Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2009 Dec 8.
Published in final edited form as: Am Heart J. 2005 Jun;149(6):1074–1081. doi: 10.1016/j.ahj.2004.08.018

Racial differences in reperfusion therapy use in patients hospitalized with myocardial infarction: A regional phenomenon

Saif S Rathore a,b, Paul D Frederick e, Nathan R Every e, Hal V Barron f, Harlan M Krumholz a,c,d,✉
PMCID: PMC2790272  NIHMSID: NIHMS155281  PMID: 15976791

Abstract

Background

Racial differences in reperfusion therapy use among patients hospitalized with myocardial infarction (MI) have been previously reported as national phenomenon. However, it is unclear whether racial differences in treatment vary by region.

Methods

Using data from the National Registry of Myocardial Infarction-2 and -3, a cohort of patients hospitalized with MI in the United States between 1994 and 2000, we sought to determine whether racial differences in reperfusion therapy use varied by geographic region in patients eligible for reperfusion therapy with no clinical contraindications to treatment (n = 204230).

Results

Black patients had lower crude rates of reperfusion therapy than white patients (66.5% vs 69.9%, −3.3% racial difference, 99% CI −4.4% to −2.2%) overall. However, racial differences in reperfusion therapy use varied by geographic region. Reperfusion therapy rates were similar for black patients and white patients in the Northeast (67.9% black vs 65.3% white, +2.7% racial difference, 99% CI −0.5% to 5.8%) and statistically comparable for patients in the Midwest (68.3% black vs 69.0% white, −0.7% racial difference, 99% CI −2.9% to 1.5%) and West (70.7% black vs 72.6% white, −1.9% racial difference, 99% CI −5.1% to 1.2%). Racial differences in reperfusion therapy use were greatest for patients hospitalized in the South (64.5% black vs 71.7% white, −7.1% racial difference, 99% CI −8.7% to −5.6%). Racial differences were reduced, but geographic variations in racial differences persisted after multivariable adjustment.

Conclusions

Lower rates of reperfusion therapy use among black patients with MI do not reflect a national pattern of racial differences in treatment, but a practice pattern predominantly attributable to the South.

Prior studies have reported that black patients with myocardial infarction (MI) have lower rates of reperfusion therapy use than white patients.1–7 Racial differences in treatment have been reported in a variety of settings and have persisted after adjustment for clinical characteristics, physician factors, and hospital type. The consistency of these findings is underscored in “Unequal treatment,” the recent Institute of Medicine report summarizing racial disparities in health care use.8 Although investigators have examined the influence of patient and provider factors, the role of geographic location and the broader question of whether racial differences in reperfusion therapy use reflect a nationally consistent pattern of lower treatment rates for black patients or instead vary based on geographic location have not been investigated.

Regional differences in MI treatment are well established.9 In a recent evaluation of quality of care provided to Medicare patients, 7 Southern states—Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, and Texas—were among the 10 worst performing states nationally.10 The large black population in these states has led to speculation that poorer quality of care in the South may further disproportionately affect black patients.11 Although selected studies have reported regional variations in racial differences in cardiac procedure use,12–18 these findings are difficult to interpret given racial differences in procedure appropriateness, patient preferences, insurance, and access to health care. In contrast, reperfusion therapy is a treatment with clear indications for use, provided to patients who demonstrated access by virtue of hospitalization and administered in an emergent fashion, thereby limiting concerns of costs or preferences. An evaluation of racial differences in reperfusion therapy, and the necessary identification of potential variations in racial differences in treatment, is particularly important given the extensive efforts proposed to redress racial differences in treatment.19 Identification of regions with larger, smaller, or no racial differences in treatment may help focus efforts directed at reducing racial differences in patient care and be a helpful adjunct to the soon-to-be-released National Health Care Disparity Report Card.20 However, no study, to our knowledge, has examined whether similar geographic variations would be observed in an assessment of racial differences in a strongly indicated clinical strategy such as reperfusion therapy.

Accordingly, we undertook an evaluation of patients enrolled in the National Registry of Myocardial Infarction (NRMI)-2 and -3 cohorts, a detailed clinical database of patients hospitalized for MI throughout the United States, to determine whether racial differences in reperfusion therapy varied by geographic region. Specifically, we sought to determine whether racial differences in reperfusion therapy were greater among patients hospitalized in the South compared with other regions of the United States.

Methods

National registry of myocardial infarction-2 and -3

NRMI-2 and -3 were prospective, phase IV observational evaluations of MI treatment sponsored by Genentech, Inc.21 Patients hospitalized with a clinically confirmed MI at one of nearly 1800 NRMI-2 or -3 participating hospitals between June 1994 and March 1998 (NRMI 2), and April 1998 and June 2000 (NRMI-3) were prospectively enrolled in the registries. Hospitals were encouraged to enroll all patients and were provided standardized materials to facilitate case identification and registry. On-site reviewers abstracted patients’ demographic characteristics, clinical presentation, in-hospital treatment and events, and discharge disposition from medical records. Data quality was promoted through the use of standardized study abstraction forms, centralized training of medical record reviewers, and auditing of forms.

