An evaluation of gender and racial disparity in the decision to treat surgically arterial disease
Article Outline
Objective
In 1994, our hospital reported a significant gender disparity in the treatment of peripheral artery disease (PAD). The objective of this study was to determine if this gender-based treatment disparity still persists after 15 years.
Methods
A retrospective review of patients with PAD and carotid artery disease based on vascular laboratory studies was performed from January 2006 to February 2008. PAD was identified by ankle-brachial index ≤ 0.9 or abnormal waveform. Treatable carotid artery disease was identified by symptomatic stenosis 60%-99% or asymptomatic stenosis 80%-99%. Patients with interventions before January 2006 were excluded. Demographics, risk factors, and interventions were recorded. Univariate and multivariate analyses were performed to identify risk factors and independent predictors of intervention.
Results
Of 2,313 peripheral artery studies, 592 patients with PAD and no prior intervention were identified. Sixty-five (21.7%) of 299 men and 47 (16.0%) of 293 women underwent angioplasty, stenting, endarterectomy, or bypass grafting. This difference was not significant (P = .077). However, by multivariate analysis of patients with critical limb ischemia, Caucasian race was an independent predictor of intervention (P = .010; odds ratio [OR] 3.363). Of 3,505 carotid duplex studies, 253 patients with treatable carotid artery disease and no prior intervention were identified. Seventy-eight (52.7%) of 148 men and 43 (41.0%) of 105 women underwent carotid endarterectomy (CEA) or stenting. This difference was not significant (P = .065). However, by multivariate analysis, Caucasian race was identified as an independent predictor of intervention (P = .015, OR 3.033). Insurance status was not a predictor of intervention in either the PAD (P = .70) or carotid artery disease cohort (P = .99).
Conclusion
Our data reveal that gender was not an independent predictor of intervention for PAD or carotid artery disease; however, Caucasian race independently predicted a greater likelihood of intervention in PAD patients with critical limb ischemia and in the carotid artery disease cohort. This study demonstrates the importance of performance assessments in uncovering unsuspected treatment disparities.
Gender and race have a significant role in determining treatment options. Male and Caucasian patients have been reported to receive more aggressive treatments in a range of cardiovascular diseases.1, 2, 3, 4, 5, 6, 7, 8 Studies have shown that African Americans are less likely to be revascularized for coronary and peripheral artery diseases (PAD).4, 5, 6, 7, 8, 9, 10, 11 Non-Caucasians have higher rates of amputations than Caucasians, which suggests that interventions on patients with critical limb ischemia may be delayed or forgone among non-Caucasians.4, 7, 8, 9, 10, 11 Additionally, population studies have shown a greater prevalence of PAD among African Americans versus Caucasians.12, 13, 14 African Americans, Hispanics, and women are less likely to receive adequate carotid artery evaluation following acute ischemic stroke, and non-Caucasians are less likely to undergo carotid endarterectomy (CEA) for stenosis.15, 16, 17 Carotid artery interventions have been utilized more for Caucasians than non-Caucasians, but it is unclear if the difference is more closely related to the varying medical risks and benefits across races, or if the problem is a public health issue. Many factors contribute to these disparities, including comorbidities, disease severity, and inherent tendencies toward atherosclerosis that differ by gender and race.18 Inequalities may also reflect that women and minorities have a misunderstanding of their treatment options, lack of access to care, or mistrust of the medical community when compared with men and Caucasians.19, 20, 21 Given these findings, it is important to uncover why gender and racial differences exist.
Determining reference points with performance reports is a crucial method that can help identify unwanted disparities and possibly reduce them.19, 22, 23 The existence of disparities is well documented, and Ayanian asserts that the current phase of treatment disparity research is focused on finding and eliminating unfair practices.19 In 1993, the National Institutes of Health Revitalization Act made it a requirement to include women and minority subjects in clinical trials, unless there was compelling evidence for their exclusion.24 Although there was a subsequent increase in the presence of women and minorities in research, both groups remain underrepresented. In 1994, an evaluation of gender disparity in the treatment of PAD was conducted at our institution and revealed a significant difference in surgical intervention rates between men and women even after controlling for disease severity and risk factors.25 Thus, the aim of this study was to track any progress, or lack thereof, toward more equitable treatment of PAD 15 years later. Our hypothesis was that a gender-based treatment disparity in PAD no longer exists because of our increased awareness of gender disparity. We also examined carotid artery disease for the first time and included racial demographics in our analysis. This study, as well as future evaluations of our practice, may serve as reference points that identify reasons for disparities.
