Journal of Vascular Surgery
Volume 45, Issue 1 , Pages 55-59, January 2007

The adverse effects of race, insurance status, and low income on the rate of amputation in patients presenting with lower extremity ischemia

  • Mohammad H. Eslami, MD

      Affiliations

    • Division of Vascular Surgery, University of Massachusetts Medical School, Worcester, Mass
    • Corresponding Author InformationReprint requests: Mohammad H. Eslami, MD, FACS, Division of Vascular Surgery, University of Massachusetts Medical School, 55 N. Lake Ave, Worcester, MA 01655.
    • ,
  • Maksim Zayaruzny, MD

      Affiliations

    • Center for Outcomes Research, Worcester, Mass.
    • ,
  • Gordon A. Fitzgerald, PhD

      Affiliations

    • Center for Outcomes Research, Worcester, Mass.

Received 30 September 2005; accepted 16 September 2006.

Article Outline

Objectives

A consequence of delay in the diagnosis of peripheral vascular disease limb loss. This study was undertaken to determine the correlation of low socioeconomic status and race on the severity of ischemic presentation and the subsequent amputation rate.

Methods

Data from the Nationwide Inpatient Sample (NIS) from 1998 to 2002 on patients from urban hospitals with the diagnosis of lower extremity ischemia were evaluated. The population was divided into two groups: the amputation group (AMP) and lower extremity revascularization group (LER). Comorbidities, age, gender, race, ischemic gangrene at presentation, insurance status (no/noncommercial or commercial), and income status at admission were determined. These variables were compared using multivariate logistic regression analyses of the data for risk adjustment.

Results

Of 691,833 patients presenting with lower extremity ischemia, 363,193 underwent revascularization (66.3%) or amputation (33.7%). Univariate analysis correlated a statistically significant (P < .0001) higher rate of amputation and multivariate analysis associated significantly higher odds of amputation with the following variables: nonwhites (1.91, 95% confidence interval [CI], 1.65, 2.20), low-income bracket (1.41, 95% CI, 1.18, 1.60), and Medicare & Medicaid (1.81, 95% CI, 1.66, 1.97). Adjusting for other variables of statistical significance, multivariate regression analysis showed a statistically significant risk for amputation based on the nonteaching status of the institution (odds ratio [OR], 1.17, 95% CI, 1.08, 1.30).

Conclusions

Primary amputation was performed with a higher frequency on patients with lower extremity ischemia who were nonwhite, low income, and without commercial insurance. The observed advanced ischemia among these economically disadvantaged patients suggests a delayed diagnosis of peripheral vascular disease, probably due to lack of access to adequate primary care or vascular surgery providers, or both. Better education of the general population and primary care providers to the symptoms and consequences of PVD may reduce the amputation rate in this group.

 

Late presentation and delay in diagnosis and treatment of peripheral vascular disease (PVD) may lead to limb loss. In addition to the immense psychologic toll of amputation, the yearly cost of providing care to amputees is estimated at $4.3 billion.1 Furthermore, 25% to 60% of amputees never ambulate with a prosthesis and another 30% have reamputation or die ≤12 months of the initial procedure.1 Despite many advances in the fields of vascular and endovascular surgery during the past decade, an alarming increase in the number of amputations has been reported.2

Studies evaluating disparities of amputation rates among PVD patients with different sociodemographic variables have shown continuous differences in the per capita rate of amputation among different races,3, 4, 5, 6, 7 income levels,6, 8 and gender.4 Some of the racial disparities can be attributed to the higher prevalence of diabetes9 or difficult-to-control diabetes among nonwhite patients,10, 11, 12, 13 or to racial differences in the severity and characteristics of PVD at the time of presentation to a vascular surgeon.14, 15, 16, 17, 18 However, there is no evidence to suggest that inherent genetic differences between races or ethnicities can account for disparities of outcome such as limb salvage vs revascularization.19

Rather, many studies have shown that a decrease in amputation rates is possible with appropriate primary care, timely referral to vascular surgical centers, adequate foot and wound care, and aggressive limb arterial revascularization therapies, regardless of the presence of cormorbidites.20, 21, 22, 23, 24, 25 Reported inconsistencies in amputation rates therefore suggest that barriers to appropriate, effective, and timely care exist across the racial and sociodemographic spectrum. The purpose of this study was to determine how income level, insurance status, and race correlate with amputation rates in patients with lower extremity ischemia who present to urban hospitals.

