Abstract
Objective:
The aim of this study was to evaluate inpatient health care utilization in the year preceding major lower extremity amputation (MLLA) among patients with chronic limb-threatening ischemia (CLTI), focusing on differences by race, ethnicity, and rurality using a large, multi-state inpatient dataset.
Methods:
A retrospective cohort study of Healthcare Cost and Utilization Project State Inpatient Databases for five states was conducted. Hospitalizations for Black, Hispanic, and White patients aged ≥40 years with a present-on-admission diagnosis of peripheral artery disease (PAD) or CLTI who underwent MLLA from 2017 to 2019 were included, and health care utilization in the year preceding their MLLA hospitalization was assessed. Differences in prior hospitalizations, minor lower limb amputations, and revascularization procedures were compared across race and/or ethnicity and rurality.
Results:
The cohort comprised 21,233 people who underwent MLLA. Most patients (>70%) underwent MLLA without prior revascularization or minor lower limb amputation, irrespective or race and/or ethnicity and rurality. Over one-half of all MLLA patients were hospitalized at least twice in the year preceding (once for PAD/CLTI specifically). Non-rural Black and Hispanic patients were more likely to be hospitalized in the year prior, with one-half of non-rural Black patients experiencing three or more hospitalizations in the year before amputation.
Conclusions:
Significant racial and/or ethnic and non-rural/rural disparities exist in inpatient preamputation health care utilization among patients with PAD. Regardless of race and/or ethnicity or rurality, most underwent MLLA without prior limb-sparing care. Findings underscore the need for targeted primary prevention interventions to reduce repeated hospitalizations and preventable amputations.
Keywords: Amputation disparities, Chronic limb-threatening ischemia, Health care utilization, Peripheral artery disease, Racial disparities
Chronic limb-threatening ischemia (CLTI) affects over 2 million adults in the United States and represents the most advanced form of peripheral artery disease (PAD).1,2 Initial management focuses on medical and lifestyle optimization in addition to wound and podiatric care when necessary. In patients with severe ischemia, presenting with ischemic rest pain or nonhealing wounds, either endovascular or surgical revascularization is indicated. When this multi-disciplinary care, including revascularization, is unsuccessful, such as in the setting of persistent infection or tissue loss, major lower extremity amputation (MLLA), defined as amputation above the ankle, may be necessary. MLLA carries a poor long-term prognosis, with an estimated 31% 4-year survival rate, highlighting the importance of timely PAD diagnosis and intervention3
In response to the high morbidity and mortality associated with MLLA, national quality initiatives have aimed to improve limb preservation.4 Despite these efforts, MLLA remains common, with 15% to 20% of patients with CLTI undergoing amputation within 1 year of diagnosis.2 More concerning, nearly one-half of patients with CLTI receive no vascular intervention prior to amputation.5 Racial and/or ethnic disparities in PAD and CLTI care are well-documented, with Black and Hispanic patients experiencing higher amputation rates and lower rates of revascularization compared with White patients6,7 Rural patients also face reduced access to vascular specialists and diagnostic resources, contributing to delays in diagnosis and treatment.8 Contributing factors include lower socioeconomic status, limited specialty care availability, and systemic barriers to timely diagnosis.9
Although disparities in revascularization and amputation rates are well-documented, less is known about health care utilization patterns in the year preceding MLLA across race and/or ethnicity and geography. Specifically, the frequency and timing of prior hospitalizations, minor amputations, and revascularization attempts remain poorly characterized. To help fill this gap, this study aimed to evaluate preamputation health care utilization across the intersection of race and/or ethnicity and rurality using a large, all-payer multi-state inpatient dataset.
METHODS
Data collection.
This study used publicly available, deidentified data and was exempt from the full Institutional Review Board review. A retrospective analysis was performed using the 2016 to 2019 Agency for Healthcare Research and Quality (AHRQ) Healthcare Cost and Utilization Project State Inpatient Databases (HCUP SIDs) for Florida, Georgia, Maryland, Mississippi, and New York. These states were selected because their SIDs include unique patient identifiers allowing longitudinal tracking of patients across facilities within each state, five-digit patient ZIP codes enabling geographic analysis, and relatively diverse racial and ethnic populations to support evaluation of sociodemographic differences.
The HCUP SIDs capture only inpatient hospitalizations and procedures; outpatient, office-based laboratory, and ambulatory surgical interventions are not included. Accordingly, all utilization and procedural estimates in this analysis reflect inpatient encounters only.
Cohort definition.
All hospitalizations of patients aged 40 years and older with a present-on-admission (POA) diagnosis of PAD or CLTI between January 1, 2017, and December 31, 2019, were identified. Hospitalizations were excluded if the patient was not a resident of the admitting state, had missing patient or hospital identifiers, or had a POA diagnosis of lower limb trauma, lower limb malignancy, pressure ulcers, or venous ulcers. Additional exclusions included patients admitted through trauma centers and those with a hospitalization for any of the exclusion diagnoses within the 365 days preceding the index admission. Diagnosis classification was based on physician-coded hospital discharge diagnoses. International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) codes were used to define inclusion criteria, exclusion criteria, and study outcomes (Supplementary Table I, online only).
For this analysis, Black, Hispanic, and White patients who underwent MLLA during a hospital stay between 2017 and 2019 were included. MLLA proportions were calculated relative to the total number of PAD/CLTI hospitalizations within each race- and rurality-stratified subgroup.
