Author + information
- Received October 15, 1996
- Revision received February 19, 1997
- Accepted February 28, 1997
- Published online June 1, 1997.
- ↵*Dr. Richard F. Gillum, Centers for Disease Control and Prevention, National Center for Health Statistics, 6525 Belcrest Road, Hyattsville, Maryland 20782.
Objectives. We sought to determine whether racial differences in rates of coronary artery bypass graft surgery (CABG), percutaneous transluminal coronary angioplasty (PTCA) and cardiac catheterization decreased after 1980.
Background. Many reports of racial differences in utilization of CABG have been published since 1982. However, changes in the relative utilization of revascularization over time have received little attention.
Methods. Data from the National Hospital Discharge Survey were examined for the years 1980 through 1993. Estimated numbers of procedures performed in nonfederal U.S. hospitals were used to compute age-adjusted rates per 100,000 population by year and race for patients 35 to 84 years old.
Results. In patients 35 to 84 years old, the rate of CABG increased in blacks and whites between 1980 and 1993. Between 1986 and 1993, there was little change in the black/white ratio of age-adjusted rates (0.23 in 1980 through 1985 combined, 0.38 in 1986 and 0.43 in 1993). An apparent increase from 0.23 in 1980 through 1985 combined may have been due to sampling variation. Despite rapid increases in rates of PTCA in both races, no increase in the black/white ratio was noted (0.57 in 1993). However, the rate of inpatient cardiac catheterization increased more rapidly in blacks than in whites. This resulted in an increase in the black/white ratio of age-adjusted rates from 0.42 in 1980 to 0.91 in 1993.
Conclusions. Rates of CABG, cardiac catheterization and especially PTCA increased between 1980 and 1993, a period during which racial disparities in the procedures became widely known. Despite apparent increases in the black/white ratio for inpatient cardiac catheterization, large racial disparities in the utilization of CABG and PTCA persist and require further evaluation and possible intervention.
(J Am Coll Cardiol 1997;29:1557–62)
Despite a long-term decline in death rates, heart disease remains the leading cause of death and an important cause of serious disability in Americans of African as well as European ancestry in the United States (). It is also a major contributor to health care costs. Hospital discharge rates for coronary heart disease have not declined appreciably (). Since the first published report in 1982 () that numbers and rates of coronary artery bypass graft surgery (CABG) and coronary angiography were grossly lower in blacks than in whites, despite similar coronary death rates, numerous studies ([4–10]) have confirmed race-related disparities in the utilization of CABG and cardiac catheterization and extended the observation to percutaneous transluminal coronary angioplasty (PTCA) and other cardiovascular surgery and diagnostic procedures. However, possible changes in the relative utilization of these procedures over time have rarely been addressed. To test the hypothesis that utilization rates in blacks became more similar to those in whites between 1980 and 1993, data from a national survey of hospital discharges were examined to describe trends in the relative utilization of CABG, PTCA and cardiac catheterization by race in the United States.
1.1 Study Patients.
The National Hospital Discharge Survey (NHDS) of the National Center for Health Statistics (NCHS) produces estimates of numbers of patients discharged from nonfederal hospitals located in the 50 states and the District of Columbia ([2, 11–13]). Eligible hospitals include those licensed facilities with at least six beds where the average length of stay for all patients was <30 days. Hospital discharges are selected for abstracting by a multistage sampling procedure. From >6,000 hospitals, a sample of >500 was selected, of which >400 consented to participate and met eligibility criteria (exact numbers varied by year). Within hospitals, the daily listing sheet of discharges was the frame from which discharges were sampled randomly. For hospitals that provided data in electronic format, sampling was automated. Because discharges were sampled, an individual patient might appear more than once in the sample. Probabilities of selection at each stage of sampling were known, so that estimates of national statistics were obtained. Either hospital staffs or representatives of the NCHS performed the sample selection and transcribed information from the hospital record face sheet to abstract forms or extracted data from computer tapes. In 1988, a new sample design was initiated, which may have affected comparisons with earlier years, as described in detail elsewhere and discussed subsequently ().
The NCHS staff coded up to seven diagnoses and four procedures according to the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) ([11, 12, 14]). Discharges were estimated as follows: those with the following any-listed procedures (at least one of the following ICD-9-CM codes) for CABG (ICD-9-CM 36.10–36.19); and those with the following all-listed procedures: removal of coronary artery obstructions (ICD-9-CM 36.0); PTCA (ICD-9-CM 36.01, 36.02 and 36.05); and cardiac catheterization (ICD-9-CM 37.21–37.23). In 1993, at all ages, there were 398,000 all-listed procedures coded 36.0, 369,000 (93%) of which were PTCA. Hereafter, PTCA will be used to refer to ICD-9-CM 36.0 because this code was in use throughout the study period and the vast majority with this code are PTCA. Use of ICD-9-CM codes 37.21 to 37.23 (cardiac catheterization) captured essentially all coronary angiography cases; however, it may include an unknown number of catheterizations without coronary angiography (). Use of more specific codes for coronary arteriography (ICD-9-CM 88.55–88.57) has been shown () to identify only a portion of such procedures because of inadequate coding by hospitals when both cardiac catheterization and coronary angiography procedures are performed. In the NHDS from 1988 to 1990, 96.1% of those discharged at ≥35 years old with either cardiac catheterization or coronary arteriography codes had cardiac catheterization codes. Therefore, for enhanced comparability with other publications, data for cardiac catheterization are presented here.
