods between regions and countries. The current narrative review focuses on recent US data to limit the variability conferred by attempting to compare multiple currencies and national health systems; however, authors recognize that high costs and disparate outcomes are a global problem. Broadly, publications were considered for inclusion if they were published after 2010, used US healthcare and/or economic data, and reported costs (rather than charges).

Shifting epidemiology of type 2 diabetes, PAD, and LEA

Diabetes and PAD affect over 37 million and 8.5 to 12 million US adults, respectively [1,6-8]. The US prevalence of type 2 diabetes is rising rapidly due to disparities in social determinants of health, high obesity rates, and improved longevity of persons with diabetes despite increasing medical complexity [8,9]. The prevalence of PAD is also rising, driven in part by high diabetes prevalence and duration, and by the aging of the US population [1,8]. Between 20% and 30% of all adults with diabetes have diagnosed PAD, including up to 50% of adults 75 years of age and older with diabetes [1,6,8]. Comorbid diabetes and PAD substantially increase the risk of poor limb outcomes including diabetic foot ulcer and minor and major LEA [1].

In the US, over 150,000 minor or major LEA are performed annually in persons with diabetes, and 30,000 to 40,000 are performed in persons with chronic limb threatening ischemia (CLTI) without diabetes [7,8]. Patterns of LEA have shifted toward a higher proportion of minor amputations, which now comprise an estimated 65% to 75% of all PAD- and diabetes-related LEA [1,2,10]. While the incidence of major LEA steadily decreased over the late 20^th and early 21^st centuries, rates have plateaued or increased in many regions over the past decade [2,7]. There has been a marked increase in rates of major non-traumatic LEA noted since 2010 among adults with diabetes aged 18 to 44 years old [11], with significant ramifications for healthcare costs, lost productivity and employment, and early mortality in this young population. Substantial attention from national medical, surgical, and public health organizations and the inclusion of reducing diabetes-related LEA among the US Healthy People 2030 health indicator goals have not meaningfully altered the trajectory of these trends [2,7,10,12].

The economic burden of diabetic limb complications and LEA

The annual direct healthcare costs of diabetes in the US are estimated to total $273 billion USD. Diabetes is responsible for an additional $90 billion per year in indirect costs, reflecting reduced productivity and lost employment (2017 USD) [13]. Diabetes-related lower extremity complications are both common and expensive. Up to one in three persons with diabetes will develop a foot ulcer during their lifetime, and one in five persons with an incident ulcer will go on to require a minor or major LEA [14]. Population-based studies suggest that diabetic foot ulcers and their sequelae account for at least 20% of the total healthcare costs of diabetes [4,13,15]; conservatively, this represents $54 billion in annual direct costs. Total PAD-related healthcare costs in the US are not well-documented but have been previously estimated to exceed $84 billion annually, upwards of $6 to $12 billion accounted for by hospitalizations related to limb events in PAD and CLTI [1,16,17]. In persons with diabetes and/or PAD, inpatient costs related to lower extremity wounds or ischemia, especially for treatment of infection, revascularization, and minor or major LEA, are estimated to account for 50% to 80% of excess healthcare expenditures compared to matched controls (i.e. compared to persons with diabetes but no foot complications, or persons without diagnosed PAD who have otherwise similar comorbidity profiles) [18,19]. Healthcare utilization in adults with PAD and/or diabetes is also higher across outpatient, home care, and pharmacy cost domains [20,21].

The incremental costs related to diabetic limb complications represent a major source of preventable healthcare spending and utilization. Retrospective claims-based analyses in persons with diabetic foot ulcers report 2- to 5.4-fold higher direct healthcare costs over the year following a foot ulcer diagnosis compared to similar persons without a foot ulcer, and these higher total costs persist for at least 2 years [19]. Several studies report higher rates of hospital admission, emergency department visits, outpatient visits and home healthcare use associated with diabetic foot complications [19,22,23]. Similarly, any diagnosis of PAD is associated with doubling of mean annual healthcare expenditure compared to age- and sex-matched persons without PAD, driven primarily by hospital admissions and out of pocket pharmacy costs [24].