Methods
Patient cohort
This study was approved by the Institutional Review Board at Northwestern University. Adults who underwent lower extremity arterial physiological studies or carotid artery duplex examinations in the vascular laboratory at Northwestern Memorial Hospital (NMH) between January 2006 and February 2008 comprised the population for this study. Blood flow studies from the laboratory were reviewed to identify patients with significant PAD or carotid artery disease. PAD was determined from laboratory records of the ankle-brachial index (ABI) and Doppler waveform analysis. The PAD cohort included all patients with an ABI less than 0.9 or an abnormal waveform. All patients with a history of intervention for PAD prior to 2006 were excluded from the analysis. Non-invasive studies are performed on all patients scheduled for lower extremity revascularization. Patients with studies from outside hospitals were only included in the analysis if they had vascular laboratory studies performed at our institution. Carotid artery disease was assessed from laboratory studies of carotid artery stenosis determined by duplex examination. Patients with carotid artery disease were identified by having greater than 60% stenosis with ipsilateral symptoms (cerebrovascular accident [CVA], transient ischemic attack [TIA], or amaurosis fugax), or greater than 80% stenosis without symptoms. All patients with a history of intervention for carotid artery disease prior to 2006 were excluded from the analysis.
Vascular laboratory studies
All vascular laboratory studies were performed by a registered vascular technologist and reviewed by a certified staff vascular surgeon (W.H.P.). The standard lower extremity noninvasive arterial physiologic study or carotid artery duplex examination was conducted and interpreted according to the Intersocietal Commission for the Accreditation of Vascular Laboratory (ICAVL) guidelines in the NMH ICAVL-accredited vascular laboratory. For measurements of the ABI, technologists used a Doppler probe to record the systolic and diastolic pressures of the brachial, posterior tibial, and dorsalis pedis arteries. Measurements were taken bilaterally whenever possible. Contraindications to measuring pressures were pain at the site of measurement or the absence of the appropriate limb. ABI was calculated by dividing the highest available lower extremity systolic pressure by the highest brachial pressure. Studies were deemed abnormal if the ABI was less than 0.9. Heavy calcification of blood vessels sometimes rendered the Doppler measurement invalid. Therefore, in the case of an invalid ABI, abnormal (non-triphasic) waveforms were considered to be indicative of PAD. For cases confirmed by ABI, severity of PAD was stratified into three classes: minimal disease (ABI, 0.5-0.9); moderate disease (ABI, 0.3-0.49); and severe disease (ABI < 0.3). The presence of rest pain, tissue loss, or gangrene anywhere on the lower extremities was defined as critical limb ischemia. For the carotid artery disease cohort, duplex examinations estimated the percentage of stenosis in the lumen of the internal and common carotid arteries bilaterally. Stenosis greater than 60% or 80% of the internal and/or common carotid arteries was the only determinant of disease in symptomatic and asymptomatic patients, respectively.
Interventions
Interventions for patients in the PAD cohort included angioplasty, stenting, endarterectomy, or bypass grafting. The interventions of patients in the carotid artery disease cohort included endarterectomy or angioplasty and stenting.
Patient demographics
The medical records of patients with documented PAD or carotid artery disease were retrospectively reviewed. There were 29 and 26 variables recorded for patients with PAD and carotid artery disease, respectively. These variables included demographics, comorbidities, severity of disease, and medications being taken at the time of the most recent pre-operative assessment in the Vascular Laboratory. Gender and race were reported by patients during their hospital registration. All patients who did not indicate Caucasian race were considered non-Caucasian.
Statistical analysis
Statistical analysis software (SAS Inc, Cary, NC) was used to analyze the data. In order to assess the decision to treat patients by surgical intervention, the PAD and carotid artery disease cohorts were separated into non-intervention and intervention groups for analysis. Variables were compared using chi-squared analysis or the Fisher exact test. Determination of risk factors and independent predictors of surgical intervention was performed with univariate and multiple logistics regression analyses. Differences between treatment groups were considered significant for P < .05. Certain demographic and clinical characteristics may have differed by gender or by race. Since the statistical significance of clinical factors influenced by gender or race may have been suppressed in univariate analysis, a multivariate analysis was performed that controlled for all variables included in univariate analysis.
Results
Peripheral artery disease
Between January 2006 and February 2008, 2,313 lower extremity arterial studies were performed in the vascular laboratory (Fig 1). Of those studies, 1,163 (50.3%) were identified as abnormal. After excluding patients with prior intervention, our PAD cohort for this study was 592 patients. In order to determine what factors influenced the decision to operate, the population was divided into non-intervention and intervention groups. Table I shows patient demographic data of the entire cohort as well as by treatment group; it also shows the statistically significant differences between non-intervention and intervention groups as determined by chi-squared analysis. In the PAD cohort, 480 (80.1%) of the patients received no intervention, while 112 (18.9%) underwent intervention, including angioplasty, stenting, endarterectomy, or bypass grafting.