1.2 Data Analysis.
To obtain more reliable estimates, given the relatively small numbers of cases in blacks in the sample, the average of three or more years of data was used in some analyses. The civilian resident population was used to compute discharge rates. No data were available on Hispanic origin; Hispanic whites were classified as white. Age-adjusted rates were computed by the direct method, with the U.S. population in 1980 as the standard ([2, 15]). Age-adjusted rates allow comparison of groups with different age distributions (e.g., blacks and whites) and minimize the effects of aging in the U.S. population over time on trend analyses. Records with race not stated were excluded (e.g., 17% of CABG in 1993). An approximation of the variance of estimates is provided using methods described previously (). For estimates of the number of procedures in whites in 1993, from the NHDS, of 50,000, 100,000 and 200,000, approximate relative standard errors were 11%, 9% and 8%, respectively (). Confidence intervals (estimate ±1.96 × SE) are given where appropriate. Rates or numbers that do not meet standards of reliability (relative SE >30%) are omitted or so noted.
Fig. 1shows the age-adjusted rate per 100,000 civilian population of CABG procedures by race in 1980 to 1993. In persons 35 to 84 years old, the rate of operation increased in blacks and whites between 1980 and 1993. There was little detectable change in the black/white ratio of age-adjusted rates (e.g., 0.38 in 1986 and 0.43 in 1993) (Fig. 2). An apparent increase from 0.23 in 1980 to 1985 combined may have been due to sampling variation, given the small estimated number of CABG procedures in blacks in this early period (22,000 in 1980 to 1985 combined). Estimated numbers of procedures and standard errors are shown in Table 1. In 1993, the number of CABG procedures at age 35 to 84 years in blacks was 10,000 (95% confidence interval [CI] 7,000 to 14,000).
Fig. 3shows the age-adjusted rate per 100,000 civilian population of PTCA procedures by race in 1980 to 1993, as approximated by ICD-9-CM 36.0 (removal of coronary artery obstruction) for consistent coding throughout the period. Before 1986, too few PTCA procedures were performed in blacks to permit reliable estimates in the NHDS. In 1993, the annual estimated numbers at age 35 to 84 years had risen to 17,000/year (95% CI 12,000 to 22,000) in blacks and 286,000/year in whites (Table 1). In subjects 35 to 84 years old, the age-adjusted rate of PTCA increased in blacks and whites between 1986 and 1993, with no evidence of a narrowing of the black–white gap (Fig. 3). The black/white ratio of age-adjusted rates was 0.66 in 1986 and 0.57 in 1993. However, because of the small number of black cases and high standard error of the estimates, differences in the ratio over time cannot be accurately assessed.
2.3 Cardiac Catheterization.
Fig. 4shows the age-adjusted rate per 100,000 civilian population of inpatient cardiac catheterization procedures by race in 1980 to 1993. In persons 35 to 84 years old, the rate of the procedure increased more rapidly in blacks than in whites, which resulted in an increase in the black/white ratio of age-adjusted rates from 0.42 in 1980 to 0.91 in 1993 (Fig. 5). In 1993, there were an estimated 70,000 (95% CI 56,000 to 85,000) in patient cardiac catheterization procedures in blacks 35 to 84 years old compared with 12,000 in 1980. In 1993, age-specific rates per 100,000 and a test for black–white difference for cardiac catheterization by age were as follows: 45 to 64 years old: black 767 (95% CI 588 to 947), white 709 (95% CI 610 to 809) (t = −0.55, p > 0.05); 65 to 84 years old: black 1,046 (95% CI 772 to 1,320), white 1,364 (95% CI 1,176 to 1,551) (t = 1.87, p > 0.05). Consistent with the age-adjusted findings, the age-specific rates for whites and blacks tended to converge in 1993 for both ages compared with rates in 1988 (the first year of the new survey design is not shown); this is consistent with trends in age-adjusted rates.
Marked increases in the rates of CABG, PTCA and cardiac catheterization have occurred since 1980 ([1, 2]). Data from the NHDS are consistent with more rapid increases in rates of inpatient cardiac catheterization in blacks compared with those in whites, resulting in convergence of inpatient utilization rates by 1993 (black/white ratio 0.91). Rates for CABG and PTCA remained lower in blacks than in whites throughout the period, with no significant change in black/white ratios. The trends in cardiac catheterization rates in blacks and whites might have been expected to lead to an increase in the black/white ratio, at least for PTCA if not for CABG; however, such a trend could not be demonstrated. Nevertheless, in 1993 blacks had 70% more PTCA than CABG procedures, whereas whites had only 24% more.