Race and/or ethnicity and rurality.
In HCUP SIDs, race and/or ethnicity are reported as a single metric (Asian or Pacific Islander, Black. Hispanic. Native American, White, and other). Rurality was determined using 2010 Rural-Urban Commuting Area codes, with ZIP codes classified as small town or rural areas (codes 7-10) categorized as rural.10
Health care utilization prior to MLLA.
Data from 2016 to 2019 were used to capture inpatient health care utilization in the 365 days preceding the hospitalization for MLLA. These included: (1) any prior hospitalization for any cause; (2) prior minor lower extremity amputation (below the ankle); and (3) prior lower extremity revascularization procedures, including both endovascular and open surgical interventions. Timing of each secondary event was categorized relative to the index MLLA hospitalization using the following intervals: <30 days, 31 to 60 days, 61 to 90 days, 91 to 120 days, 121 to 180 days, and 181 to 365 days prior.
Patient and neighborhood characteristics.
Patient characteristics of interest included patient age at MLLA hospitalization, sex, chronic conditions, infection history, admission type, transfer status, and whether they were admitted with a diagnosis of CLTI. Chronic conditions and infection history were defined based on diagnoses present on admission during the MLLA hospitalization or during any hospitalization in the prior 365 days. ICD-10-CM codes used to identify these conditions have been previously published.9
Over 40 neighborhood-level sociodemographic characteristics and health care access measures were evaluated by linking the HCUP SIDs to the AHRQ Social Determinants of Health (SDOH) Databases using patient ZIP codes. Sociodemographic variables, such as population density, racial and ethnic composition, and socioeconomic indicators, were measured at the ZIP code level. Health care access variables, including distance to health care facilities, hospital bed density, and physician availability, were measured at the county level. Data in the AHRQ SDOH Database are derived from the American Community Survey, Area Health Resource Files, County Health rankings, and Center for Medicare and Medicaid Provider of Services files. To allow for neighborhood changes over time, hospitalizations were linked to the AHRQ SDOH file for the previous calendar year.
Statistical analysis.
All data analyses were conducted in SAS version 9.4 (SAS Institute Inc). Descriptive statistics were used to summarize clinical characteristics, health care utilization patterns, and neighborhood-level sociodemographic factors. Categorical variables were reported as counts and percentages. Continuous variables were summarized as means with standard deviations (SDs). Descriptive analyses were stratified by race and/or ethnicity and rurality. Additional stratification was performed based on prior health care utilization history, including presence or absence of prior revascularization, hospitalization, or minor amputation.
No inferential statistical tests were performed due to the large sample size, which would likely produce statistically significant results even for clinically negligible differences. All analyses were descriptive and intended to characterize care patterns and population distributions among patients with MLLAs.
RESULTS
Among 1,577,061 total hospitalizations with a POA diagnosis of PAD/CLTI, 20,172 MLLA hospitalizations among Black, Hispanic, and White patients were included (Supplementary Table II, online only; Fig 1). Stratified patient demographics and clinical characteristics are reported in Table I.
Fig 1.
Cohort selection diagram. CLTI, Chronic limb-threatening ischemia; MLLA, major lower extremity amputation; PAD, peripheral artery disease.
Table I.
Characteristics of patients with major lower extremity amputation (MLLA) stratified by race and/or ethnicity and rurality
| Rural Black | Rural Hispanic | Rural White | Non-rural Black | Non-rural Hispanic | Non-rural White | |
|---|---|---|---|---|---|---|
| Unique patients/MLLA | 542 | 39 | 698 | 7424 | 2093 | 9376 |
| Age, years | 65 (11.7) | 58 (11.7) | 64 (12.1) | 65 (12.0) | 66 (12.4) | 65 (11.9) |