More granular data show limb treatment costs correlate with increasing severity of limb threat. For persons with PAD and/or diabetes requiring inpatient admission for the treatment of tissue loss or infection, even early (stage 1) wounds, as classified by the Society for Vascular Surgery’s Wound, Ischemia and foot Infection (Wifi) index, are associated with mean index admission costs near $15,000 (2017 USD) [25]. Advanced (Wifi stage 3 to 4) wounds are estimated to cost at least $50,546±$4,887 (2016 USD) from diagnosis to healing or LEA (Table 1) [20,25]. More frequent and costly inpatient admissions and more inpatient and outpatient procedures required to treat advanced staged wounds contribute to these high costs [18,20,25].

Table 1.

Reported costs associated with limb preservation based on wound stage/severity of limb threat

Low stage wounds	Cost methodology	Cost and utilization findings	Additional notable findings
Ramanan et al. 2017 [25] Single-center, retrospective analysis of consecutive CLTI-related admissions (n=157 limbs) to a multidisciplinary limb preservation service from 2013-2014. WIfI 1 n=20 limbs WIfI 2 n=48 limbs	Inpatient hospitalization costs for index admission and total 1-year hospital costs from date of index admission. Not including outpatient costs. Costs reported in USD, not standardized to a single year.	WIfI 1 Index hospitalization costs (mean): $14,919 Admissions per patient (mean): 1.8 1-year costs (mean): $27,057 WIfI 2 Index hospitalization costs (mean): $25,233 Admissions per patient (mean): 1.9 1-year costs (mean): $30,588	Major LEA performed in 8% of WIfI 1 wounds and 7% of WIfI 2 wounds. No significant difference in ambulatory status outcomes between white, Asian, Black, and Hispanic groups
Hicks et al. 2018 [20] Single-center retrospective analysis of consecutive patients (n=248 patients, 319 wound episodes) treated in a multidisciplinary diabetic limb preservation service from 2012-2016. WIfI 1 n=98 wounds WIfI 2 n=50 wounds	Inpatient and outpatient costs per wound episode (endpoint healing, major LEA, or death). Costs reported in 2016 USD.	WIfI 1 Total costs (mean±SD): $3,995±1,047 Admissions per patient (mean±SD): 2.07±0.48 WIfI 2 Total costs (mean±SD): $9462±$1661 Admissions per patient (mean±SD): 1.50±0.18	Major LEA performed in 0% of WIfI 1 wounds and 3.1%±3.1% of WIfI 2 wounds.
High stage wounds	Cost methodology	Cost and utilization findings	Additional notable findings
Ramanan et al. 2017 (see above) WIfI 3 n=41 limbs WIfI 4 n=48 limbs	See above. Reported in USD, not standardized to a single year.	WIfI 3 Index hospitalization costs (mean): $25,522 1-year costs (mean): $29,092 Admissions per patient (mean): 1.9 WIfI 4 Index hospitalization costs (mean): $34,055 1-year costs (mean): $52,729 Admissions per patient (mean): 2.1	Major LEA performed in 5% of WIfI 3 wounds and 8% of WIfI 4 wounds. No significant difference in ambulatory status outcomes between white, Asian, Black, and Hispanic groups
Hicks et al. 2018 (see above) WIfI 3 n=95 wounds WIfI 4 n=76 wounds	See above. Reported in 2016 USD.	WIfI 3 Total costs (mean±SD): $31,810±$4,939 Admissions per patient (mean±SD): 2.24±0.20 WIfI 4 Total costs (mean±SD): $50,546±$4,887 Admissions per patient: 3.40±0.27	Major LEA performed in 6.9%±4.3% of WIfI 3 wounds and 6.4%±3.1% of WIfI 4 wounds.

Standard deviation or 95% confidence interval were not uniformly reported, and missing values reflect non-reporting in the study cited. All reported data is in costs (rather than charges or other non-cost expenditure metrics.)

CLTI=chronic limb threatening ischemia; WIfI=Society for Vascular Surgery’s Wound, Ischemia, and Foot Infection classification system; USD=United States dollars; LEA=lower extremity amputation

LEA is especially costly. Minor LEA is an important element of functional limb preservation and is associated with mean perioperative costs near $45,000 per minor LEA, although the reported range is wide and the definition of costs and description of included services are highly variable (Table 2) [26]. Major LEA represents an even greater expense, with mean reported perioperative costs ranging from $55,000 to $83,000 (2016 and 2013 USD, respectively) (Table 2) [26,27].

Table 2.