Back to Article Outline

Methods 

Data source 

The Nationwide Inpatient Sample (NIS), a stratified, cross-sectional database resulting from a federal, state, and private industry partnership, includes approximately 20% of all nonfederal hospital discharges in the United States.26 The database, which includes information on 100% of discharges at participating hospitals, is stratified by geographic region, urban or rural hospital location, hospital teaching status, hospital ownership status, and hospital size. Data for this study were derived from the NIS database from 1998 to 2002, inclusive.

The databases were searched initially to identify all inpatient admissions for which the primary International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) indicated a diagnosis code for peripheral vascular disease (443.9), atherosclerosis, and lower extremity ischemia with or without gangrene (440.20, 21, 23, 24; 440-30, 31, 32). Patients presenting with traumatic ischemia or arterial embolism were excluded. To minimize the potential for bias owing to coding variations among the participating hospitals and to address the possibility of overlap between groups, 13 distinct codes were used to identify patients.

To assess the effect of race, the study was limited to urban hospitals, which generally have a more racially diverse patient population than rural settings (www.census.org).

Patients were divided into two groups by operative procedure. The first group included patients with primary ICD-9-CM procedure codes for lower extremity amputation (84.1, 84.10, 15-19) (AMP). Patients with toe amputations or partial foot or ankle amputations were excluded. The second group included those with a lower extremity revascularization (39.25, 39.29) (LER). Patients who had both procedures—revascularization and amputation—at the index admission were also excluded from this analysis.

Statistical analysis 

All statistical analyses were performed using SAS 9.2 (SAS Institute, Cary, NC) software. Because the NIS databases are stratified probability samples of US community hospitals, calculations were adjusted to reflect the survey sampling characteristics (probability weights, cluster sampling, and stratification). Probability weights were provided for each record in the NIS database. Stratification was performed for geographic region, teaching status, ownership, and hospital size; hospitals were used as cluster units.

Factors such as hospital type, initial presentation, comorbidities, age, gender, race, income status, and insurance payer mix were compared among the two procedure groups. Race was identified as a dichotomous variable: white vs nonwhite. Univariate analyses, using χ2 testing, were performed to compare theses variables among groups during the index admission, and statistically significant variables (P < .05) were analyzed further. The elimination multiple logistic regression analyses of statistically significant variables—controlling and adjusting risk across each study population—were performed, and (odds ratios [ORs], 1.17, 95% CI, 1.08, 1.30) were generated.

Back to Article Outline

Results 

According to NIS data, 691,833 patients presented with the primary diagnosis of lower extremity ischemia or gangrene to urban hospitals during the study period. Of these 363,193 underwent either lower extremity revascularization (240,740; 66.3%) or amputation (122,453; 33.7%) as the primary procedure during the index admission. Selected comorbidities are summarized in Table I. The average patient age was 70.6 years, with a statistically significant difference (P < .0001) in the mean ages of the LER (68.9, 95%) and AMP (73.2, 95%) groups. Although most patients were men (54.6%), more AMP patients were women (49% vs 43.9%, P < .001); conversely, more men were in the LER group (57% vs 51%, P < .001). Diabetes and chronic renal failure were more common among the AMP group, and chronic obstructive pulmonary disease, hypertension, and coronary artery disease were more common in the LER group.