| Male | 340 (62.7) | 28 (71.8) | 494 (70.8) | 4522 (60.9) | 1415 (67.6) | 6611 (70.5) |
| Insurance type | ||||||
| Medicare | 395 (73.0) | 20 (51.3) | 465 (66.7) | 5188 (69.9) | 1400 (66.9) | 6413 (68.4) |
| Medicaid | 67 (12.4) | <11 | 86 (12.3) | 1240 (16.7) | 416 (19.9) | 1072 (11.4) |
| Private insurance | 46 (8.5) | <11 | 90 (12.9) | 629 (8.5) | 172 (8.2) | 1311 (14.0) |
| Self-pay | 11 (2.0) | <11 | 33 (4.7) | 163 (2.2) | 68 (3.3) | 313 (3.3) |
| Other | 22 (4.1) | <11 | 23 (3.3) | 200 (2.7) | 37 (1.8) | 266 (2.8) |
| Emergency admission | 254 (46.9) | 22 (56.4) | 356 (51.0) | 5102 (68.8) | 1673 (79.9) | 6087 (64.9) |
| Transferred in | 82 (15.2) | <11 | 125 (18.0) | 941 (12.8) | 208 (10.0) | 1116 (11.9) |
| Admitted with CLTI | 459 (84.7) | 33 (84.6) | 554 (79.4) | 6368 (85.8) | 1769 (84.5) | 7357 (78.5) |
| Chronic conditionsa | ||||||
| Cancer | ||||||
| Metastatic | <11 | 0 | 14 (2.0) | 98 (1.3) | 13 (0.6) | 110 (1.2) |
| Non-metastaticb | 29 (5.4) | 0 | 28 (4.0) | 245 (3.3) | 54 (2.58) | 323 (3.4) |
| Cerebrovascular disease | 84 (15.5) | <11 | 73 (10.5) | 1114 (15.0) | 260 (12.4) | 1109 (11.8) |
| Chronic pulmonary disease | 124 (22.9) | <11 | 238 (34.1) | 1732 (23.2) | 547 (26.1) | 3307 (35.3) |
| Congestive heart failure | 158 (29.2) | <11 | 230 (33.0) | 2642 (35.6) | 798 (38.1) | 3221 (34.4) |
| Coronary artery disease | 211 (38.9) | <11 | 344 (49.3) | 3337 (45.0) | 1102 (52.7) | 5046 (53.8) |
| Dementia | 84 (15.5) | <11 | 69 (9.9) | 1097 (14.8) | 323 (15.4) | 930 (9.9) |
| Diabetes | ||||||
| With complications | 336 (62.0) | 27 (69.2) | 453 (64.9) | 5348 (72.0) | 1652 (78.9) | 6129 (65.4) |
| Without complications | 73 (13.5) | <11 | 78 (11.2) | 748 (10.1) | 206 (9.8) | 852 (9.1) |
| Foot deformity | 13 (2.4) | 0 | 61 (8.7) | 354 (4.8) | 120 (5.7) | 869 (9.3) |
| Hemiplegia/paraplegia | 11 (2.0) | <11 | <11 | 179 (2.4) | 40 (1.9) | 150 (1.6) |
| HIV/AIDS | <11 | <11 | <11 | 97 (1.3) | <11 | 19 (0.2) |
| Hypercholesteremia | 304 (56.1) | 22 (56.4) | 458 (65.6) | 4737 (63.8) | 1391 (66.5) | 6259 (66.8) |
| Hypertension | 325 (60.0) | 23 (59.0) | 439 (62.9) | 4157 (56.0) | 1230 (58.8) | 5947 (63.4) |
| Liver disease | ||||||
| With complications | <11 | <11 | <11 | 59 (0.8) | 23 (1.1) | 115 (1.2) |
| Without complications | <11 | <11 | 19 (2.7) | 102 (1.4) | 36 (1.7) | 166 (1.8) |
| Myocardial infarction | 53 (9.8) | <11 | 114 (16.3) | 1155 (15.6) | 399 (19.1) | 1860 (19.8) |
| Peptic ulcer disease | 22 (4.1) | <11 | 19 (2.7) | 264 (3.6) | 74 (3.5) | 325 (3.5) |
| Renal disease | 285 (52.6) | 18 (46.2) | 274 (39.3) | 4320 (58.2) | 1128 (53.9) | 4102 (43.8) |
| Rheumatologic disease | 20 (3.7) | <11 | 30 (4.3) | 207 (2.8) | 77 (3.7) | 386 (4.1) |
| Admitted with infection | ||||||
| Superficial | 108 (19.9) | 15 (38.5) | 200 (28.7) | 1890 (25.5) | 711 (34.0) | 3197 (34.1) |
| Deep | 266 (49.1) | 24 (61.5) | 376 (53.9) | 4300 (57.9) | 1318 (63.0) | 5523 (58.9) |
| Gangrene | 308 (56.8) | 23 (59.0) | 353 (50.6) | 4494 (60.5) | 1290 (61.6) | 4343 (46.3) |
| History of infectionc | ||||||
| Superficial | 104 (19.2) | 12 (30.8) | 200 (28.7) | 1929 (26.0) | 722 (34.5) | 3174 (33.9) |
| Deep | 175 (32.3) | 17 (43.6) | 249 (35.7) | 3179 (42.8) | 1016 (48.5) | 3919 (41.8) |
| Gangrene | 155 (28.6) | 18 (46.2) | 166 (23.8) | 2427 (33.3) | 800 (38.2) | 2306 (24.6) |
| Patients with MLLA per state | ||||||
| Florida | 39 (7.2) | 8 (20.5) | 98 (14.0) | 1856 (25.0) | 1084 (51.8) | 3796 (40.5) |
| Georaia | 177 (32.7) | 12 (30.8) | 141 (20.2) | 2074 (27.9) | 157 (7.5) | 1656 (17.7) |
| Maryland | 10 (1.9) | 1 (2.6) | 24 (3.4) | 990 (13.3) | 47 (2.3) | 804 (8.6) |
| Mississippi | 303 (55.9) | 7 (18.0) | 191 (27.4) | 896 (12.1) | 8 (0.4) | 621 (6.6) |
| New York | 13 (2.4) | 11 (28.2) | 244 (35.0) | 1608 (21.7) | 797 (38.1) | 2499 (26.7) |
CLTI, Chronic limb-threatening ischemia.
Data are presented as number (%) or mean (standard deviation).
Diagnosis present on any hospitalization in the prior 365 days or flagged as present on admission (POA) on MLLA hospitalization.
Excluding skin cancer.
Diagnosis present on any hospitalization in the prior 365 days.