Reported US costs of minor and major lower extremity amputation (LEA)

Minor LEA	Cost methodology	Cost findings	Additional notable findings
Franklin et al. 2014[27] Cross-sectional analysis of VHA clinic users with diabetes who underwent LEA during fiscal year 2010 (n=3,403) Toe amputation n=1,454	One year inpatient medical, inpatient surgical, outpatient care and pharmacy costs among persons with prevalent LEA. Reported in 2012 USD.	Toe amputation 1-year costs [mean (95% CI)]: $41,484 ($40,075, $42,943)	Toe amputation was the most common amputation in this cohort. Average costs of all true minor LEA (i.e., including other inframalleolar amputations) were not reported.
Sareh et al. 2020[62] Retrospective review of all persons who underwent minor LEA in the US Nationwide Readmission Database (n= 302,798 adults, mean age 61.8 years, 80% diabetes prevalence and 41% PAD prevalence), comparing 90-day outcomes, including inpatient costs, between Frail and not-Frail adults. Race and ethnicity were not reported.	Inpatient costs for index discharge through 90 days post-discharge (index and readmission inpatient costs). Reported in 2015 USD.	Not-Frail adults 90-day costs (mean): $27,244 Crude 90-day readmission rate: 36% Frail adults 90-day costs (mean): $39,417 Crude 90-day readmission rate: 44%	Frail adults were more likely to have Medicare, and less likely to have private insurance. Admission to a rural hospital (aOR 1.23) predicted readmission and major re-amputation, but no comparison of costs based on major reamputation or rural status were reported.
Stewart et al. 2022 [53] Retrospective review of all persons who underwent LEA at a single public safety-net hospital in San Francisco during the year 2013. Minor LEA n=76	Index admission hospital costs and one year hospital and outpatient costs, not including outpatient rehabilitation or prosthetics. Reported in USD, not standardized to year.	Index hospitalization costs (mean): $45,513 1-year costs (mean): $59,549	Persons who underwent LEA also had high cumulative admission costs in the 1 year prior to LEA admission.
Major LEA	Cost methodology	Cost findings	Additional notable findings
Franklin et al. 2014 (see above)[27] BKA (including TMA) n=1,303 AKA n=646	One year inpatient, outpatient, and pharmacy costs among persons with prevalent LEA. Reported in 2012 USD.	BKA (including TMA, but not including toe) 1-year total costs [mean (95% CI)]: $71,067 (68,449–73,785) AKA 1-year total costs [mean (95% CI)]: $82,758 (78,063–87,736)	Marked increase in inflation-adjusted costs from FY2004 to FY2010, with total LEA-related costs in the VHA population with diabetes increasing from $170 million to $206 million from 2004 to 2010 (both 2012 USD).
Mustapha et al. 2018[17] Adults (n=72,199) enrolled in Medicare FFS with an incident CLTI diagnosis in 2011.	Costs (inpatient, outpatient, and physician costs) from incident CLTI diagnosis through 4 years following diagnosis. Costs reported in 2016 USD and standardized to person year of follow-up/survival.	Total healthcare costs per person year following LEA (mean): $55,700	Black race was associated with increased risk of primary and any subsequent major LEA [aOR (95% CI) 1.60 (1.44,1.79)] or mortality [1.05 (1.00, 1.11)] across all severities of presentation.
Stewart et al. 2022 [53] See above. Major LEA n=14 (includes 3 LEA for trauma)	Index admission hospital costs and one year hospital and outpatient costs, not including outpatient rehabilitation or prosthetics. Reported in USD, not standardized to year.	Index hospitalization costs (mean): $83,881 1-year post-index costs (mean): $82,642	Index admission costs included LEA for trauma (3/14), and costs for traumatic versus non-traumatic LEA were not reported separately. Post-index admission costs did not include traumatic LEA individuals.

Standard deviation or 95% confidence interval were not uniformly reported, and missing values reflect non-reporting in the study cited. All reported data is in costs (rather than charges or other non-cost expenditure metrics.)