Table I. Nationwide Inpatient Sample data (1998-2000) for patients presenting to urban hospitals with the diagnosis of lower extremity ischemia who subsequently had a procedure
CharacteristicsOverall⁎ (%)Amputation⁎ (%)Vascularization⁎ (%)
Weighted frequency (n)363,193122,453240,740
Average age (year)70.674.5⁎68.9⁎
Female sex45.449.7⁎43⁎
Race
White74.1⁎61.2⁎80⁎
Nonwhite25.9⁎38.8⁎20⁎
Diabetes42.149.7⁎38.9⁎
CRF2.74.8⁎1.7⁎
Hypertension50.340.7⁎54.3⁎
CHD10.89.3⁎11.5⁎
COPD1.81.1⁎2.2⁎
Ischem w/o gangrene66.323.6⁎84.3⁎
Ischemic gangrene37.276.4⁎15.6⁎
Hospital type
Teaching53.839.9⁎60.1⁎
Nonteaching46.245.2⁎54.8⁎

CRF, Chronic renal failure; CHD, chronic heart disease; COPD, chronic obstructive pulmonary disease.

⁎P < .001.

The racial composition of the groups is also summarized in Table I. At 80.0%, the LER group included a significantly higher percentage of white patients (P < .001) compared with the general population (74.1%) and the AMP (61.2%) group. Conversely, nonwhite patients were more prominently represented in the AMP group (38.8% vs 25.9% and 20% for LER; P < .001). Initial ischemic presentation significantly differed among the two procedure groups, as a significantly higher percentage of AMP patients (76.4%) presented with ischemic gangrene compared with the LER group (15.6%, P < .001).

More than half of the patients (53.6%) were treated at teaching hospitals, and 60.1% underwent revascularization. The reverse was true at nonteaching hospitals, where amputation rates were statistically significantly higher (45.2% vs 39.9%, P < .0001). Adjusting for other variables of statistical significance, multivariate regression analysis for OR estimates showed a statistically significant risk (OR = 1.17, 95% CI, 1.08, 1.30) for amputation based on the nonteaching status of the institution where patients were cared for (Table II).

Table II. Multiple logistic regression analysis results for amputation vs. lower extremity revascularization using Nationwide Inpatient Sample data, 1998-2000
Sociodemographic factorsOR95% Wald CL
Age >701.431.13,1.55
Female1.1391.02,1.11
Nonwhite1.901.75,2.05
Diabetes1.151.09,1.22
CRF1.391.19,1.64
Hypertension0.670.63,0.68
COPD0.610.52,0.71
Ischemic gangrene15.0412.14,14.32
Income ZIP⁎
<$24,9991.411.18,1.60
$25,000-$34,9991.221.10,1.36
$35,000-$44,9991.151.01,1.27
Medicaid1.911.65,2.20
Medicare1.811.66,1.97

OR, Odds ratio; CL, confidence limits; CRF, chronic renal failure; COPD, chronic obstructive pulmonary disease.

⁎Patient’s postal ZIP code was used as a proxy for socioeconomic status.

Socioeconomic factors such as patient income and the insurance payer mix were also found to be significantly different between the AMP and LER groups (Fig 1). The NIS-reported median income of each patient’s postal ZIP code was used as a proxy for socioeconomic status to divide the patient population was into three economic groups. Of these, 36.1% were in the highest income group (>$45,000/year), and 8% were in the lowest (<$24,999/year). The LER group, however, contained fewer patients from the lowest income category (6.5%) compared with the AMP group (10.6%), a statistically significant difference (P < .0001).

The two procedure subgroups also had disparate primary payer compositions. Among this patient population, Medicare was the most common provider (71.2%), and the second most common was a combination of private insurance or Health Maintenance Organization (private/HMO, 21.4%). Medicaid beneficiaries were 5.8% of the group; the remaining 1.6% was excluded because no insurance type was documented. Notably, the AMP group included 82.2% Medicare patients and 10.6% private/HMO patients, and the LER group had 66.4% Medicare and 26% private/HMO patients. Univariate analysis showed a statistically significant association between the procedure performed during the index admission and the primary payer (P < .0001). Similarly, Medicare recipients had the highest rate of amputation and lowest rate of revascularization (34.1% AMP, 65.1% LER) compared with Medicaid (31%, 68.9%) and private/HMO beneficiaries (14.8%, 85.2%; Fig 1).