Hispanic patients had the highest rates of diabetes with complications regardless of rurality (78.9% non-rural; 69.2% rural). Heart conditions were most common among non-rural White and non-rural Black patients, with congestive heart failure present in 34.4% and 35.6%, coronary artery disease in 53.8% and 45.0%, and prior myocardial infarction in 19.8% and 15.6%, respectively. Deep infections were present in 49.1% to 63.0% of patients across all groups, highest in non-rural Hispanic patients. Gangrene was most common among non-rural Black (60.5%), non-rural Hispanic (61.6%), and rural Hispanic (59.0%) patients (Table I).
The mean age at the time of MLLA ranged from 58 years in rural Hispanic patients to 66 years in non-rural Hispanic patients. Male patients represented over 60% of all groups, highest among rural Hispanic (71.8%) and rural White (70.8%) patients. MLLA prevalence, stratified by race and/or ethnicity and rurality, is reported in Table I.
Inpatient health care utilization prior to MLLA
Minor lower limb amputations.
Overall, most patients had no history of prior minor amputation in the year before their MLLA (range, 78.0%-87A%). Rural and non-rural White patients were the least likely to have no prior minor lower limb amputation history (87.4% and 84.0%, respectively) (Table II). When evaluating timing of prior procedures, rural Hispanic patients had the highest proportion of minor amputations within 30 days of MLLA (12.8%), followed by non-rural Hispanic (10.5%) and rural Black patients (8.3%).
Table II.
Inpatient health care utilization in the year prior to major lower extremity amputation (MLLA), stratified by race and/or ethnicity and rurality
| Rural Black | Rural Hispanic | Rural White | Non-rural Black | Non-rural Hispanic | Non-rural White | |
|---|---|---|---|---|---|---|
| Minor lower limb amputation | ||||||
| <30 days prior | 45 (8.3) | <11 | 36 (5.2) | 594 (8.0) | 220 (10.5) | 630 (6.7) |
| 31-60 days prior | 45 (8.3) | <11 | 26 (3.7) | 659 (8.9) | 217 (10.4) | 677 (7.2) |
| 61-90 days prior | 30 (5.5) | <11 | 30 (4.3) | 493 (6.6) | 194 (9.3) | 479 (5.1) |
| 91-120 days prior | 26 (4.8) | <11 | 31 (4.4) | 408 (5.5) | 166 (7.9) | 430 (4.6) |
| 121-180 days prior | 31 (5.7) | <11 | 37 (5.3) | 494 (6.7) | 170 (8.1) | 554 (5.9) |
| 181-365 days prior | 30 (5.5) | <11 | 41 (5.9) | 603 (8.1) | 198 (9.5) | 710 (7.6) |
| No prior minor amputation | 449 (82.8) | 31 (79.5) | 610 (87.4) | 6030 (81.2) | 1632 (78.0) | 7872 (84.0) |
| Revascularization | ||||||
| <30 days prior | 48 (8.9) | <11 | 52 (7.5) | 817 (11.0) | 306 (14.6) | 933 (10.0) |
| 31-60 days prior | 48 (8.9) | <11 | 34 (4.9) | 760 (10.2) | 290 (13.9) | 858 (9.2) |
| 61-90 days prior | 32 (5.9) | <11 | 40 (5.7) | 610 (8.2) | 258 (12.3) | 701 (7.5) |
| 91-120 days prior | 20 (3.7) | <11 | 27 (3.9) | 523 (7.0) | 214 (10.2) | 579 (6.2) |
| 121-180 days prior | 37 (6.8) | <11 | 31 (4.4) | 662 (8.9) | 246 (11.8) | 738 (7.9) |
| 181-365 days prior | 27 (5.0) | <11 | 41 (5.9) | 792 (10.7) | 289 (13.8) | 946 (10.1) |
| No prior revascularization | 454 (83.8) | 30 (76.9) | 595 (85.2) | 5714 (77.0) | 1493 (71.3) | 7381 (78.7) |
| No. prior hospitalizations | 4.6 (8.4) | 3.1 (3.6) | 4.5 (7.2) | 6.1 (11.0) | 6.9 (11.1) | 5.9 (15.4) |
| No. prior PAD hospitalizations | 3.4 (6.6) | 2.7 (3.3) | 3.4 (6.1) | 4.6 (8.7) | 5.6 (9.5) | 4.7 (13.8) |
PAD, Peripheral artery disease.
Data are presented as number (%) or mean (standard deviation).
Revascularization.
The absence of any prior inpatient revascularization was highest among rural White (85.2%) and rural Black (83.8%) patients, and lowest among non-rural Hispanic patients (71.3%). In contrast, when examining timing among those revascularized inpatients, non-rural Hispanic patients had the highest proportion within 30 days of MLLA (14.6%), followed by non-rural Black (11.0%) and non-rural White (10.0%) patients (Table II).
Number of prior hospitalizations.
The mean number of hospitalizations in the year prior to MLLA was 6.9 ± 11.1 among non-rural Hispanic and 6.1 ± 11.0 among non-rural Black patients, compared with 5.9 among non-rural White patients (SD, 15.4). Mean number of hospitalizations among rural patients ranged from 3.1 in rural Hispanic patients to 4.6 in rural Black patients. The median number of PAD-specific hospitalizations ranged from 2.7 to 5.6 across all patient populations. PAD-specific hospital admission with CLTI exceeded 75% in all groups, highest in non-rural Black patients (85.8%).