Fig 1.
Patients in the peripheral arterial disease cohort by gender and intervention status.
Table I. Demographics of patients with peripheral artery disease
| Characteristics | All patients (n = 592) | No intervention (n = 480) | Intervention (n = 112) | P |
|---|---|---|---|---|
| Mean age (years ± standard deviation) | 70.1 ±12.0 | 70.2±12.1 | 69.5±11.7 | .611 |
| <50 | 39(6.6%) | 30(6.3%) | 9(8.0%) | |
| 50-65 | 137(23.1%) | 110(22.9%) | 27(24.1%) | |
| 65-75 | 167(28.2%) | 140(29.2%) | 27(24.1%) | |
| 75-85 | 188(31.8%) | 148(30.8%) | 40(35.7%) | |
| > 80 | 61(10.3%) | 52(10.8%) | 9(8.0%) | |
| Male gender | 299(50.5%) | 234(48.8%) | 65(58.0%) | .077 |
| Insurance | 581(97.8%) | 470(97.9%) | 111(99.1%) | .696 |
| Caucasian race | 336(56.8%) | 263(54.8%) | 73(65.2%) | .046 |
| Non-Caucasian race | 256(43.2%) | 217(45.2%) | 39(34.8%) | .046 |
| PAD | ||||
| ABI <0.9 | 441(74.5%) | 353(73.5%) | 88(78.6%) | <.001 |
| ABI <0.3 | 57(9.6%) | 37(7.7%) | 20(17.9%) | |
| ABI 0.3-0.49 | 108(18.2%) | 76(15.8%) | 32(28.6%) | |
| ABI 0.5-0.9 | 276(46.6%) | 240(50.0%) | 36(32.1%) | |
| Abnormal waveform only | 151(25.5%) | 127(26.5%) | 24(21.4%) | |
| Type I diabetes mellitus | 32(5.4%) | 25(5.2%) | 7(6.3%) | .664 |
| Type II diabetes mellitus | 231(39.0%) | 187(39.0%) | 44(39.3%) | .949 |
| Hypertension | 490(82.8%) | 392(81.7%) | 98(87.5%) | .152 |
| Arrhythmia | 121(20.4%) | 92(19.2%) | 29(25.9%) | .112 |
| Coronary artery disease | 300(50.7%) | 227(47.3%) | 73(65.2%) | <.001 |
| Myocardial infarction | 93(15.7%) | 75(15.6%) | 18(16.1%) | .907 |
| Congestive heart failure | 104(17.6%) | 80(16.7%) | 24(21.4%) | .233 |
| Coronary artery bypass graft | 122(20.1%) | 89(18.5%) | 33(29.5%) | .010 |
| Carotid artery disease | 144(24.3%) | 112(23.3%) | 32(28.6%) | .245 |
| Renal insufficiency | 91(15.4%) | 73(15.2%) | 18(16.1%) | .820 |
| Renal failure | 63(10.6%) | 47(9.8%) | 16(14.3%) | .165 |
| Current or past smoker | 346(58.4%) | 266(55.4%) | 80(71.4%) | .002 |
| Chronic obstructive pulmonary disease | 64(10.8%) | 47(9.8%) | 17(15.2%) | .098 |
| Hyperlipidemia | 361(61.0%) | 286(59.6%) | 75(67.0%) | .066 |
| Critical limb ischemia | 195(32.9%) | 136(28.3%) | 59(52.7%) | <.001 |
| Preoperative clinic visit | 338(57.1%) | 271(56.5%) | 67(59.8%) | .517 |
| Medications | ||||
| Aspirin | 338(57.1%) | 260(54.2%) | 78(69.6%) | .003 |
| Clopidogrel | 149(25.2%) | 98(20.4%) | 51(45.5%) | <.001 |
| Warfarin | 94(15.9%) | 73(15.2%) | 21(18.8%) | .356 |
| Statin | 360(60.8%) | 280(58.3%) | 80(71.4%) | .011 |
| ACE inhibitor | 256(43.2%) | 203(42.3%) | 53(47.3%) | .333 |
| Beta blocker | 331(55.9%) | 259(54.0%) | 72(64.3%) | .048 |
| Other anti-hypertensive | 260(43.2%) | 203(42.3%) | 57(50.9%) | .099 |
Of the entire PAD cohort, 50.5% were men and 49.5% were women. Of those who did not receive an intervention, 48.8% were men, 51.2% were women. Of those who received an intervention, 58.0% were men and 42.0% were women (Table I). The difference between intervention rates of men and women for PAD did not reach statistical significance by univariate analysis (P = .078; OR 1.454) or multivariate analysis (P = .157; OR 1.412). However, relatively fewer Caucasians were in the non-intervention group (263 [54.8%]) compared with the intervention group (73 [65.2%]; Fig 2). This difference was significant by univariate analysis (P = .047; OR 1.544; Table II), but not by multivariate analysis (P = .165; OR 1.432).