Table II summarizes the multiple logistic regression analyses data. Independently, age >70, female sex, diabetes, chronic renal failure, hospital type, and ischemic gangrene were found to significantly increase the risk of undergoing amputation. Initial presentation of ischemic gangrene was the most important determinant, increasing the odds of having an amputation 15-fold. This analysis, which controls for the simultaneous effect of diabetes, gangrene, and chronic renal failure, indicates that the race and socioeconomic factors evaluated here (income and payer mix) independently increase the risk of having amputation. Low-income patients have a 1.41-fold risk of having an amputation compared with high-income patients (OR, 1.41, 95% CI, 1.18, 1.60). This odds ratio gradually decreases as patient income increases (Table II).

Compared with private/HMO patients, both Medicaid (1.9-fold) and Medicare (1.8-fold) patients were more likely to have amputations (Table II); among the latter two, Medicaid patients have the highest OR for amputation than Medicare recipients (P < .001). Although Medicare patient age (74.6 years) was significantly higher (P < .001) than Medicaid (59.2 years) and private/HMO patients (61.3 years), the Medicaid patients had the highest OR for amputation (P < .001).

Fig 2 shows that diabetes, an independent predictor of amputation, was more common among nonwhites (53.1% vs 37.1%, P < .001), Medicaid (45.9% vs 39.4% [private], or 44.4% [Medicare], P < .001 and P < .05, respectively), and the lowest-income group (P < .001). The diabetes rate decreases as income increases. As shown in Fig 3, nonwhite (50.7% vs 33%, P < .001) and low-income patients (46.3% vs 36.9%, P < .001) presented more commonly with ischemic gangrene, as did Medicaid patients compared with Medicare recipients (50.7% vs 23%, P < .001).

Back to Article Outline

Discussion 

The purpose of this study, which examined data collected in urban hospitals from 1998 to 2002, was to determine the correlation of low socioeconomic status and race on the severity of ischemic presentation and subsequent amputation rate in patients presenting with lower extremity ischemia. Comorbidities, including diabetes, chronic renal failure, and gangrene, were significantly associated with amputation. Age, gender, income level, and insurance status also were found to independently affect amputation rates.

Age as a risk factor for amputation may be due to the prevalence of PVD in older patients, reported by Selvin and Erlinger27 to increase from <7% in adults younger than age 70 to >14.5% in older adults. Increased amputation risk for women, also reported previously6, may reflect the inclusion of more women in this older, PVD-susceptible group. Underprivileged patients in this study, such as nonwhites, low-income, or Medicaid recipients, presented more commonly with diabetes, chronic renal failure, or gangrene, as reported previously.3, 6, 7, 9, 12, 13, 19, 21, 28 The observed advanced ischemia among these economically disadvantaged patients suggests a delayed diagnosis of PVD, probably owing to lack of access to adequate primary care or vascular surgery providers, or both, as well cultural distrust.30, 31, 32 Better education of the general population, particularly minorities, and primary care providers on the symptoms and consequences of PVD may reduce the amputation rate in this group.

Medicare and Medicaid patients exhibited a higher rate of amputation compared with patients with private insurance. Age and diabetes in these two groups may partly account for this trend, but delayed presentation appears to play a significant role, especially among Medicaid patients, who often receive care in the emergency department.33

Several shortcomings of this study stem from errors inherent in using NIS data, including coding errors,34 changes in sampling and weighting strategies owing to an expansion in the number of participating sites during the study, administrative errors resulting from changes in ICD-9 codes (partially addressed by using 13 different codes), and missing data on racial makeup. In addition, the inability to track prior or subsequent hospitalization/revascularization of individual patients may affect the findings because a single patient may at different time points be a member of one or another group. From that perspective, the study as designed provides a snapshot of the NIS database, which can be used to draw conclusions about population tendencies. The results obtained here are consistent with previous reports using different methodology and databases,3, 6, 7 suggesting that they accurately reflect the overall picture.