CLTI=chronic limb threatening ischemia; USD=United States dollars; WIfI=Society for Vascular Surgery’s Wound, Ischemia, and Foot Infection classification system; FY=fiscal year; VHA=Veteran’s Health Administration; BKA=below knee (transtibial) amputation; TMA=transmetatarsal amputation; AKA=above knee (transfemoral) amputation

The cumulative costs of limb preservation and minor or major LEA in persons with diabetes and PAD are enormous, increasing, and under-studied [7,27]. Differences in study populations, variability in definitions and source of cost data, and changing epidemiology make it difficult to evaluate trends in LEA-related expenditures over time. A study of the burden of diabetes-related LEA in the Veteran’s Health Administration reported a 20% increase in inflation-adjusted costs between 2004 and 2010, with cumulative expenditure for LEA admissions exceeding $200 million in 2010 (2010 USD) [27]. A 2015 analysis of emergency department encounters for diabetic foot complications found the annual costs of diabetes-related emergent LEA admissions exceeded $2 billion (2014 USD)[22]. Using the most recently available data on LEA incidence, the total annual US health expenditure for diabetic LEA ranges from approximately $1.7 billion to $5.4 billion for minor LEA, and from $3.7 billion to $5.6 billion for major LEA, with broad ranges reflecting the significant variation in reported amputation costs [1,2,8,28].

Indirect costs (defined here as costs related to lost productivity, missed work, unemployment, and early disability or mortality) of diabetic foot disease, PAD, and major LEA are more challenging to measure. Diabetes-related LEA is among the top 10 causes of global disability, and estimates of the disability burden of diabetes and PAD-related limb complications have risen significantly in the past three decades [29]. Data on the indirect costs of LEA are lacking, but a retrospective analysis of the economic burden of diabetic foot ulcers found annual absenteeism and disability costs of $6,311±$9,288 (2012 USD) among privately insured adults, which is equal to twice the indirect costs associated with diabetes alone in matched controls [19]. Studies on functional outcomes report that up to 30% of persons who undergo any level (minor or major) LEA require assistance with activities of daily living [30], and approximately 35% do not return to full time work [31]. Of those patients with LEA who do return to work, 35% to 78% change occupation because of limitations related to amputation [31]. Low socioeconomic status, limited educational attainment, and lower-skill jobs predispose to job insecurity and are associated with more significant decline in employment status following LEA [31].

The association between LEA and premature mortality is well established, with an estimated five-year mortality following major LEA ranging from 40% to 90% [3]. Given the rising incidence of major LEA among young and middle-aged adults with diabetes, the lost work and premature mortality associated with LEA can be expected to continue to rise if there are not significant reductions in LEA.

Disparities in revascularization, LEA, and costs

LEA occur at disproportionately high rates among racial and ethnic minorities, persons with low socioeconomic status, and persons in geographically vulnerable areas (e.g., low-resource metropolitan neighborhoods and rural areas without proximity to specialist care) [5,6,10,32]. These disparities are inextricably linked to inequities in access to healthcare overall and, more specifically, to unequal access to limb preservation services [5,10,33].

Clinical risk factors for primary major LEA, including an infected, ischemic, or otherwise advanced-stage lower extremity wound at presentation [34], poor glycemic control, PAD, chronic kidney disease (especially end stage kidney disease and hemodialysis), and cardiovascular disease affect non-white and socioeconomically disadvantaged groups at higher rates than white or high income populations [8-10], and are also associated with higher healthcare costs irrespective of limb outcomes [35]. Controlling for clinical risk factors attenuates but does not erase the independent contributions of demographic factors toward risk for poor outcomes and higher costs in limb preservation [36].

Racial, ethnic, socioeconomic, and geographic disparities in limb preservation and LEA

Racial and ethnic minority patients with diabetes and/or PAD have a consistently lower rate of successful limb revascularization than their otherwise similar white counterparts [33,36]. In a large Nationwide Inpatient Sample analysis of admissions for lower extremity ischemia, non-white persons were 1.9 times more likely to undergo primary LEA without revascularization compared to white persons after controlling for gangrene, CLTI, income, and insurance status [37]. Medicaid insurance and low-income status in the same analysis were also associated with primary LEA. While lack of anatomic data in large retrospective claims-based studies is often cited as a major limitation of these studies, more granular studies of revascularization patterns confirm the differences. A systematic review and meta-analysis of outcomes following interventions for PAD found lower amputation-free survival, including higher rates of major LEA at 30 days and 1 year, following endovascular or open surgical bypass in Black and Hispanic patients compared to non-Hispanic white patients, despite similar rates of secondary patency [38]. Mortality was lower among non-white patients compared to non-Hispanic whites in this study despite higher rates of major LEA, suggesting that the cumulative burden of comorbidities cannot fully explain differences in limb outcomes.