Fig 2.
Percentage of Caucasian and Non-Caucasian patients with PAD and carotid artery disease by intervention status. *Significant by univariate analysis (P = .047; OR 1.544); **Significant by multivariate analysis (P = .015; OR 3.033).
Table II. Logistics regression analysis of risk factors and independent predictors for intervention on patients with peripheral arterial disease
| Characteristics | Odds ratio | P |
|---|---|---|
| Univariate analysis | ||
| Caucasian race | 1.544 | .047 |
| Ankle-brachial index <0.3 | 2.603 | .001 |
| Ankle-brachial index 0.3-0.49 | 2.126 | .002 |
| Ankle-brachial index 0.5-0.9 | 0.474 | .001 |
| Coronary artery disease | 2.086 | .001 |
| Coronary artery bypass graft | 1.835 | .011 |
| Current or past smoker | 2.011 | .002 |
| Critical limb ischemia | 2.787 | <.001 |
| Aspirin | 1.941 | .003 |
| Clopidogrel | 3.259 | <.001 |
| Statin | 1.786 | .011 |
| Beta blocker | 1.536 | .048 |
| Multivariate analysis | ||
| Ankle-brachial index <0.3 | 3.052 | .005 |
| Ankle-brachial index 0.3-0.49 | 3.042 | .001 |
| Coronary artery disease | 1.954 | .020 |
| Current or past smoker | 2.036 | .007 |
| Hyperlipidemia | 1.719 | .034 |
| Critical limb ischemia | 3.910 | <.001 |
A sub-analysis was performed on patients with critical limb ischemia. The results of the multivariate analysis showed that Caucasian patients were far more likely to receive revascularization than non-Caucasian patients (P = .010; OR 3.363). There was no difference between men and women by the multivariate analysis of patients with critical limb ischemia. Table II shows statistically significant results from the univariate and multivariate analyses that were performed on all demographic variables. By univariate analysis, positive risk factors for intervention included coronary artery disease (CAD) (P = .001; OR 2.086), prior coronary artery bypass (P = .011; OR 1.835), history of smoking (P = .002; OR 2.011), critical limb ischemia (P < .001; OR 2.787), and usage of aspirin (P = .003; OR 1.941), clopidogrel (P < .001; OR 3.259), statins (P = .011; OR 1.786), and beta-blockers (P = .048; OR 1.536). The intervention group was notable for a greater proportion of severe (ABI < 0.3; 17.9% versus 7.7%; P = .001; OR 2.603) and moderate disease (ABI 0.3-0.49; 28.6% versus 15.8%; P = .002; OR 2.126), but a smaller proportion of minimal disease (ABI 0.5-0.9; P = .001; OR 0.474). Multivariate analysis revealed CAD (P = .020; OR 1.954), history of smoking (P = .007; OR 2.036), hyperlipidemia (P = .034; OR 1.719), critical limb ischemia (P < .001; OR 3.910), ABI < 0.3 (P = .005; OR 3.052), and ABI 0.30-0.49 (P = .001; OR 3.042) as independent predictors of intervention among this cohort (Table II).
Carotid artery disease
Between January 2006 and February 2008, 3,505 carotid artery duplex examinations were performed in the vascular laboratory (Fig 3). Of those studies, 2,498 (71.3%) were identified as abnormal (stenosis 1%-100%), among which 371 (14.9%) had significant disease (stenosis 60%-100%). The final population of the carotid artery disease cohort was limited to 253 patients who had treatable disease (stenosis < 100%) and did not have a history of a prior surgical intervention of their carotid artery stenosis. In order to determine what factors influenced the decision to operate, the population was divided into non-intervention and intervention groups. Table III shows the demographic data of the entire cohort as well as by treatment group; it also shows significant differences between the non-intervention and intervention groups as determined by chi-squared analysis. In the carotid artery disease cohort, 132 (52.2%) of the patients received no intervention, while 121 (47.8%) underwent intervention by stenting or carotid endarterectomy (CEA).
Fig 3.