We restricted our evaluation of the 1310030 hospitalizations in the NRMI-2 and -3 cohorts to the 378654 patients who had an ST-segment elevation or a left-bundle branch block on their admission electrocardiogram and had no specific chart-documented clinical contraindications to reperfusion therapy (eg, arrived within 12 hours of chest pain onset, no risk of bleeding, etc). Patients who transferred to the hospital (n = 83478) and those discharged alive within 24 hours of admission (n = 49433) were excluded from analysis because we could not reliably assess if they had received reperfusion therapy. Patients whose race was missing (n = 15775) and those with a race other than black or white (n = 15889) were also excluded from analysis. In total, 153020 patients met one or more of the above criteria, leaving 225634 patients eligible for analysis.

Of the 225634 cases eligible for analysis, we excluded those treated at 129 hospitals for which 1994 American Hospital Association Annual Survey data22 (n = 12032) were not available and patients treated outside of the 50 United States and the District of Columbia (n = 13). Patients with missing data for clinical covariates previously identified as predictors of use of reperfusion therapy23 in the NRMI-2 population and systolic blood pressure (n = 21391) were also excluded from analysis. In total, 21404 eligible patients (9.5%) met one or more of the above criteria; the remaining 204230 patients comprised the study cohort.

Reperfusion therapy use

All patients in the study cohort were eligible for reperfusion therapy and had no clinical treatment contraindications. Patients’ medical records were evaluated for the use of any reperfusion therapy within 12 hours of admission. Patients who received reperfusion therapy were further classified based on their form of treatment as having received thrombolytic agents, undergone primary percutaneous coronary intervention (PCI), or coronary artery bypass surgery. Patients who received more than one form of reperfusion therapy were assigned to the therapy they received first.

Geographic location

Patients were assigned to one of 4 US Census regions (Northeast, South, Midwest, West) based on the state in which the hospital they were treated in was located (Appendix A).24 Patients located in the South US Census region were further stratified into its 3 constituent US Census divisions (South Atlantic, East South Central, West South Central)24 to establish the consistency of any observed variations. To specifically test the hypothesis that racial variations in reperfusion therapy would differ in the South, patients hospitalized in the Northeast, Midwest, and West were aggregated into a single “non-South” region.

Statistical analysis

Clinical and hospital characteristics were compared between black patients and white patients using χ2 and Wilcoxon rank-sum tests. Crude rates of reperfusion therapy use overall, by thrombolytic therapy, and by primary PCI were compared between black patients and white patients in the overall cohort using χ2 analyses. Racial differences in crude rates of reperfusion therapy use were compared across the 4 US Census regions, the 3 US Census divisions within the US Census South region, and between the US Census South region and the previously defined “non-South” region. Mantel-Hanzel tests of heterogeneity were used to determine whether racial differences in overall reperfusion therapy use, thrombolytic therapy use, and primary PCI use varied by geographic location.

Multivariable logistic regression analyses were conducted to assess whether racial variations in reperfusion therapy use across US Census regions were independent of other racial or geographic differences in the NRMI cohort. Models of reperfusion therapy, thrombolytic, and primary PCI use adjusted for patient factors previously identified as predictors of reperfusion therapy use in the NRMI cohort, including age, sex, presence of left bundle branch block, absence of chest pain, medical history (MI, hypertension, stroke/cerebrovascular disease, smoking status, congestive heart failure, angina, diabetes, coronary revascularization), Killip class, infarct location, time to presentation, and admission systolic blood pressure.23 Logistic regression models were also adjusted for hospital characteristics, including teaching status, ownership, MI volume, cardiac care facilities, and location in a rural area. Regional variations in racial differences in reperfusion therapy use were assessed by the inclusion of patient race * US Census region interaction terms and the comparison of likelihood ratios between logistic regression models with and without patient race * US Census region interaction terms.

To account for hospital confounding of the association between race and use of reperfusion therapy, analyses were repeated modeling the effect of race as a combination of between-hospital variation and within-hospital variation using the methods outlined by Localio et al.25 Risk-standardized treatment rates were derived by multiplying the ratio of regions’ observed to predicted treatment rate by the overall cohort’s treatment rate. Because of the size of the NRMI cohort, findings are reported with 99% CIs. All analyses were conducted using the SAS 8.2 (SAS Institute Inc, Cary, NC) and STATA 6.0 (STATA Inc, College Station, Tex) software packages.

Results

The study cohort consisted of 12823 (6.3%) black patients and 191407 white patients. Black patients were younger and a smaller proportion were men or had commercial insurance compared with white patients. Compared with white patients, fewer black patients had a medical history of angina or prior coronary revascularization and a lower in-hospital mortality rate. A greater proportion of black patients experienced an anterior infarction, did not present with chest pain, were smokers, had diabetes, hypertension, and presented with elevated systolic blood pressure than white patients; other clinical characteristics were generally comparable between white patients and black patients. Black patients were more likely to be treated at urban, teaching, and coronary bypass surgery–capable hospitals. More than half of all black patients were hospitalized in the South US Census region (Table I).