Disparities in limb salvage are further evidenced by two to four times higher rates of any LEA and more proximal LEA in Black, Native, Hispanic, and other minority populations compared to non-Hispanic whites in the US [33,39-41]. Several studies examining limb outcomes in adults with diabetes and/or CLTI have demonstrated that risks conferred by minority race and ethnicity overlap with risk conferred by geographic disadvantage, defined either by neighborhood or community deprivation or rurality, and this has been proposed as an explanation for racial and ethnic outcomes disparities. A recent analysis of 14,816 cases of LEA in the Vascular Quality Initiative found that a higher distressed community index was independently associated with higher rates of above-knee LEA, but non-white race remained a risk factor for minor or major LEA even when controlling for community disadvantage [33,39,40]. In a large Medicare sample, significantly higher rates of major LEA were observed in Black persons with diabetes, persons living in disadvantaged neighborhoods, and persons living in rural counties relative to their white, non-disadvantaged, and urban-living counterparts, respectively [42]. The latter study was additionally notable for the finding that Black rural adults had risks of major LEA and death that exceeded the additive risk of Black race and rural residency status, suggesting that membership in multiple disadvantaged groups magnifies poor limb salvage outcomes [42].

Geographic differences in revascularization and LEA by neighborhood, state, region, and rurality are closely related to racial and socioeconomic disparities [33,43]. Neighboring hospital referral regions have been found to have similar rates of LEA [44]. LEA are less common in regions with a lower prevalence of diabetes, and more common in regions with lower socioeconomic status, more Black patients, and a higher prevalence of diabetic foot ulcers [44]. An analysis of regional variation in LEA among Medicare beneficiaries between 2010 and 2018 found that ZIP codes with >75% Black residents had rates of LEA in the uppermost quartile [33]; many urban ZIP codes in close proximity to large tertiary academic centers were included in this top-quartile LEA group despite the protective effects of tertiary academic care in other analyses of LEA risk [37,45]. These data suggest that geographic proximity alone is not sufficient to guarantee access to high quality limb preservation care.

Limited access to the healthcare system among patients of lower socioeconomic status almost certainly contributes to later presentation and greater severity of disease [46], but the association of socioeconomic status and race/ethnicity with low revascularization rates and high rates of LEA persists even after adjusting for these factors [41,47]. Non-white race was associated with a 2.4-fold risk of LEA in a retrospective review of consecutive admissions for diabetic foot ulcer in adjacent public, private, and Veterans Administration hospitals when controlling for care setting, gangrene, and comorbidities [48]. In an analysis of 144,647 US veterans receiving care through the Veterans Administration between 2003-2014, Black race was associated with greater LEA risk than white race within each SES stratum after controlling for comorbidities, PAD severity, and medication use [20].

Cost and healthcare utilization disparities in limb preservation and LEA

Despite scant literature focusing directly on disparities in costs, there are demonstrated differences in short- and long-term healthcare expenditures related to lower extremity revascularization and LEA that are worth examining (Table 3). One study that analyzed hospital costs for lower extremity revascularization among adults in Maryland from 2009 to 2015 found significantly higher costs for Black and other non-white adults compared to non-Hispanic white adults regardless of revascularization strategy, and higher costs for adults with Medicaid insurance, compared to private insurance or Medicare, for open surgical revascularization (Figure 1)[49]. Elevated costs were partially driven by higher likelihood of intensive care unit admission and prolonged hospital stay among non-white adults, even controlling for comorbidities. Another analysis of 225,180 National Inpatient Sample admissions for limb ischemia between 2012 and 2018 found similarly higher risk-adjusted and hospital-adjusted costs for non-white compared to non-Hispanic white adults: in multivariable models, non-white race was associated with $4,810 (95% CI $3,280, $6350) excess hospital costs per person, and higher costs in the non-white sample persisted across major LEA, endovascular intervention, and open intervention treatment groups (Tables 2 and 3) [50]. Non-white race was associated with significantly longer time to treatment and higher odds of primary major LEA, the latter of which was associated with excess costs of $17,800 (95% CI $16,049–$19,580) per hospitalization compared to limb preservation. Delays in care and disparities in major LEA contributed to significant excess costs of treating non-white patients [50].

Table 3.