Patients in the carotid artery disease cohort by symptoms, gender, and intervention status. Symptoms include ipsilateral stroke, transient ischemic attack, or amaurosis fugax.
Table III. Demographics of patients with carotid artery disease
| Characteristics | All (n = 253) | No intervention (n = 132) | Intervention (n = 121) | P |
|---|---|---|---|---|
| Mean age (years ± standard deviation) | 72.0±11.2 | 74.5±11.1 | 69.3±10.6 | .001 |
| <50 | 6(2.4%) | 2(1.5%) | 4(3.3%) | |
| 50-65 | 59(23.3%) | 23(17.4%) | 36(29.8%) | |
| 65-75 | 83(32.8%) | 42(31.8%) | 41(33.9%) | |
| 75-85 | 67(26.5%) | 34(25.8%) | 33(27.3%) | |
| >80 | 38(15.0%) | 31(23.5%) | 7(5.8%) | |
| Male gender | 148(58.5%) | 70(53.0%) | 78(64.5%) | .065 |
| Caucasian race | 200(79.1%) | 99(75.0%) | 101(83.5%) | .098 |
| Non-Caucasian race | 53(20.9%) | 33(25.0%) | 20(16.5%) | .098 |
| Insurance | 249(98.4%) | 130(98.5%) | 119(98.3%) | .999 |
| Type I diabetes mellitus | 14(5.5%) | 7(5.3%) | 7(5.8%) | .867 |
| Type II diabetes mellitus | 70(27.7%) | 40(30.3%) | 30(24.8%) | .328 |
| Hypertension | 205(81.0%) | 101(76.5%) | 104(86.0%) | .059 |
| Arrhythmia | 29(11.5%) | 14(10.6%) | 15(12.4%) | .655 |
| Coronary artery disease | 122(48.2%) | 68(51.5%) | 54(44.6%) | .274 |
| Myocardial infarction | 43(17.0%) | 26(19.7%) | 17(14.0%) | .232 |
| Congestive heart failure | 22(8.7%) | 12(9.1%) | 10(8.3%) | .816 |
| Coronary artery bypass graft | 68(26.9%) | 40(30.3%) | 28(23.1%) | .199 |
| Renal insufficiency | 24(9.5%) | 15(11.4%) | 9(7.4%) | .287 |
| Renal failure | 8(3.2%) | 7(5.3%) | 1(0.8%) | .068 |
| Current or past smoker | 103(40.7%) | 44(33.3%) | 59(48.8%) | .013 |
| Chronic obstructive pulmonary disease | 17(6.7%) | 5(3.8%) | 12(9.9%) | .052 |
| Hyperlipidemia | 181(71.5%) | 83(62.9%) | 98(81.0%) | .001 |
| Symptomatic | 54(21.3%) | 20(15.2%) | 34(28.1%) | .012 |
| Preoperative clinic visit | 187(73.9%) | 75(56.8%) | 112(92.6%) | <.001 |
| Medications | ||||
| Aspirin | 167(66.0%) | 62(47.0%) | 105(86.8%) | <.001 |
| Clopidogrel | 72(28.5%) | 28(21.2%) | 44(36.4%) | .008 |
| Warfarin | 25(9.9%) | 16(12.1%) | 9(7.4%) | .212 |
| Statin | 166(65.6%) | 79(59.8%) | 87(71.9%) | .044 |
| ACE inhibitor | 78(30.8%) | 40(30.3%) | 38(31.4%) | .850 |
| Beta blocker | 123(48.6%) | 65(49.2%) | 58(47.9%) | .835 |
| Other anti-hypertensive | 104(41.1%) | 49(37.1%) | 55(45.5%) | .178 |
Of the entire carotid artery disease cohort, 58.5% were men and 41.5% were women. Of those who did not receive an intervention, 53.0% were men and 47.0% were women. Of those who received an intervention, 64.5% were men and 35.5% were women (Table III). The difference between intervention rates of men and women for carotid artery disease did not reach statistical significance by univariate analysis (P = .066; OR 1.607) or multivariate analysis (P = .168; OR 1.631). However, further analysis revealed Caucasian race as an independent predictor of intervention. Relatively fewer Caucasians were in the non-intervention group (99 [75.0%]), while more were in the intervention group (101 [83.5%]; Fig 2), and this difference was significant by multivariate analysis (P = .015; OR 3.033; Table IV). As shown in Fig 3, when analyzed by symptoms, a more striking difference was observed between the sexes. Of 199 asymptomatic patients, 56.8% were men and 43.2% were women. Of the asymptomatic patients, 46.9% of men received an intervention, whereas only 39.5% of the women did. This difference did not reach statistical significance (P = .299). Of 54 symptomatic patients, 64.8% were men and 35.2% were women. Of the symptomatic patients, 71.4% of men received an intervention, whereas only 47.4% of women did. This large difference did not reach statistical significance (P = .080), most likely due to the relatively small sample size of patients with symptomatic disease (n = 54).