Table I.

Patient characteristics

Characteristics Overall Black White P
n 204230 12823 191407 –
Percent of population 100.0 6.3 93.7
Demographic characteristics
 Age <.0001
  <55 years 25.7 36.9 24.9
  55–64 years 21.8 25.2 21.6
  65–74 years 24.9 21.9 25.1
  75 years and older 27.6 16.1 28.4
 Female sex 35.7 40.9 35.4 <.0001
 NRMI 2 registry 66.9 67.2 66.8 .41
 Primary payer <.0001
  Age <65 and commercial insurer 23.1 20.1 23.3
  Age <65 and HMO 8.6 10.9 8.4
  Medicare 55.6 44.9 56.4
  Age <65 and Medicaid 1.9 5.7 1.7
 Age <65 and other type of insurance 10.8 18.5 10.3
Clinical presentation
 ST segment elevation 90.3 91.8 90.1 <.0001
 Left bundle branch block 11.5 9.6 11.7 <.0001
 Anterior infarction 32.8 36.3 32.5 <.0001
 No chest pain at presentation 11.9 13.0 11.8 <.0001
 Median systolic blood pressure, mm Hg 141 143 141 <.0001
 Median diastolic blood pressure, mm Hg 82 84 82 <.0001
 Median heart rate, beats/min 80 81 80 <.0001
Admission Killip class .034
  I 81.4 81.0 81.5
  II 12.9 12.8 12.9
  III 5.7 6.2 5.7
 Time to presentation <.0001
  Less than 3 h after symptoms 72.5 69.2 72.8
  3–6 h after symptoms 17.6 19.3 17.5
  >6–12 h after symptoms 9.9 11.5 9.8
 Arrived at night (12:00 AM–7:59 AM) 26.8 30.7 26.5 <.0001
Medical history
 Smoker 32.9 39.9 32.4 <.0001
 Diabetes 22.2 31.5 21.5 <.0001
 Congestive heart failure 10.1 11.0 10.0 .0003
 Stroke 5.1 6.5 5.0 <.0001
 Myocardial infarction 22.4 23.4 22.3 .004
 Hypertension 48.5 64.5 47.4 <.0001
 Angina 13.5 11.6 13.6 <.0001
 Coronary bypass surgery 9.2 5.2 9.5 <.0001
 Percutaneous coronary intervention 8.9 8.4 8.9 .052
Hospital characteristics
 Teaching status <.0001
  Non-teaching 56.7 43.1 57.6
  Residency affiliated 31.3 30.9 31.3
  Council of Teaching 12.0 25.9 11.1
  Hospital member
 Ownership <.0001
  Public 11.9 19.0 11.4
  Not-for-profit 81.8 75.9 82.2
  For-profit 6.3 5.1 6.4
 Annual myocardial infarction volume quartiles <.0001
  78 cases or fewer 6.1 5.5 6.1
  79–133 cases 14.6 15.6 14.5
  134–223 cases 27.3 30.4 27.1
  224 cases or more 52.0 48.4 52.2
 Cardiac care facilities <.0001
  No invasive capabilities 14.8 8.2 15.2
  Cardiac catheterization capable 31.2 29.4 31.3
  Coronary bypass surgery capable 54.0 62.3 53.5
 Located in metropolitan statistical area 85.4 91.9 85.0 <.0001
 Geographic location—US Census region <.0001
  Northeast 19.3 11.8 19.8
  South 32.6 53.0 31.2
  Midwest 27.6 24.4 27.9
  West 20.4 10.9 21.1
Open in a new tab

Overall, black patients had lower rates of reperfusion therapy use (66.5% vs 69.9%, P < .0001), thrombolytic therapy use (52.3% vs 53.9%, P = .0004), and primary PCI use (13.8% vs 15.3%, P < .0001) than white patients (Table II). Racial differences in reperfusion therapy use varied by US Census region (P = .0001 Mantel-Hanzel test of heterogeneity). Black patients, relative to white patients, were more likely to receive reperfusion therapy in the Northeast (+2.7%, 99% CI −0.5% to 5.8%) and there were no significant racial differences in treatment use in the Midwest (−0.7%, 99% CI −2.9% to 1.5%), or West (−1.9%, 99% CI −5.1% to 1.2%). Lower rates of reperfusion therapy use in black patients were only observed in the South (−7.1%, 99% CI −8.7% to −5.6%). Black patients in the South were less likely to receive both thrombolytic agents (−4.5%, 99% CI −6.1% to −2.8%) and primary PCI (−2.4%, 99% CI −3.5% to −1.3%), whereas there were no racial differences in thrombolytic use in any other regions. Black patients had lower rates of reperfusion therapy use in the 3 Census Divisions that comprised the US South: South Atlantic (63.9% black vs 71.1% white, P < .0001), West South Central (66.4% black vs 73.1% white, P < .0001), and the East South Central (64.5% black vs 71.17% white, P < .0001).

Table II.