Neighborhood sociodemographic and health care access characteristics by prior inpatient health care utilization
Patients who underwent a minor lower limb amputation in the year prior to MLLA were more likely to live in a ZIP code with a higher percentage of Black residents (21.2% vs 19.3%), more residents with less than a high school education (14.3% vs 13.6%), and more Supplemental Nutrition Assistance Program-receiving households (16.8% vs 16.0%), compared with those who did not undergo a minor amputation in the year prior. Additionally, patients without prior minor amputation were more likely to reside in ZIP codes with higher mean population densities (6950.7 vs 6855.7 persons/mile2, respectively) and a higher percentage of White residents (51.7% vs 49.1%).
Patients who underwent inpatient revascularization in the years prior were more likely to live in areas with higher population density (7367.3 vs 6809.4 persons/mile2), neighborhoods with a higher percentage of Hispanic residents (10.2% vs 7.8%), and a higher percentage of foreign-born residents (17.4% vs 14.3%). Patients with prior hospitalization lived in areas with slightly higher population density (7514.6 vs 6693.5 persons/mile2), compared with those who were not hospitalized in the year prior to MLLA.
Neighborhood socioeconomic indicators (unemployment, internet access, income inequality) and county-level health care access metrics (distance to nearest emergency department, hospital bed density, physician office availability, and cardiovascular specialist density) showed minimal differences across groups (Table III).
Table III.
Neighborhood sociodemographic and health care access characteristics, stratified by prior inpatient health care utilization
| MLLA with prior minor limb amputation |
MLLA without prior minor limb amputation |
MLLA with prior revascularization |
MLLA without prior revascularization |
MLLA with prior hospitalization |
MLLA without prior hospitalization |
|
|---|---|---|---|---|---|---|
| Neighborhood characteristicsa | ||||||
| Population density, persons per mile2 | 6855.7 (16599.2) | 6950.7 (16647.9) | 7367.3 (16756.1) | 6809.4 (16603.5) | 6693.5 (16127.1) | 7514.6 (17802.1) |
| Demographic makeup | ||||||
| American Indian/Alaska Native | 0.3 (1.5) | 0.2 (1.1) | 0.2 (0.6) | 0.3 (1.3) | 0.2 (1.2) | 0.3 (1.1) |
| Asian | 3.0 (4.9) | 3.2 (5.4) | 3.2 (5.1) | 3.1 (5.4) | 3.1 (5.1) | 3.3 (5.8) |
| Black/African American | 29.0 (25.4) | 27.9 (25.7) | 27.2 (25.1) | 28.4 (25.8) | 28.0 (25.6) | 28.4 (28.9) |
| Native Hawaiian/Pacific Islander | 0.0 (0.1) | 0.0 (0.1) | 0.0 (0.1) | 0.0 (0.1) | 0.0 (0.1) | 0.0 (0.1) |
| Hispanic | 16.3 (18.3) | 14.7 (17.6) | 17.4 (18.9) | 14.3 (17.3) | 15.4 (18.0) | 14.2 (16.9) |
| White | 49.1 (27.4) | 51.7 (28.2) | 49.7 (27.7) | 51.7 (28.1) | 51.0 (27.9) | 51.8 (28.4) |
| Foreign-born residents | 15.5 (14.5) | 14.7 (14.6) | 16.5 (14.9) | 14.4 (14.4) | 14.9 (14.6) | 14.6 (14.5) |
| Limited-English speaking households | 5.4 (7.7) | 5.0 (7.6) | 5.8 (8.1) | 4.8 (7.4) | 5.1 (7.7) | 4.8 (7.3) |
| Residents with <high school education | 15.2 (7.6) | 14.8 (7.6) | 14.9 (7.7) | 14.9 (7.6) | 14.9 (7.6) | 14.8 (7.6) |
| Residents with only high school education | 30.7 (7.6) | 30.5 (7.8) | 30.3 (7.6) | 30.6 (7.8) | 30.6 (7.7) | 30.4 (7.8) |
| Unemployed residents | 7.7 (4.0) | 7.5 (3.7) | 7.5 (3.7) | 7.5 (3.8) | 7.5 (3.7) | 7.5 (3.9) |
| GINI index of income inequality | 0.5 (0.1) | 0.5 (0.1) | 0.5 (0.1) | 0.5 (0.1) | 0.5 (0.1) | 0.5 (0.1) |
| Households receiving food stamps/SNAP | 18.6 (11.1) | 17.8 (10.8) | 18.3 (11.2) | 17.9 (10.8) | 18.1 (10.9) | 17.6 (10.8) |
| Households without internet access | 19.7 (10.2) | 19.9 (10.2) | 19.1 (9.9) | 20.0 (10.3) | 19.8 (10.2) | 20.0 (10.4) |
| Households with no vehicle available | 13.8 (16.3) | 13.9 (16.6) | 14.3 (17.0) | 13.8 (16.5) | 13.7 (16.2) | 14.2 (17.4) |
| Households with no health insurance | 12.2 (6.0) | 11.8 (6.0) | 12.1 (6.2) | 11.8 (5.9) | 11.9 (6.0) | 11.6 (6.0) |
| Grocery stores per 1000 populationb | 0.3 (0.2) | 0.3 (0.2) | 0.3 (0.2) | 0.3 (0.2) | 0.3 (0.2) | 0.3 (0.2) |
| Neighborhood health care accessb | ||||||
| Distance to nearest ED, miles | 3.7 (3.8) | 3.7 (3.9) | 3.6 (3.7) | 3.8 (3.9) | 3.7 (3.9) | 3.8 (4.0) |
| Distance to nearest health clinic, miles | 2.8 (3.3) | 2.9 (3.5) | 2.7 (3.2) | 2.9 (3.5) | 2.9 (3.4) | 3.0 (3.6) |
| Hospital beds per 1000 population | 3.3 (3.1) | 3.3 (2.3) | 3.3 (2.9) | 3.3 (2.4) | 3.3 (2.5) | 3.3 (2.4) |
| Physician offices per 1000 | 0.8 (0.3) | 0.8 (0.3) | 0.8 (0.3) | 0.8 (0.3) | 0.8 (0.3) | 0.8 (0.3) |
| Federally qualified health centers per 1000 | 0.0 (0.0) | 0.0 (0.1) | 0.0 (0.0) | 0.0 (0.1) | 0.0 (0.1) | 0.0 (0.1) |
| Cardiovascular disease specialists per 1000 | 0.1 (0.1) | 0.1 (0.1) | 0.1 (0.1) | 0.1 (0.1) | 0.1 (0.1) | 0.1 (0.1) |
ED, Emergency department; MLLA, major lower extremity amputation; SNAP, Supplemental Nutrition Assistance Program.