Table IV. Logistics regression analysis of risk factors and independent predictors for intervention on patients with carotid artery disease
| Characteristics | Odds ratio | P |
|---|---|---|
| Univariate analysis | ||
| Current or past smoker | 1.903 | .013 |
| Hyperlipidemia | 2.515 | .002 |
| Prior amaurosis fugax | 4.300 | .028 |
| Symptomatic disease | 2.188 | .013 |
| Preoperative clinic visit | 9.458 | <.001 |
| Aspirin | 7.410 | <.001 |
| Clopidogrel | 6.988 | .008 |
| Statin | 1.717 | .045 |
| Multivariate analysis | ||
| Age | 0.937 | <.001 |
| Caucasian race | 3.033 | .015 |
| Current or past smoker | 2.192 | .038 |
| Chronic obstructive pulmonary disease | 6.159 | .032 |
| Hyperlipidemia | 4.227 | <.001 |
| Prior transient ischemic attack | 0.289 | .034 |
| Preoperative clinic visit | 31.324 | <.001 |
Table IV shows statistically significant results from the univariate and multivariate analyses that were performed on all demographic variables. By univariate analysis, positive risk factors for intervention included history of smoking (P = .013; OR 1.903), hyperlipidemia (P = .002; OR 2.515), prior amaurosis fugax (P = .028; OR 4.300), symptomatic disease (P = .013; OR 2.188), preoperative clinic visit (P < .001; OR 9.458) and usage of aspirin (P < .001; OR 7.410), clopidogrel (P = .008; OR 6.988), and statins (P = .045; OR 1.717). Multivariate analysis also revealed age (P < .001; OR 0.937), history of smoking (P < .038; OR 2.192), chronic obstructive pulmonary disorder (P = .032; OR 6.159), hyperlipidemia (P < .001; OR 4.227), prior TIA (P = .034; OR 0.289), and a preoperative clinic visit (P < .001; OR 31.324) as independent predictors of intervention among this cohort (Table IV).
Discussion
Clinical research in the field of surgery tends to focus on the outcomes of successful operations. The decision to treat, however, can be impeded by public health or socioeconomic factors. This project was conceived as an acknowledgement of the efforts of past research that has shown that treatment is not always equitable, and—more importantly—that concerted efforts are required to ensure and uphold its fairness.19, 22 We hypothesized that, irrespective of medical history, women would be as likely as men to receive surgical intervention for PAD and carotid artery disease. Our research determined that gender was not a determinant of intervention, but surprisingly that Caucasian race was an independent predictor of carotid artery disease intervention. Caucasian race was also an independent predictor of revascularization in patients with critical limb ischemia. Thus, while we corrected the gender disparity identified in 1994, we have now identified a disparity between races.
There is no current evidence that suggests, given the same indications, that intervention rates for PAD or carotid artery disease should differ by gender or race. Many studies have confirmed the negative consequences of delayed intervention for PAD on African Americans.4, 7, 9, 10, 11 These data on limb loss show that minority patients with PAD should be considered as much as non-minorities for revascularization. As for carotid artery disease, in 2008 the Society for Vascular Surgery appointed a committee of experts to create clinical guidelines for treatment of carotid stenosis. The guidelines indicated that symptoms and degree of stenosis are proper indicators of intervention, but not gender or race.26 However, it should be noted that the risk-benefit ratio for CEA has been reported to be different between men and women; therefore, a treatment disparity may be warranted.27
Disparities have been identified in the treatment of PAD and carotid artery disease. Several studies have identified that minorities have higher amputation rates and are less likely to undergo revascularization as opposed to amputation.4, 7, 8 These differences are most common among African Americans with PAD. Feinglass et al found that amputation disparities can exist in hospitals with varying vascular surgery capacity and further suggested that socioeconomic inequality had a major influence on racial disparities.10, 11 Kennedy et al found that non-Caucasians had lower rates of initial CEA utilization and higher rates of in-hospital death and stroke.16 Such findings underline how disparities can skew complications toward minorities. Jha et al aggregated Medicare data in 158 referral areas to determine if racial disparities in intervention rates of nine different procedures abated.22 Over 10 years the disparities actually increased for most procedures, and in none of the procedures was disparity eliminated. It is clear that treatment disparities are prevalent and can lead to severe negative consequences in under-treated groups.