Reported cost disparities in limb preservation and major lower extremity amputation (LEA)

Costs of limb preservation vs LEA	Cost methodology	Cost disparities	Additional notable findings
Ramanan et al. 2017 [25] Single-center, retrospective analysis of consecutive CLTI-related admissions (n=157 limbs) to a multidisciplinary limb preservation service from 2013-2014.	Inpatient hospitalization costs for index admission and total 1 year hospital costs from date of index admission, not including outpatient costs. The cost of preserved vs all limbs was reported but no specific cost reported for LEA. Reported in USD, not standardized to a single year.	WIfI 2 Index hospitalization costs preserved limbs vs all limbs (mean): $25,233 vs $26,794 1-year costs preserved limbs vs all limbs (mean): $30,588 vs $37,267 WIfI 3 Index hospitalization costs preserved limbs vs all limbs (mean): $25,644 vs $25,522 1-year costs preserved limbs vs all limbs (mean): $29,430 vs $29,092 WIfI 4 Index hospitalization costs preserved limbs vs all limbs (mean): $35,881vs $34,055 1-year costs preserved limbs vs all limbs (mean): $50,581 vs $52,729	No major LEA in WIfI 1 limbs within 1 year from index admission, thus no difference in all limbs versus preserved limb cost. Mortality was 12% at mean follow up of 270-300 days, but was not reported by WIfI stage, and costs were not adjusted based on follow up duration or survival.
Mustapha et al. 2018[17] Adults (n=72,199) enrolled in Medicare FFS with an incident CLTI diagnosis in 2011. Costs compared in a propensity-matched group (n=9,942) undergoing endovascular LER, open surgical LER, or major LEA as first procedural treatment between 2011-2015	Costs (inpatient, outpatient, and physician costs) from incident CLTI diagnosis through 4 years following diagnosis. Costs reported in 2016 USD and standardized to person year of follow-up/survival.	Costs per person year (mean) Surgical LER: $49,200 Endovascular LER: $49,700 Major LEA: $55,700	The difference between person year costs was significant at p<0.001 (major LEA compared to each LER). Black race was associated with primary major LEA [aOR (95% CI) 1.60 (1.44,1.79)] and mortality [1.05 (1.00, 1.11)] compared to other non-white race/ethnicity (reference). Black vs white race not directly compared, and cost differences by race/ethnicity not examined.
Gandjian et al. 2021[50] Retrospective review of all NIS admissions (n=225,000) for ALI from 2012-2018. PAD present in >50% of white and non-white groups, diabetes present in 29% and 60% of white and non-white groups, respectively.	Index hospitalization costs, adjusted for comorbidities and clinical and hospital-level variables, reported in 2017 USD.	Excess adjusted costs of major LEA compared to any revascularization [mean (95% CI) +$17,800 ($16,049, $19,580)	Non-white race was associated with significantly longer time to revascularization, increased odds of primary major LEA, and increased odds of LEA after revascularization compared to white race.
Cost disparities in limb preservation and/or LEA	Cost methodology	Cost disparities	Additional notable findings
Nejim et al. 2018 [49] Adults (n=17,186) undergoing inpatient LER for PAD/CLTI (48.6% with comorbid diabetes) in the state of Maryland from 2009-2015.	Index hospitalization costs, outcomes primarily reported as cost difference rather than absolute costs. Reported in 2016 USD.	Higher costs [mean (95% CI)] associated with Black race compared to white race Surgical LER cost excess: $6,092 ($4,682, $7,501) Endovascular LER cost excess: $2,642 ($1,574, $3,711) “Other” non-white race/ethnicity compared to white race: Surgical LER cost excess: $3,324 ($437, $6,212) Endovascular LER cost excess: $4,124 ($2,091, $6,157) Cost difference [mean (95% CI)] associated with Medicaid compared to Medicare insurance Surgical LER cost excess: $4,325 ($1,441, $7,209) Cost difference associated with commercial insurance compared to Medicare: Open LER cost reduction: −$4,755 (−$6,304, −$3,207) Endovascular LER cost reduction: −$3,345 (−$4,700, −$1,991)	Higher costs and more frequent ICU admission and 30-day readmission were found despite risk-adjustment Black race associated with higher costs across every patient complexity level Implementation of a capped per-capita payment model in the state of MD (2014) curbed overall costs but not disparities in costs or outcomes
Gandjian et al. 2021 [50] See above.	Index hospitalization costs, adjusted for comorbidities and clinical and hospital-level variables, reported in 2017 USD.	Excess costs associated with non-white race compared to white race [mean cost difference (95% CI)] All admissions: +$4,810 ($3,280, $6,350) Primary amputation: +$5,087 ($152, $10,022) Angioplasty or stenting: +$4,424 ($2,199, $6,649 Catheter-directed thrombolysis: +$4,899 ($2,252, $7,546) Open thrombectomy: +$6,058 ($1,967, $10,148) Open bypass, replacement, or repair: +$7,991 ($4,598, $11,385)	Non-white race was associated with significantly longer time from admission to revascularization, and increased odds of primary or any major LEA compared to white race.