Data are presented as number (%) or mean (standard deviation).
Based on patient billing address; all neighborhood characteristics were measured at the ZIP code level unless otherwise noted.
Measured at the county level.
DISCUSSION
Our findings reveal gaps in preamputation inpatient health care utilization. Most patients undergoing MLLA had multiple hospitalizations in the preceding year, but these hospital encounters alone were not effective in preventing limb loss. Across all racial and/or ethnic and geographic populations, most patients underwent MLLA without antecedent inpatient revascularization or minor amputation in the year prior. Additionally, over one-half of MLLA hospitalizations were emergency admissions, indicating late presentation and missed inpatient and outpatient intervention opportunities. Although rural White patients had the highest proportion without any prior inpatient revascularization (85.2%), rural Black and Hispanic patients experienced the highest overall rates of amputation. This suggests that, although limited access to vascular care affects all rural populations, minority patients, both rural and non-rural, face compounded disparities stemming from delayed or fragmented intervention. This distinction between any prior revascularization and the timing of intervention helps reconcile these patterns: rural White patients most often lacked revascularization entirely, whereas non-rural Black and Hispanic patients more commonly underwent procedures within 30 days of MLLA, suggesting late or rescue-phase attempts at limb salvage.
These patterns reflect disparities in access, referral, and longitudinal management for vulnerable populations with PAD. Inferential statistical testing was intentionally not performed due to the large sample size, which would likely yield statistically significant results for clinically negligible differences. Our focus was instead on describing overall patterns of care utilization to identify broad gaps in preamputation management. By focusing on preamputation health care utilization rather than amputation incidence alone, this study offers new insight into upstream contributors to PAD-related limb loss.
Prior studies have consistently found racial and/or ethnic disparities in PAD management. A 2021 comprehensive review by Hackler et al synthesized data from multiple epidemiologic and clinical studies and found that Black and Hispanic patients are more likely to present with advanced PAD, less likely to undergo revascularization, and more likely to receive amputation compared with non-Hispanic White patients.11 More recently, a 2023 study by Alhuneafat et al, which analyzed over 1.3 million patients hospitalized with CLTI using the National Inpatient Sample, found that Black and Hispanic patients were significantly less likely than White patients to undergo any revascularization and more likely to undergo primary amputation.12 These disparities persist even after accounting for comorbidity burden and disease severity. For instance, Kalbaugh et al demonstrated that, among patients undergoing elective revascularization at comparable disease stages. Black and Hispanic patients experienced worse limb-related outcomes than White patients.13
Geographic disparities compound these inequities. Rural patients with CLTI are less likely to see vascular specialists, undergo diagnostic testing, or receive revascularization. In a retrospective cohort of over 73,000 Medicare beneficiaries with CLTI-related amputations, rural residence was associated with 18% lower odds of vascular specialist evaluation and 11% lower odds of revascularization compared with urban residence.14 Additionally, analysis of the Vascular Quality Initiative amputation dataset by Minc et al demonstrated that rural residence significantly increases the risk of primary amputation without attempted revascularization among non-White patients, but not among White patients, underscoring the intersectional impact of race and geography on amputation disparities.15 However, in our assessment of prior inpatient PAD management across the intersection of race and/or ethnicity and rurality, we found that rural patients (irrespective of race and/or ethnicity) consistently had lower health care utilization compared with their non-rural counterparts. Moreover, smaller differences in prior health care utilization were seen across race and/or ethnicity among non-rural patients.
Although our study used ZIP code level classification to define rurality, which enhances geographic specificity, limitations inherent in broader rural and urban classification systems may contribute to disparities identified in other studies. For example. Minc et al demonstrated significant heterogeneity in amputation risk within rural counties in West Virginia, identifying high-risk “pockets” at the ZIP code level that were not captured by broader county-level rurality designations.16 Similarly, a national analysis of metropolitan ZIP codes from 2010 to 2018 found that neighborhoods in the highest amputation quartile had significantly lower median household incomes, with each $10,000 decrease in income associated with a 4.4% higher amputation rate.8 These findings highlight that poor, predominantly minority communities in both rural and urban settings continue to face disproportionately high amputation rates, likely reflecting compounded barriers in access to vascular care and preventive services.