Many different etiologies account for treatment disparities. These include limited insurance coverage, patient mistrust of healthcare providers and hospitals, patient misunderstanding of disease severity and treatment options, physician bias, and a lack of access to adequate health care facilities and qualified personnel.4, 7, 8, 9, 10, 11, 19, 20, 21, 27, 28, 29, 30, 31, 32, 33, 34 The results of our gender analysis suggest that performance assessments and increased self-awareness of treatment trends can reduce inequities. Awareness of what etiologies account for disparities is necessary to mitigating them.
Census data have consistently shown that higher levels of poverty and lower rates of insurance exist in Hispanic and African American communities when compared with Caucasian communities.31 Our study showed that insurance was not predictive of intervention in our patient cohorts. However, we did not record different levels of coverage, including which patients may be underinsured. Unfortunately, underinsurance rates are difficult to measure, but they nonetheless serve as a significant barrier to healthcare access.33, 34 Patients with less coverage face higher costs for surgery and thus they are less likely to opt for intervention.29 Therefore, a lack of insurance and underinsurance both contribute to treatment disparities.
Patients' mistrust of healthcare and misunderstandings of disease severity can contribute to their disinclination to undergo surgery. Minorities and the poor often live in urban areas where disadvantages to accessing healthcare are greatest.29 Urban hospitals in these communities are overworked and are generally less able to provide adequate treatment for severely diseased patients.29, 30, 32 These settings can lead to hastening of disease progression, negatively affect patient attitudes, and reduce patient understanding of the consequences of delayed intervention.
The physician can have a positive or negative effect on treatment disparities. We recorded which patients had preoperative visits to our outpatient clinic in order to determine if consultation with one of our institution's vascular surgeons influenced the surgeon's decision to intervene. Patients who were seen in clinic but did not undergo intervention may have declined an option offered by their physician. Unfortunately, we were unable to determine who may have declined an offer for intervention. For the PAD cohort, being evaluated preoperatively in our clinic was not significantly associated with intervention. For the carotid artery disease cohort, in which race was identified as an independent predictor of intervention, a preoperative clinic evaluation was an independent predictor of intervention. Therefore, the barriers between minorities and intervention may be reinforced by not referring patients to the appropriate specialist. In sum, physicians may be able to reduce disparities by encouraging minority patients to understand all the options available to them.
Recent efforts that aim to increase the presence of minorities in healthcare professions offer a promising opportunity to encourage members of disadvantaged communities to take advantage of better treatments. Placing members of minority communities in positions to provide healthcare establishes a trust between provider and patient that has been largely absent from the minority medical experience in America. These efforts can also work to successfully reduce or eliminate treatment disparities.
As with any retrospective study, there were several limitations to this research. This was a single-institution study, and it may not reflect national trends. For PAD, we were not able to record the extent to which symptoms affected patients. The severity of disease was determined by the ABI, which cannot solely quantify the disruption and inconvenience that a patient experiences from claudication due to PAD. Furthermore, the medical records we reviewed relied on patients' reporting of their surgical and medical histories and may not have been completely accurate. Also complicating the analysis of our study was the sample size. While our PAD cohort was large (n = 592), the carotid artery disease cohort was smaller (n = 253). We observed significant trends in treatment disparity among gender for this cohort, yet it did not reach statistical significance. In particular, there was a seemingly large but insignificant difference in intervention rates between men and women with symptomatic carotid artery disease. A greater sample size may have clarified whether or not gender contributed to the treatment algorithm. It is also possible that some of the patients with disease identified at our hospital sought treatment at other institutions. There were 410 of 592 PAD patients and 93 of 253 patients with carotid artery disease who had only one vascular laboratory study and no intervention. These patients may have declined to be treated, were lost to follow-up, or may have sought treatment at outside institutions that we were unable to record. Lastly, we were unable to determine the number of patients lost to follow-up due to the retrospective nature of this study.
In conclusion, our data revealed no evidence of a gender disparity in the treatment of PAD or carotid artery disease. However, our data did identify the presence of a statistically significant difference in the treatment of Caucasians versus non-Caucasians for carotid artery disease and critical limb ischemia. When compared with data from 15 years ago, these findings are encouraging as they show a more equitable treatment of men and women. However, the findings are alarming because they reveal an unexpected racial disparity. Physicians can encourage patients with severe disease to seek treatment, especially minority patients who are faced with negative external factors like underinsurance or mistrust of healthcare systems. Thus, staying aware of disparities and the issues facing minority populations may work to improve equity.