Back to Article Outline

Conclusion 

The data suggest there is a disparity of care provided to patients with PVD on the basis of race, income, and insurance status. Unlike more affluent white patients who undergo revascularization, those who are nonwhite, poor, and Medicaid recipients are more likely to undergo amputation. The elimination of barriers to appropriate medical care along with better education of the consequences of diabetes and the signs and symptoms of PVD may lead to earlier presentation and significantly decrease amputation rates.29

Back to Article Outline

Author contributions 


Conception and design: ME

Analysis and interpretation: ME, MZ, GF

Data collection: MZ

Writing the article: ME

Critical revision of the article: ME

Final approval of the article: ME

Statistical analysis: MZ, GF

Obtained funding: Not applicable

Overall responsibility: ME

Back to Article Outline

References 

  1. Dillingham TR, Pezzin LE, Shore AD. Reamputation, mortality, and health care costs among persons with dysvascular lower-limb amputations. Arch Phys Med Rehab. 2005;86:480–486
  2. Dillingham TR, Pezzin LE, MacKenzie EJ. Limb amputations and limb deficiencies; epidemiology and recent trends in the United States. South Med J. 2002;95:875–883
  3. Guadagnoli E, Ayanian JZ, Gibbons G, McNeil BJ, LoGerfo FW. The influence of race on the use of surgical procedures for treatment of peripheral vascular disease of the lower extremities. Arch Surg. 1995;130:381–386
  4. Tunis SR, Bass EB, Klag MJ, Steinberg EP. Variation in utilization of procedures for treatment of procedures for treatment of peripheral arterial disease: a look at patient characteristics. Arch Intern Med. 1993;153:991–998
  5. McBean AM, Gornick M. Differences by race in the rates of procedures performed in hospitals for medicare beneficiaries. Health Care Financ Rev. 1994;15:77–90
  6. Feinglass J, Kaushik S, Handel D, Kosifas A, Martin GJ, Pearce WH. Peripheral bypass surgery and amputation—northern Illinois demographics, 1993 to 1997. Arch Surg. 2000;135:75–80
  7. Feinglass J, Rucker-Whitaker C, Lindquist L, McCarthy WJ, Pearce WH. Racial differences in primary and repeat lower extremity amputation: results from a multihospital study. J Vasc Surg. 2005;41:823–829
  8. Gornick ME, Eggers PW, Reilly TW, et al. Effects of race and income on mortality and use of services among Medicare beneficiaries. New Engl J Med. 1996;355:791–799
  9. Harris MI, Flegal KM, Cowie CC, Eberhardt MS, Goldstein DE, et al. 1998 prevalence of diabetes, impaired fasting glucose, and impaired glucose tolerance in U.S. adults. Diabetes Care. 1998;21:518–524
  10. Erdman DM, Cook CB, Greenlund KJ, Giles WH, El-Kebbi I, Ryan GJ, et al. The impact of outpatient diabetes management on serum lipids in urban African-Americans with type 2 diabetes. Diabetes Care. 2002;25:9–15
  11. Chin MH, Zhang JS, Merrell K. Diabetes in the African-American Medicare population. Diabetes Care. 1998;21:1090–1095
  12. Resnick HE, Valsania P, Halter JB, Lin X. Differential effects of BMI on diabetes risk among black and white Americans. Diabetes Care. 1998;21:1828–1835
  13. Resnick HE, Valsania P, Phillips CL. Diabetes mellitus and nontraumatic lower extremity amputation in black and white Americans: the National Health and Nutrition Examination Survey epidemiologic follow-up study, 1971-1992. Arch Intern Med. 1999;159:2470–2475
  14. Newman AB, Siscovik DS, Manolio TA, Polak J, Fried LP, Bohari NO, et al. Ankle-arm index as a marker of atherosclerosis in the Cardiovascular Health Study. Circulation. 1993;88:837–845
  15. Collins TC, Petersen NJ, Suarez-Almazor M, Ashton CM. The prevalence of peripheral disease in a racially diverse population. Arch Intern Med. 2003;163:1469–1474