Crude reperfusion therapy treatment rates by race and US Census region

US Census regions
Overall Northeast Midwest West South Non-South US
n 204230 39445 56466 41765 66554 137676
Any reperfusion therapy
 Overall 69.6 65.4 69.0 72.5 70.9 69.0
 Black 66.5 67.9 68.3 70.7 64.5 68.8
 White 69.9 65.3 69.0 72.6 71.7 69.0
 P <.0001 .032 .39 .11 <.0001 .65
 Racial difference (99% CI) −3.3 (−4.4, −2.2) 2.7 (−0.5, 5.8) −0.7 (−2.9, 1.5) −1.9 (−5.1, 1.2) −7.1 (−8.7, −5.6) −0.3 (−1.8, 1.3)
Thrombolytic agents
 Overall 53.8 59.3 49.6 51.5 55.6 52.9
 Black 52.3 59.9 49.3 54.4 51.5 53.1
 White 53.9 59.3 49.6 51.4 56.0 52.9
 P .0004 .66 .79 .028 <.0001 .77
 Racial difference (99% CI) −1.6 (−2.8, −0.4) 0.6 (−2.7, 3.9) −0.2 (−2.6, 2.1) 3.0 (−0.5, 6.5) −4.5 (−6.1, −2.8) 0.2 (−1.5, 1.9)
Percutaneous coronary intervention
 Overall 15.2 5.9 18.7 20.2 14.7 15.5
 Black 13.8 8.0 18.7 15.6 12.5 15.3
 White 15.3 5.8 18.7 20.3 14.9 15.5
 P <.0001 .0004 1.00 <.0001 <.0001 .71
 Racial difference (99% CI) −1.5 (−2.3, −0.7) 2.2 (0.4, 4.0) 0.0 (−1.8, 1.9) −4.8 (−7.3, −2.2) −2.4 (33.5, 31.3) −0.2 (−1.4, 1.0)
Open in a new tab

Treatment rates reported as the proportion of patients who received the therapy.

Racial difference reported as the absolute difference in treatment rates between black patients and white patients with negative numbers indicating lower treatment rates among blacks compared with whites.

P < .0001 or lower for Mantel-Hanzel tests of heterogeneity in effect of race on treatment across 4 US Census regions and for the comparison of the South vs the rest of the United States for all 3 reperfusion therapy endpoints.

Note: Treatment by thrombolytic agents and percutaneous coronary intervention does not equal “any reperfusion therapy ” as 1300 (0.6%) patients who underwent immediate coronary bypass surgery for reperfusion therapy are not reported.

The race and US Census region interaction for use of reperfusion therapy, specifically the larger racial difference in treatment use among patients in the South US Census region, was confirmed in multivariable analysis (P < .0001 for specific interaction term and likelihood ratio tests). Although multivariable adjustment reduced the magnitude of the racial difference in treatment use in the South, black patients continued to have a 5.4% absolute lower risk-standardized rate of reperfusion therapy use than white patients. In contrast, there were no racial differences in reperfusion therapy use in the Northeast and small, statistically nonsignificant racial differences in the Midwest and West (Table III). Black patients in the South had lower risk-standardized rates of thrombolytic therapy use than white patients, whereas there were no racial differences in thrombolytic therapy use in the other regions. Racial differences in risk-standardized rates of primary PCI were observed in all regions. Findings were similar when analyses were repeated separating the effect of patient race into its between-hospital and within-hospital components (results not shown).

Table III.

Risk-standardized acute reperfusion therapy rates by race and US Census region

US Census regions
Overall Northeast Midwest West South Non-South US
n 204230 39445 56466 41765 66554 137676
Any reperfusion therapy
 Overall 69.6 69.0 69.7 71.1 68.9 69.9
 Black 65.7 68.1 67.0 68.3 64.0 67.5
 White 69.9 69.0 69.8 71.3 69.5 70.0
 P <.001 .69 .06 .19 <.001 .022
 Racial difference −4.2 −0.8 −2.9 −3.0 −5.4 −2.5
Thrombolytic agents
 Overall 53.8 55.1 51.9 51.7 55.8 52.8
 Black 54.0 56.8 53.1 55.0 53.5 54.6
 White 53.8 55.0 51.9 51.6 56.1 52.8
 P .74 .36 .37 .08 .009 .07
 Racial difference 0.2 1.7 1.2 3.4 −2.5 1.8
Percutaneous coronary intervention
 Overall 15.2 11.2 16.9 18.5 12.9 16.5
 Black 12.2 11.0 14.3 13.2 11.0 13.5
 White 15.4 11.2 17.0 18.7 13.2 16.7
 P <.001 .85 <.001 <.001 <.001 <.001
 Racial difference −3.3 −0.2 −2.7 −5.5 −2.2 −3.2
Open in a new tab

Treatment rates reported as the proportion of patients who received the therapy.

Racial difference reported as the absolute difference in treatment rates between black patients and white patients with negative numbers indicating lower treatment rates among blacks compared with whites.

P < .0001 or lower for Mantel-Hanzel tests of heterogeneity in effect of race on treatment across 4 US Census regions and for the comparison of the South vs the rest of the United States for all 3 reperfusion therapy end points.