Standard deviation or 95% confidence interval were not uniformly reported, and missing values reflect non-reporting in the study cited. All reported data is in costs (rather than charges or other non-cost expenditure metrics.)

LEA=lower extremity amputation; CLTI=chronic limb threatening ischemia; USD=United States dollars; WIfI=Society for Vascular Surgery’s Wound, Ischemia, and Foot Infection classification system; LER=lower extremity revascularization; FFS=Fee for Service; NIS=National Inpatient Sample; ALI=acute limb ischemia; PAD=peripheral artery disease; ALI=acute limb ischemia; VHA=Veteran’s Health Administration; BKA=below knee (transtibial) amputation; TMA=transmetatarsal amputation; AKA=above knee (transfemoral) amputation.

Figure 1.

Racial and Ethnic Disparities in Mean Hospital Costs for Open and Endovascular Revascularization Procedures, Maryland 2009 to 2015 Footnote: Adapted from Nejim et al. Ann Vas Surg. 2018 [49].

The lifetime cost of LEA is up to 3 times as high as the cost of limb preservation over a patient’s lifetime after considering rehospitalizations, post-acute care, and prosthesis-related costs, regardless of the etiology of limb loss [51]. This is true despite the high mortality and limited life expectancy associated with major LEA. Older cost analyses found excess lifetime costs of $509,275 for LEA compared to $163,282 for limb preservation (both 2002 USD) [51]. Modern cost studies in patients with CLTI undergoing limb preservation versus amputation have demonstrated similar results, with higher direct costs per patient year for patients who underwent primary amputation (mean costs $55,700 per year) relative to both endovascular or surgical revascularization (mean costs $49,700 and $49,200 per year, respectively, 2016 USD) (Table 3) [17].

The cost and healthcare burden related to LEA are concentrated in public “safety net” hospitals and resource-poor communities [52]. An analysis of LEA within the state of Texas found a small number of urban safety net hospitals perform almost half of the state’s diabetes-related LEA, serve largely under-insured or uninsured populations, and demonstrate more surgically and medically complex caseload than the statewide average [52]. A single-center analysis of short- and long-term costs associated with LEA performed at an urban community hospital found that incident LEA was associated with an average of 2.7 readmissions, more than 30 days of total inpatient hospital length of stay, high rates of re-amputation, and mean per-patient costs of $114,000 over a 2-year follow-up period (2013-2014 USD) [53]. Cumulative costs for patients undergoing LEA over the 2-year study period exceeded $10 million (2013-2014 USD), of which only 46% was reimbursed. Reimbursed costs were paid almost exclusively by Medicare and Medicaid (56% and 41%, respectively), reflecting the burden on older and underinsured adults.

Mitigating costs and outcome disparities in limb preservation

Equitable and affordable access to high quality care is a necessary foundation for efforts to reduce disparities in limb preservation [10,54]. Expanding health insurance and establishing multidisciplinary limb preservation programs are two cost-effective paradigms to attenuate disparities in LEA through improved healthcare access.

Racial and ethnic minority groups comprise the vast majority of uninsured or government-insured persons in the US [55]. Non-white Americans are significantly more likely than socioeconomically similar white persons to be uninsured, and are less likely to receive recommended preventive care for diabetes [55]. Subsidized preventive services through expanded public insurance access have been proposed to bridge this gap in care. Passage of the Affordable Care Act led to the narrowing of racial and ethnic disparities in uninsured status between 2010 and 2016. This was particularly notable among Hispanic adults, for whom the uninsured rate dropped by 30% after the Affordable Care Act was enacted [55]. While these changes in insurance status did not coincide with a population-level reduction in major LEA among minority adults, they did coincide with a shift toward increasing minor and decreasing major LEA rates among Hispanic and Native adults [43]. Specific data on care received by newly insured minority adults is lacking, but use of more minor LEA has been previously associated with improved limb salvage efforts [2].