Author contributions
References
- . Ambulatory hypercholesterolemia management in patients with atherosclerosis. J Gen Intern Med. 2005;20:123–130
- . Gender differences in vascular access in hemodialysis patients in the United States: Developing strategies for improving access outcome. Gend Med. 2007;4:193–204
- Racial disparities in access to renal transplantation – clinically appropriate or due to underuse or overuse?. N Engl J Med. 2000;343:1537–1544
- Impact of race on the treatment for peripheral arterial occlusive disease. J Vasc Surg. 1999;30:417–426
- . Race and sex differences in the management of coronary artery disease. Am Heart J. 2000;139:848–857
- . Racial variation in the use of coronary-revascularization procedures – are the differences real? (Do they matter?). N Engl J Med. 1997;336:480–486
- . The adverse effects of race, insurance status, and low income on the rate of amputation in patients presenting with lower extremity ischemia. J Vasc Surg. 2007;45:55–59
- Effects of race and income on mortality and use of services among medicare beneficiaries. N Engl J Med. 1996;335:791–799
- . Explaining racial variation in lower extremity amputation: a 5-year retrospective claims data and medical record review at an urban teaching hospital. Arch Surg. 2003;138:1347–1351
- . A census-based analysis of racial disparities in lower extremity amputation rates in Northern Illinois, 1987-2004. J Vasc Surg. 2008;47:1001–1007
- . Racial differences in primary and repeat lower extremity amputation: results from a multihospital study. J Vasc Surg. 2005;41:823–829
- . Incidence of peripheral vascular disease in women: is it different from that in men?. J Thorac Cardiovasc Surg. 2004;127:314–317
- . Arterial vascular disease in women. J Vasc Surg. 2007;46:1295–1302
- Lower extremity arterial disease assessed by ankle-brachial index in a middle-aged population of African Americans and Whites: the Atherosclerosis Risk in Communities (ARIC) Study. Am J Prev Med. 2005;29(5 Suppl 1):42–49
- The association of race and sex with the underuse of stroke prevention measures. J Stroke Cerebrovasc Dis. 2008;17:226–234
- . Elective and isolated carotid endarterectomy: health disparities in utilization and outcomes, but not readmission. J Natl Med Assoc. 2007;99:480–488
- Disparities in the treatment and outcomes of vascular disease in Hispanic patients. J Vasc Surg. 2007;46:971–978
- Intrinsic contribution of gender and ethnicity to normal ankle-brachial index values: The Multi-Ethnic Study of Atherosclerosis (MESA). J Vasc Surg. 2007;45:319–327
- . Determinants of racial and ethnic disparities in surgical care. World J Surg. 2008;32:509–515
- . Patient preferences and health disparities. JAMA. 2001;286:1506–1509
- . Gender differences in patient preferences may underlie differential utilization of elective surgery. Am J Med. 1997;102:524–530
- . Racial trends in the use of major procedures among the elderly. N Engl J Med. 2005;353:683–691
- . Gender disparities in managed care: it's time for action. Women's Health Issues. 2007;17:116–119
- . Adherence to federal guidelines for reporting of sex and race/ethnicity in clinical trials. J Womens Health (Larchmt). 2006;15:1123
- . Gender differences in interventional management of peripheral vascular disease: evidence from a blood flow laboratory population. Ann Vasc Surg. 1994;8:343–349
- Management of atherosclerotic carotid artery disease: clinical practice guidelines of the Society for Vascular Surgery. J Vasc Surg. 2008;48:480–486
- . Executive Committee for the Asymptomatic Carotid Atherosclerosis Study. JAMA. 1995;273:1421–1428
- . The effect of patients' preferences on racial differences in access to renal transplantation. N Engl J Med. 1999;341:1661–1669
- . Black and disadvantaged health, health reform, and the future (An American health dilemma, Volume II: Race, medicine, and health care in the United States: 1900-2000). In: New York: Routledge; 2002;p. 569–587
- . Getting political: racism and urban health. Am J Public Health. 2000;90:841–842
- . Income, poverty, and health insurance coverage in the United States: 2006. Washington, D.C: US Government Printing Office: U.S. Department of Commerce; 2007;
- . To mitigate, resist, or undo: addressing structural influences on the health of urban populations. Am J Public Health. 2000;90:867–872
- . Association between underinsurance and access to care among children with special health care needs in the United States. Pediatrics. 2005;116:1162–1169
- Underinsurance in primary care: a report from the State Networks of Colorado Ambulatory Practices and Partners (SNOCAP). J Am Board Fam Med. 2008;21:309–316
Competition of interest: none.
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 competition of interest.
PII: S0741-5214(09)01562-6
doi:10.1016/j.jvs.2009.07.089
Published by Elsevier Inc.