Note: Treatment by thrombolytic agents and percutaneous coronary intervention does not equal “any reperfusion therapy” as 1300 (0.6%) patients who underwent immediate coronary bypass surgery for reperfusion therapy are not reported.

Discussion

As reported in prior studies, we found that black patients hospitalized with MI were less likely to receive reperfusion therapy than white patients.1–7 However, our data indicate that racial differences in reperfusion therapy use vary by geographic location. We found no racial differences in the use of reperfusion therapy in the Northeast and small, statistically nonsignificant racial differences for patients hospitalized in the Midwest and West. Racial differences in reperfusion therapy use were greatest among patients hospitalized in the US South. Thus, previous reports of lower rates of reperfusion therapy use among black patients with MI do not reflect a nationally consistent pattern of racial differences in treatment, but rather a pattern that varies based on geographic location.

This study is the first, to our knowledge, to specifically examine whether racial variations in reperfusion therapy use vary by geographic region. The use of the NRMI database and our evaluation of reperfusion therapy provide a significant methodological advancement upon previous studies of geographic variations in racial differences in treatment. Previous studies identified geographic variations in racial differences in cardiac procedure use and other potentially discretionary treatments.12–18 Thus, it was unclear whether racial differences in treatment reflected shortfalls in the use of necessary therapies. We evaluated a guideline-recommended therapy with proven mortality benefits in a cohort of patients confirmed as eligible for treatment with no therapy-specific clinical contraindications. Thus, treatment differences in our population provide definitive evidence of disparities in appropriate clinical care. Prior studies also lacked detailed clinical data and were based on patients drawn from selected payer groups.12–18 The NRMI registry contains detailed, chart-abstracted data and is not restricted to specific patient populations. Our analysis accounted for hospital cardiac care facilities, MI volume, and teaching status, suggesting racial differences are independent of hospital characteristics. Further, our secondary analyses indicated racial differences in treatment use were not attributable to black patients’ disproportionate use of hospitals less likely to provide reperfusion therapy use. Thus, our study provides strong evidence that racial differences in reperfusion therapy use vary by geographic location.

Our study confirms previous reports of geographic variations in racial differences in treatment use.16 Racial differences in the use of coronary artery bypass surgery, although observed nationwide, were more than twice as large in the 7 contiguous states of Alabama, Arkansas, Georgia, Louisiana, Mississippi, North Carolina, and South Carolina in an evaluation of 1986 Medicare data.14 Racial differences in the use of cardiac catheterization, angioplasty, and coronary bypass surgery were similarly greatest in the South within the Veterans Affairs Health Care System and later Medicare cohorts.12,17,18 Geographic variations in racial differences in treatment have also been observed for noncardiac services. Escarce et al reported that black Medicare patients were less likely to receive 23 of 32 common services, with racial differences larger in the South than in other Census regions.13 Lee et al similarly demonstrated that racial differences in Medicare patients’ use of services were greatest in the South and further exacerbated when expressed in terms of resource use.15 The consistent pattern of greater racial differences in treatment use in the US South observed in our study and prior evaluations of cardiac procedure suggests the existence of a disparate practice pattern that appears specific to the South.

Racial differences in reperfusion therapy use may not necessarily reflect the effect of race per se, but rather racial differences in patient presentation. Racial differences in presentation may further vary by geographic region. Black patients hospitalized with MI in the South reported different symptoms than patients hospitalized in other regions.26 It is unclear what factors may account for this pattern, although region-specific cultural differences in patients’ subjective perceptions of symptoms have been suggested.26 Racial differences in clinical presentation, including electrocardiographic features, have previously been shown to account for racial differences in the use of reperfusion therapy.27 Further, black patients experiencing MI are less likely than white patients to present with “classical” symptoms such as angina or chest discomfort or to attribute their symptoms to cardiac causes.28,29 Although our analysis accounted for racial differences in the prevalence of chest pain, the NRMI registry did not collect detailed patient presentation data, and thus we cannot preclude the influence of other aspects of patient presentation. In addition to racial differences in presentation, lower rates of arteriographic evidence of coronary artery disease30 and MI among black patients presenting with chest pain31 may decrease physicians’ suspicions of cardiac ischemia among black patients and result in missed diagnoses of cardiac ischemia.32 This phenomenon may be particularly prevalent in the South because of its large population of black persons. Finally, larger racial differences in treatment in the South may be attributable to higher rates of treatment refusal by black patients in the South compared with white patients in the South or black patients in other regions. However, we know of no data that suggest regional variations in racial differences in treatment preferences concerning reperfusion therapy.