Results from the Translating Research into Action for Diabetes (TRIAD) Study, a large multicenter cohort study evaluating structural barriers to care, found that racial and ethnic minorities were 40% less likely than non-Hispanic whites to develop a diabetic foot ulcer, and had similar risk of LEA, when continuously enrolled in a managed care plan that placed an emphasis on preventive care and limited out of pocket costs [56].

Multidisciplinary care, limb preservation, and cost

Multidisciplinary lower extremity wound and limb preservation programs have been extensively described elsewhere, and are increasingly considered standard of care for persons with diabetes and peripheral artery disease [57]. The components of these programs vary widely, but several systematic reviews have reported that multidisciplinary limb care reduces the risk major LEA by 10% to 80% in this population, with one meta-analysis reporting approximately 30% reduction in risk of major LEA among adults with diabetic foot ulcers after the implementation of multidisciplinary programs that include podiatry [57,58]. Notably, these improved outcomes are achieved with similar or lower costs per wound episode compared to traditional care models [18,59,60]. Many centers with multidisciplinary limb preservation programs have demonstrated similar outcomes in guideline directed medical therapy, wound healing, ulcer recurrence, rehospitalization and LEA for white and non-white persons [25,60]. Ultimately, expanding access to multidisciplinary limb preservation has the potential to address both cost and outcome disparities [61].

Conclusions

Drivers of high individual and health system-wide costs of LEA in the US are multifactorial, but disparities in access to care, limb salvage practices, and major LEA contribute significantly to the economic burden of diabetes and PAD. Persons with multiple disadvantaged characteristics (minority race or ethnicity, low socioeconomic status, resource-poor geographic location, and limited or no health insurance) experience compounded risk for poor outcomes and high healthcare costs.

At present, the inequities in outcomes and costs of diabetes-related lower extremity disease lead to a disproportionate burden on safety net hospitals and vulnerable communities where resources are limited and access to preventive care is poor. It is important to acknowledge that systematic inequalities in limb preservation result in enormous excess resource utilization by local and national health systems, and that these resources could be more effectively used to promote prevention if they were not urgently needed to address impending limb loss or treat the health consequences of LEA.

Although reducing major LEA is among the health indicators included in US National Healthy People 2030 goals [12], reimbursements for complex limb care and federal funding for research related to diabetic foot complications and PAD do not reflect the immense cost, morbidity, and mortality of these diseases [4]. Multidisciplinary care for patients with diabetes and PAD should be incentivized as a mechanism to achieve better clinical outcomes, mitigate disparities in limb preservation, and reduce excess expenditures related to preventable LEA. Expansion of subsidized health insurance coverage, funding for high-quality research, implementation of targeted screening and surveillance interventions, and outreach strategies to improve access to care in high-risk communities are urgently needed. Mitigating disparities in limb care has the potential to produce sizeable reductions in the overall incidence and economic burden of major LEA and should be a practice, policy, and funding priority.

Highlights.

• Treatment of lower extremity disease in diabetes and peripheral artery disease conservatively represents an estimated $60 billion in annual US healthcare spending.

• Costs of limb preservation are higher in non-white adults and uninsured or underinsured adults, in rural care settings, and in geographic regions with a high density of racial and ethnic minority populations.

• Non-white race, Hispanic ethnicity, low socioeconomic status, and lack of access to care based on geography or insurance status are associated with higher rates of primary and secondary major lower extremity amputation (LEA) in persons with diabetes and/or peripheral artery disease (PAD).

• Major LEA is more expensive and less cost-efficient than limb preservation by both short- and long-term cost metrics.

• Disparities in treatment patterns, including higher rates of LEA among minority and socioeconomically disadvantaged populations, contribute to enormous excess healthcare expenditures.

Abbreviations:

AKA

above knee (transfemoral) amputation

BKA

below knee (transtibial) amputation

CLTI

chronic limb-threatening ischemia

LEA

lower extremity amputation; unless specified, this refers to both minor and major non-traumatic lower extremity amputation

PAD

peripheral artery disease

SVS WIfI

Society for Vascular Surgery Wound, Ischemia, and Foot Infection Classification

TMA

transmetatarsal amputation

USD

United States dollars