  16. Rucker-Whitaker C, Feinglass J, Pearce WH. Explaining racial variation in lower extremity amputation. Arch Surg. 2003;138:1347–1351
  17. Sidawy AN, Schweitzer EJ, Neville RF, Alexander EP, Temeck BK, Curry KM. Race as a factor in the severity of infragenicular occlusive disease: study of an urban hospital population. J Vasc Surg. 1990;11:536–543
  18. Toursarkissian B, Jones WT, D’Ayala MD, Shireman PK, Harrison A, Schoolfield J, et al. Does the efficacy of dorsalis pedis artery bypasses among diabetic patients of different ethnic backgrounds?. Vasc Endovasc Surg. 2002;36:207–212
  19. Pearce N, Foliaki S, Spoele A, Cunnigham C. Genetics, race, ethnicity, and health. BMJ. 2004;328:1070–1072
  20. Reiber GE, Recoraro RE, Koepsell TD. Risk factors for amputation in patients with diabetes: a case-control study. Ann Intern Med. 1992;117:97–105
  21. Larsson J, pelqvist J, Agrdh CD, Stenstrom A. Decreasing incidence of major amputation in diabetic patients: a consequence of multidisciplinary foot care team approach. Diabetes Med. 1995;12:770–776
  22. Bild DE, SElby JV, Sinnock P, Browner WS, Braveman P, Showstack JA. Lower extremity amputation in people with diabetes; epidemiology and prevention. Diabetes Care. 1989;12:24–31
  23. Weaver FM, Burdi MD, Pinzur MS. Outpatient foot care: correlation to amputation level. Foot Ankle Int. 1994;15:498–501
  24. Ebskov LB, Schroeder TV, Holsten PE. Epidemiology of leg amputation: the influence of vascular surgery. Br J Surg. 1994;81:1600–1603
  25. Hallett JW, Byrne J, Gayari MM, Listrup DM, Jacobsen SJ, Gary DT. Impact of arterial surgery and balloon angioplasty on amputation; a population-based study of 1155 procedures between 1973 and 1992. J Vasc Surg. 1997;25:29–38
  26. Nationwide Inpatient Sample (NIS). Agency for Health Care Policy and Research, Rockville, MD. Available from http://www.hcup-us.ahrq.gov/nisoverview.jsp. Accessed Sept 16, 2006.
  27. Selvin E, Erlinger TP. Prevalence and risk factors for peripheral arterial disease in the United States; results from the National Health and Nutrition Examination Survey, 1999-2000. Circulation. 2004;110:738–743
  28. Lavery LA, von Houtum WH, Armstrong DG, Harkless LB, Ashry HR, Walker SC. Mortality following lower extremity amputation in minorities with diabetes mellitus. Diabetes Res Clin Pract. 1997;37:41–47
  29. Davey Smith G, Neaton JD, Wentworth D, Stamler R, Stamler J MRFIT Research Group. Mortality differences between black and white men in the USA: Contribution of income and other risk factors among patients screened for MRFIT (Multiple Risk Factor Intervention Trial). Lancet. 1998;351:934–939
  30. Doeschner MP, Saver BG, Franks P, Fiscella K. Racial and ethnic disparities in perceptions of physician style and trust. Arch Fam Med. 2000;9:1156–1163
  31. Cooper-Patrick L, Gallo JJ, Gonzales JJ, Vu HT, Powe NR, Nelson C, et al. Race, gender, and partnership in the patient-physician relationship. JAMA. 1999;282:583–589
  32. Whittle J, Conigliaro J, Good CB, Joswiak M. Do patient preferences contribute to racial differences in cardiovascular procedure use?. J Gen Intern Med. 1997;12:267–273
  33. Sox CM, Burstin HR, Edwards RA, O’Neil AC, Brennan TA. Hospital admissions through the emergency department: does insurance status matter?. Am J Med. 1998;105:506–512
  34. Fisher ES, Whaley FS, Kurshat WM, Malenka DJ, Fleming C, Baron JA, et al. The accuracy of Medicare’s claims data: progress has been made, but problems remain. Am J Public Health. 1992;82:443–448

 Competition of interest: none.

PII: S0741-5214(06)01749-6

doi:10.1016/j.jvs.2006.09.044

Journal of Vascular Surgery
Volume 45, Issue 1 , Pages 55-59, January 2007

Article Tools

* Image Usage & Resolution