Regional factors specific to the South may also explain its larger racial differences in reperfusion therapy use. Despite increasing racial diversity in the United States, the 2000 Census indicates blacks in the South continue to cluster in counties in which they comprise near majorities of the population.33 Residence in these areas may exert an “environmental” influence not accounted for in our evaluation of racial confounding by hospitals. The southern United States has historically had a high incidence of cardiovascular disease, although it is unclear why this increased burden would result in greater racial differences in treatment.34 Black persons in the South reside predominantly in rural areas, and minorities in rural areas have historically poorer quality health care than rural white or minorities in urban areas.35 However, we accounted for hospital-specific reperfusion therapy practices, suggesting racial variations are not attributable to black patients disproportionately presenting for treatment at lower quality hospitals than white patients. Alternatively, if racial differences in reperfusion therapy use are attributable to bias, then greater racial disparities in care observed in the US South may be an artifact of the greater prevalence of prejudicial attitudes toward minority groups documented in the region.36 Additional research is needed to determine whether and how these regional characteristics may contribute to racial differences in treatment use observed in the South.

The consequences of regional variations in racial differences in treatment are substantial. The 5.4% racial difference in risk-standardized rates of reperfusion therapy in the South is more than twice as large as the racial difference in the rest of the United States. This increased disparity in treatment is particularly important given that a majority of black persons in the United States reside in the US South. Of the more than 100000 black patients hospitalized for MI annually,37 the NRMI cohort suggests more than 50000 of these patients would be hospitalized in the US South. Extrapolating the 3.0% greater racial difference in risk-standardized treatment rates in the South to these more than 50000 patients, our data suggest that more than 1500 black patients would be undertreated primarily because they were hospitalized in the US South. Because of the greater concentration of black patients in the US South and the larger racial difference in treatment rates, our data underscore the importance of considering geographic location when considering racial differences in treatment.

Our study has certain issues to consider. We examined racial differences in US Census regions and did not explore variations in smaller geographic units such as states or counties. However, our purpose was not to identify small area variations in racial differences in treatment, but rather to examine the extent to which specific regions, and their associated practice patterns and cultures, might modify racial differences in treatment use. Although there is undoubtedly heterogeneity in racial differences in treatment use within regions, our findings suggest a general pattern of greater differences in the US South. The NRMI database does not contain measures of access or patient socioeconomic data, and thus racial differences observed in the US South may be attributable to socioeconomic factors, rather than race itself. However, it is unclear how access to care or socioeconomic factors would influence the use of an emergent therapy among hospitalized patients. Further, race and socioeconomic status are highly correlated, suggesting a common social influence that makes their distinction potentially arbitrary.38 The NRMI registry does not contain admitting physician data, thus we are unable to assess whether physician specialty, experience, or other characteristics may explain racial disparities in treatment. However, any such physician characteristics would have to manifest differentially in the South to explain the regional variation we observed. The NRMI represents a selective group of hospitals that are generally larger and have greater cardiac capabilities than other hospitals nationwide.39 Similarly, the proportion of black patients in the NRMI cohort is smaller than suggested by epidemiologic surveys of coronary disease. Although these differences may limit the generalizability of our findings, the selected nature of participating hospitals and underrepresentation of black patients suggests that the racial differences in treatment we observed likely reflect a “best-case” scenario given the shortfalls in quality of care documented at smaller volume hospitals without advanced cardiac care capabilities and the likely selected nature of black patients in the NRMI cohort.

Conclusion

Our evaluation of reperfusion therapy use among patients hospitalized with MI confirmed prior reports of racial differences in treatment use. However, racial differences in treatment varied by geographic region, as there were no differences in treatment in the Northeast, whereas racial differences were greatest in the US South. Our data, along with previous reports of larger racial differences in health care use in the South, suggest a consistent pattern of poorer quality of care provided to black persons that merits further examination. Our analysis further underscores the importance of accounting for geographic location in evaluations of racial disparities in care and specifically the manner in which regions may exacerbate, mitigate, or potentially explain racial differences in treatment observed at the national level. In the interim, greater efforts are needed to address shortfalls in reperfusion therapy use to ensure that all patients, regardless of race, receive high quality care.

Acknowledgments

The National Registry of Myocardial Infarction-2 and -3 is supported by Genentech, Inc, South San Francisco, Calif. Mr Rathore is supported by National Institutes of Health National Institute of General Medical Sciences Medical Scientist Training Grant GM07205.

Appendix A

Hospitals were grouped into the following 4 US Census regions based on the state in which they were located.

Northeast—Maine, New Hampshire, Vermont, Massachusetts, Rhode Island, Connecticut, New York, Pennsylvania, New Jersey

Midwest—Ohio, Michigan, Indiana, Illinois, Wisconsin, Missouri, Iowa, Minnesota, North Dakota, South Dakota, Nebraska, Kansas

West—Montana, Idaho, Wyoming, Colorado, Utah, New Mexico, Arizona, Nevada, California, Oregon, Washington, Alaska, Hawaii

South—Delaware, Maryland, District of Columbia, West Virginia, Virginia, North Carolina, South Carolina, Georgia, Florida are located in the South Atlantic Census division; Kentucky, Tennessee, Alabama, Mississippi are in the East South Central division; Arkansas, Louisiana, Oklahoma, Texas are in the West South Central Division.

Footnotes

Reprints not available from authors.