Although prior research has shown higher post-revascularization amputation rates among Black and Hispanic patients even after adjustment for disease severity, few have characterized preamputation utilization patterns across diverse populations.13 Our findings show that most patients, regardless of race and/or ethnicity or geography, underwent amputation without documented inpatient limb-salvaging intervention. Moreover, over one-half of all patients who underwent MLLA in our study were hospitalized at least twice in the year prior (at least once for PAD/CLTI specifically), highlighting missed opportunities for limb salvage, even among patients with repeated health care system contact. Similar patterns have been described in other high-risk populations. For example, in a cross-sectional study using Medicare claims data for patients with diabetic foot ulcers or infections. Black and Native American patients had higher rates of emergency department visits, outpatient encounters, and inpatient days, yet were significantly less likely to undergo revascularization and more likely to undergo amputation compared with White patients.17 In an analysis of Medicare claims data for beneficiaries with multiple hospital encounters related to PAD, rates of diagnostic arterial testing and revascularization remained low overall and were particularly reduced among racial and/or ethnic minority populations.5 These findings suggest that disparities may not be solely driven perhaps by access but also by deficiencies in care coordination, delayed referrals, and inconsistent inpatient diagnostic evaluation.
Although absolute differences in neighborhood sociodemographic and health care access characteristics were modest across groups in our study, small disparities across multiple social determinants may have a cumulative impact. In a retrospective cohort study using Medicare claims linked with the Distressed Communities Index, individuals residing in the most socioeconomically distressed areas had significantly higher odds of major lower extremity amputation, even when individual socioeconomic indicators appeared modest.18 Similarly, the United Kingdom’s Index of Multiple Deprivation, which incorporates 37 socioeconomic and environmental indicators, has been associated with increased risk of major amputation and death within 1 year of revascularization.19 Given our large sample size, even small percentage differences in exposure to socioeconomic risk factors may translate into meaningful population-level disparities in preamputation care and limb outcomes.
Despite the strengths of using a large, multi-state administrative dataset, this study has limitations. The HCUP SIDs capture only inpatient hospitalizations and lack outpatient claims or clinical registry data, limiting assessment of outpatient diagnostic testing, medication use, and office-based interventions. Care received in noninpatient settings (or outside of the included states) is not captured in these data and likely led to some underestimation of prior intervention use. Additionally, ZIP code information may reflect mailing addresses or PO boxes rather than patients’ actual residences, which could result in misclassification of rurality. We also lack detailed clinical information, such as anatomic disease severity (eg, Wound, Ischemia, and foot Infection [WIfI] staging). Coding variability and potential misclassification of diagnoses and procedures are inherent limitations of administrative data.
Although the analysis captured hospitalizations with PAD or CLTI as the admitting diagnosis, comorbid conditions such as congestive heart failure or end-stage renal disease may also contribute to higher overall hospitalization frequency and were not separately accounted for in these utilization estimates. Additionally, the analysis was restricted to five states with longitudinal identifiers, limiting generalizability. Our focus was instead on describing overall patterns of care utilization to identify broad gaps in preamputation management. Finally, the small number of rural patients, particularly rural Hispanic patients (n = 39), limits statistical power for subgroup analyses and warrants cautious interpretation.
CONCLUSIONS
Our findings highlight significant gaps in preamputation inpatient health care utilization across racial, ethnic, and geographic groups. Most patients underwent major lower extremity amputation without antecedent inpatient revascularization or minor amputation, underscoring widespread underuse of limb-preserving care prior to amputation. However, because this analysis was limited to inpatient encounters, outpatient and office-based vascular interventions were not captured; therefore, the reported rates of nonintervention represent the absence of inpatient procedures rather than total revascularization attempts. These descriptive patterns reveal persistent disparities in preamputation care utilization, particularly among rural and minority populations. Further work is needed to investigate the underlying causes of these disparities and to develop targeted interventions that improve access to timely, limb-salvaging care.
Supplementary Material
ARTICLE HIGHLIGHTS.
Type of Research: Multicenter retrospective cohort study using the Healthcare Cost and Utilization Project State Inpatient Databases
Key Findings: Among 21,233 patients who underwent major lower extremity amputation, over 70% had no prior inpatient revascularization or minor amputation. More than one-half were hospitalized at least twice in the preceding year. Non-rural Black and Hispanic patients had the highest proportion of prior hospitalizations, with 50% of non-rural Black patients hospitalized ≥3 times.
Take Home Message: Most patients undergo major lower extremity amputation without prior limb-salvaging intervention, with rural and minority populations disproportionately affected. These findings highlight missed opportunities for prevention and the need for targeted upstream vascular care strategies.
FUNDING
This research was supported in part by the Intramural Research Program of the National Institutes of Health (NIH). The contributions of the NIH author(s) were made as part of their official duties as NIH federal employees, are in compliance with agency policy requirements, and are considered Works of the United States Government. However, the findings and conclusions presented in this paper are those of the author(s) and do not necessarily reflect the views of the NIH or the United States Department of Health and Human Services.
DISCLOSURES
K.M. receives consulting fees from Medtronic and has research support from NovoNordisk Foundation. The remaining authors report no conflicts.