References

  • 1.Petersen LA, Wright SM, Peterson ED, et al. Impact of race on cardiac care and outcomes in veterans with acute myocardial infarction. Med Care. 2002;40:I86–96. doi: 10.1097/00005650-200201001-00010. [DOI] [PubMed] [Google Scholar]
  • 2.Pashos CL, Normand S-LT, Garfinkle JB, et al. Trends in the use of drug therapies in patients with acute myocardial infarction: 1988 to 1992. J Am Coll Cardiol. 1994;23:1023–30. doi: 10.1016/0735-1097(94)90585-1. [DOI] [PubMed] [Google Scholar]
  • 3.Manhapra A, Khaja F, Syed M, et al. Electrocardiographic presentation of blacks with first myocardial infarction does not explain race differences in thombolysis administration. Am Heart J. 2000;140:200–5. doi: 10.1067/mhj.2000.107173. [DOI] [PubMed] [Google Scholar]
  • 4.Syed M, Khaja F, Rybicki BA, et al. Effect of delay on racial differences in thrombolysis for acute myocardial infarction. Am Heart J. 2000;140:643–50. doi: 10.1067/mhj.2000.109644. [DOI] [PubMed] [Google Scholar]
  • 5.Rathore SS, Berger AK, Weinfurt KP, et al. Race, sex, poverty, and the medical treatment of acute myocardial infarction in the elderly. Circulation. 2000;102:642–8. doi: 10.1161/01.cir.102.6.642. [DOI] [PubMed] [Google Scholar]
  • 6.Canto JG, Allison JJ, Kiefe CI, et al. Relation of race and sex to the use of reperfusion therapy in Medicare beneficiaries with acute myocardial infarction. N Engl J Med. 2000;342:1094–100. doi: 10.1056/NEJM200004133421505. [DOI] [PubMed] [Google Scholar]
  • 7.Allison JJ, Kiefe CI, Centor RM, et al. Racial differences in the medical treatment of elderly Medicare patients with acute myocardial infarction. J Gen Intern Med. 1996;11:736–43. doi: 10.1007/BF02598987. [DOI] [PubMed] [Google Scholar]
  • 8.Institute of Medicine Unequal treatment. Washington (DC): Institute of Medicine; 2002. [Google Scholar]
  • 9.Pilote L, Califf RM, Sapp S, et al. Regional variation across the United States in the management of acute myocardial infarction. N Engl J Med. 1995;333:565–72. doi: 10.1056/NEJM199508313330907. [DOI] [PubMed] [Google Scholar]
  • 10.Jencks SF, Cuerdon T, Burwen DR, et al. Quality of medical care delivered to Medicare beneficiaries: a profile at state and national levels. JAMA. 2000;284:1670–6. doi: 10.1001/jama.284.13.1670. [DOI] [PubMed] [Google Scholar]
  • 11.Etchason J, Armour B, Ofili E, et al. Racial and ethnic disparities in health care [letter] JAMA. 2001;285:883. doi: 10.1001/jama.285.7.883-a. [DOI] [PubMed] [Google Scholar]
  • 12.Daniels MJ, Gatsonis C. Hierarchical polytomous regression models with applications to health services research. Stat Med. 1997;16:2311–25. doi: 10.1002/(sici)1097-0258(19971030)16:20<2311::aid-sim654>3.0.co;2-e. [DOI] [PubMed] [Google Scholar]
  • 13.Escarce JJ, Epstein KR, Colby DC, et al. Racial differences in the elderly’s use of medical procedures and diagnostic tests. Am J Public Health. 1993;83:948–54. doi: 10.2105/ajph.83.7.948. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Goldberg KC, Hartz AJ, Jacobsen SJ, et al. Racial and community factors influencing coronary artery bypass graft surgery rates for all 1986 Medicare patients. JAMA. 1986;267:1473–7. [PubMed] [Google Scholar]
  • 15.Lee AJ, Gehlbach S, Hosmer D, et al. Medicare treatment differences for blacks and whites. Med Care. 1997;35:1173–89. doi: 10.1097/00005650-199712000-00002. [DOI] [PubMed] [Google Scholar]
  • 16.Mirvis DM, Burns R, Gaschen L, et al. Variation in utilization of cardiac procedures in the Department of Veterans Affairs Health Care System: effect of race. J Am Coll Cardiol. 1994;24:1297–304. doi: 10.1016/0735-1097(94)90112-0. [DOI] [PubMed] [Google Scholar]
  • 17.Whittle J, Conigliaro J, Good CB, et al. Racial differences in the use of invasive cardiovascular procedures in the Department of Veterans Affairs Medical System. N Engl J Med. 1993;329:621–7. doi: 10.1056/NEJM199308263290907. [DOI] [PubMed] [Google Scholar]
  • 18.Gatsonis CA, Epstein AM, Newhouse JP, et al. Variations in the utilization of coronary angiography for elderly patients with an acute myocardial infarction: an analysis using hierarchical logistic regression. Med Care. 1995;33:625–42. doi: 10.1097/00005650-199506000-00005. [DOI] [PubMed] [Google Scholar]