Footnotes
The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest.
Additional material for this article may be found online at www.jvascsurg.org.
REFERENCES
- 1.Conte MS, Bradbury AW, Kolh P, et al. Global vascular guidelines on the management of chronic limb-threatening ischemia. J Vasc Surg. 2019;69:36–125S.e40. 10.1016/j.jvs.2019.02.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Duff S, Mafilios MS, Bhounsule P, Hasegawa JT, The burden of critical limb ischemia: a review of recent literature. Vasc Health Risk Manag. 2019;15:187–208. 10.2147/VHRM.S209241 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Gooneratne T, Cassim R, Wijeyaratne M. Major lower limb amputation for chronic limb-threatening ischemia is associated with poor long-term survival: 4-year follow-up of a single-center experience. Ann Vasc Dis. 2024;17:234–240. 10.3400/avd.oa.23-00078 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Allison MA, Armstrong DG, Goodney PP, et al. Health disparities in peripheral artery disease: a scientific statement from the American Heart Association. Circulation. 2023;148:286–296. 10.1161/CIR.0000000000001153 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Goodney PP, Travis LL, Nallamothu BK, et al. Variation in the use of lower extremity vascular procedures for critical limb ischemia. Circ Cardiovasc Qual Outcomes. 2012;5:94–102. 10.1161/CIRCOUTCOMES.111.962233 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Holman KH, Henke PK, Dimick JB, Birkmeyer JD. Racial disparities in the use of revascularization before leg amputation in Medicare patients. J Vasc Surg. 2011;54:420–426.e1. 10.1016/j.jvs.2011.02.035 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Huber TS, Wang JG, Wheeler KG, et al. Impact of race on the treatment for peripheral arterial occlusive disease. J Vasc Surg. 1999;30:417–425. 10.1016/s0741-5214(99)70068-6 [DOI] [PubMed] [Google Scholar]
- 8.Fanaroff AC, Yang L, Nathan AS, et al. Geographic and socioeconomic disparities in major lower extremity amputation rates in metropolitan areas. J Am Heart Assoc. 2021;10:e021456. 10.1161/JAHA.121.021456 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Strassle PD, Minc SD, Kalbaugh CA. Donneyong MM, Ko JS, McGinigle KL. Disaggregating health differences and disparities with machine learning and observed-to-expected ratios: application to major lower limb amputation. Epidemiology. 2025;36:841–848. 10.1097/EDE.0000000000001892 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.United States Department of Agriculture, Economic Research Service. Rural-urban commuting area codes [Internet]. Washington (DC): USDA: 2010. Accessed June 29. 2025. https://www.ers.usda.gov/data-products/rural-urban-commuting-area-codes/. [Google Scholar]
- 11.Hackler EL, Hamburg NM, White Solaru KT. Racial and ethnic disparities in peripheral artery disease. Circ Res. 2021;128:1913–1926. 10.1161/CIRCRESAHA.121.318243 [DOI] [PubMed] [Google Scholar]
- 12.Alhuneafat L, Omar YA, Naser A, et al. Racial and ethnic disparities in peripheral vascular disease admissions using a nationally representative sample. Am J Cardiol. 2023;202:74–80. 10.1016/j.amjcard.2023.06.055 [DOI] [PubMed] [Google Scholar]
- 13.Kalbaugh CA, Witrick B, Sivaraj LB, et al. Non-hispanic black and hispanic patients have worse outcomes than white patients within similar stages of peripheral artery disease. J Am Heart Assoc. 2022;11:e023396. 10.1161/JAHA.121.023396 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ramadan OI, Yang L, Shultz K, et al. Racial, socioeconomic, and geographic disparities in preamputation vascular care for patients with chronic limb-threatening ischemia. Circ Cardiovasc Qual Outcomes. 2025;18:e010931. 10.1161/CIRCOUTCOMES.124.010931 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mine SD, Goodney PP, Misra R, Thibault D, Smith GS, Marone L. The effect of rurality on the risk of primary amputation is amplified by race. J Vasc Surg. 2020;72:1011–1017. 10.1016/j.jvs.2019.10.090 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Mine SD, Hendricks B, Misra R, et al. Geographic variation in amputation rates among patients with diabetes and/or peripheral arterial disease in the rural state of West Virginia identifies areas for improved care. J Vasc Surg. 2020;71:1708–1717.e5. 10.1016/j.jvs.2019.06.215 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Tan TW, Armstrong DG, Concha-Moore KC, et al. Association between race/ethnicity and the risk of amputation of lower extremities among medicare beneficiaries with diabetic foot ulcers and diabetic foot infections. BMJ Open Diabetes Res Care. 2020;8;e001328. 10.1136/bmjdrc-2020-001328 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Witrick B, Shi L, Mayo R, Hendricks B, Kalbaugh CA. The association between socioeconomic distress communities index and amputation among patients with peripheral artery disease. Front Cardiovasc Med. 2022;9:1021692. 10.3389/fcvm.2022.1021692 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Heikkilä K, Loftus IM, Waton S, Johal AS, Boyle JR, Cromwell DA. Association of neighbourhood deprivation with risks of major amputation and death following lower limb revascularisation. Atherosclerosis. 2020;306:11–14. 10.1016/j.atherosclerosis.2020.06.017 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.