  • 19.Agency for Healthcare Research and Quality. Excellence centers to eliminate ethnic/racial disparities (EXCEED) Agency for Healthcare Research and Quality; 2002. [Accessed November 17, 2004]. Available at: http://www.ahrq.gov/research/exceed.htm. [Google Scholar]
  • 20.Guidance for the National Healthcare Disparities Report. Washington (DC): The National Academies Press; 2002. Committee on Guidance for Designing a National Healthcare Disparities Report. [PubMed] [Google Scholar]
  • 21.National Registry of Myocardial Infarction. Genentech, Incorporated; 2002. [Google Scholar]
  • 22.American Hospital Association. 1994 Annual survey of hospitals. Chicago (Ill): American Hospital Association; 2002. [Google Scholar]
  • 23.Barron HV, Bowlby LJ, Breen T, et al. Use of reperfusion therapy for acute myocardial infarction in the United States: data from the National Registry of Myocardial Infarction 2. Circulation. 1998;97:1150–6. doi: 10.1161/01.cir.97.12.1150. [DOI] [PubMed] [Google Scholar]
  • 24.U.S. Census. Census regions and divisions. Washington (DC): US Department of Commerce; 2002. [Google Scholar]
  • 25.Localio AR, Berlin JA, Ten Have TR, et al. Adjustments for center in multicenter studies: an overview. Ann Intern Med. 2001;135:112–23. doi: 10.7326/0003-4819-135-2-200107170-00012. [DOI] [PubMed] [Google Scholar]
  • 26.Summers RL, Cooper GJ, Carlton FB, et al. Prevalence of atypical chest pain descriptions in a population from the Southern United States. Am J Med Sci. 1999;318:142–5. doi: 10.1097/00000441-199909000-00008. [DOI] [PubMed] [Google Scholar]
  • 27.Borzak S, Joseph C, Havstad S, et al. Lower thrombolytics use for African Americans with myocardial infarction: an influence of clinical presentation? Am Heart J. 1999;137:338–45. doi: 10.1053/hj.1999.v137.92523. [DOI] [PubMed] [Google Scholar]
  • 28.Racyznski JM, Taylor H, Cutter G, et al. Diagnoses, symptoms, and attribution of symptoms among black and white inpatients admitted for coronary heart disease. Am J Public Health. 1994;84:951–6. doi: 10.2105/ajph.84.6.951. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Lee H-O. Typical and atypical clinical signs and symptoms of myocardial infarction and delayed seeking of professional care among blacks. Am J Crit Care. 1997;6:7–13. [PubMed] [Google Scholar]
  • 30.Parish DC, Klekamp J. Arteriographic incidence of coronary artery disease in black men with chest pain. South Med J. 1994;87:33–7. doi: 10.1097/00007611-199401000-00008. [DOI] [PubMed] [Google Scholar]
  • 31.Maynard C, Beshansky JR, Griffith JL, et al. Causes of chest pain and symptoms suggestive of acute cardiac ischemia in African-American patients presenting to the emergency department: a multicenter study. J Natl Med Assoc. 1997;89:665–71. [PMC free article] [PubMed] [Google Scholar]
  • 32.Pope JH, Aufderheide TP, Ruthazer R, et al. Missed diagnoses of acute cardiac ischemia in the emergency department. N Engl J Med. 2000;342:1163–70. doi: 10.1056/NEJM200004203421603. [DOI] [PubMed] [Google Scholar]
  • 33.Schmitt E. US now more diverse, ethnically and racially. NY Times. 2001 April 1;:A20. [Google Scholar]
  • 34.Cooper R, Cutler J, Desvigne-Nickens P, et al. Trends and disparities in coronary heart disease, stroke, and other cardiovascular diseases in the United States. Circulation. 2000;102:3137–47. doi: 10.1161/01.cir.102.25.3137. [DOI] [PubMed] [Google Scholar]
  • 35.Slifkin RT, Goldsmith LJ, Ricketts TC. Working Paper No. 66: Race and place: urban-rural differences in health for racial and ethnic minorities. Chapel Hill (NC): Cecil G. Sheps Center for Health Services Research North Carolina Rural Health Research Program; 2002. [Google Scholar]
  • 36.Edsall TB. 25% of US view Chinese Americans negatively, poll says. Washington Post. 2001 April 16;:A4. [Google Scholar]
  • 37.American Heart Association. 2002 heart and stroke statistical update. Dallas (Tex): American Heart Association; 2001. [Google Scholar]
  • 38.Kaufman JS, Cooper RS, McGee DL. Socioeconomic status and health in blacks and whites: the problems of residual confounding and the resiliency of race. Epidemiology. 1997;8:621–8. [PubMed] [Google Scholar]
  • 39.Rogers WJ, Bowlby LJ, Chandra NC, et al. Treatment of myocardial infarction in the United States (1990 to 1993): observations from the National Registry of Myocardial Infarction. Circulation. 1994;90:2103–14. doi: 10.1161/01.cir.90.4.2103. [DOI] [PubMed] [Google Scholar]

RESOURCES