The increase in black/white ratios for cardiac catheterization may well be due to public and physician education efforts and resulting publications, symposia and continuing medical education programs to increase awareness of coronary heart disease as a major problem among blacks and the inequalities in treatment of coronary heart disease ([3–10]). Combined with public information, such efforts may have increased patient acceptance of cardiac catheterization procedures when recommended and increased physician willingness to recommend cardiac catheterization to black as well as white patients. Furthermore, a greater awareness among physicians of the difficulty in accurately assessing the probability of coronary stenosis in blacks on the basis of history and noninvasive tests alone may have provided a stimulus to greater utilization of cardiac catheterization for the diagnosis of coronary artery disease in blacks. Clearly, further studies are needed to confirm these trends and more adequately test these hypotheses.
Mechanisms for the repeatedly observed lower rates of CABG and PTCA in blacks than in whites have been discussed at length elsewhere ([4–10, 16–24]). They include a possible lower prevalence of multivessel coronary obstructive disease, lesser ability to pay, sociocultural preference for nonsurgical care and differences in care provided by physicians. Although all these may play a role, studies have demonstrated a persistent racial difference after controlling for coronary anatomy, income and health insurance coverage.
In 1990, CABG rates in U.S. blacks were similar to rates in the United Kingdom and Germany, the lowest among eight industrialized countries; PTCA rates in U.S. blacks were similar to those in Australia (). Age-adjusted ischemic heart disease mortality rates were similar in U.S. black men and German men and higher in men in the United Kingdom and Australia than in U.S. black men (). Rates of all these procedures were much higher in U.S. whites than in the residents of any of the other seven countries, all of which had provisions for providing the procedures to patients unable to afford them (). These findings suggest that high utilization in U.S. whites as well as underutilization in U.S. blacks contributes to racial disparities.
Limitations of NHDS data for the study of patterns and trends in cardiovascular procedures have been discussed elsewhere and include possible inaccuracies in listing and coding of procedures ([2, 4, 11–13, 16]). The NHDS samples discharges, not individual patients; therefore, it is not possible to distinguish first from repeat procedures, which are particularly likely for PTCA. Overall, in 1988 to 1990, only 6.3% of patients discharged at age ≥35 years with PTCA codes also had CABG codes.
In recent years, some cardiac catheterization or coronary arteriography procedures were performed in outpatient settings (). Unfortunately, the NHDS had no information on outpatient procedures. Thus, the total number of cardiac catheterization procedures is underestimated by NHDS data in more recent years. Preliminary analyses of data from the recently completed National Survey of Ambulatory Surgery indicated that in 1994, the first year of this survey, 21.6% of cardiac catheterization procedures in blacks and 23.2% in whites were done without hospital admission (NCHS, unpublished data). In patients 35 to 84 years old, the age-adjusted rate per 100,000 of total cardiac catheterization procedures (outpatient and inpatient) was 947 in blacks and 1,404 in whites (ratio 0.67). Thus, when outpatient cardiac catheterization was also considered, the increase in the black/white ratio was less striking than for inpatient cardiac catheterization (1993 ratio 0.91). However, the difference in ratios could also be due in part to chance variation in the less stable outpatient estimate. Because Veterans Affairs hospitals were not included, the tendency of blacks to use federal sources of care more so than whites might also have a small effect on the results. However, similar patterns of lower utilization by blacks have been reported within the Veterans Affairs hospital system ().
Because of the redesign of the NHDS in 1988, caution is required in comparing more recent with earlier numbers and rates (). The redesign of the NHDS reduced the percentage of discharges with race not stated. The estimated number of CABG and cardiac catheterization procedures was slightly higher and the number of PTCA procedures slightly lower with the new than the old method, according to a special study (). Race-specific analyses suggested that, if anything, the increase in the black/white ratio reported here might be understated (). The introduction of prospective payment for Medicare patients and diagnosis-related groups in 1983 to 1986 could also have had some impact (). The role of utilization review in the trends in blacks between 1985 and 1988 is also worthy of study, as is a possible ongoing effect of managed care ().
A limitation for analyses by race is the substantial number of discharges with race not stated (e.g., 17% of discharges with CABG in 1993) that were excluded from this study. One study () indicated that for the period 1988 to 1992, discharges with race not stated affected estimates for whites to a greater extent than for blacks, resulting in a possible greater underestimation of the number and rate of procedures in whites than in blacks (). Thus, the racial disparity in procedures may be, if anything, even greater than that shown. This problem is not limited to the present survey, because data on race were missing for 9% of procedures in a study () in which the medical records of Medicare patients were abstracted. Furthermore, no reliable data for Hispanic ethnicity or racial groups other than white and black are available in the NHDS, although a recent report () suggests different utilization in these groups compared with that in whites. Confounding by region cannot be excluded because analysis of trends by race and region was not possible.
Despite apparent increases in the black/white ratio for inpatient cardiac catheterization between 1980 and 1993, large racial disparities in utilization of CABG and PTCA persist and require further evaluation and possible intervention. Continued monitoring of trends in CABG, PTCA and cardiac catheterization will be useful in assessment of the impact of public and physician education efforts and clinical trials and in studies of the appropriateness of technology utilization. Trends may also reflect effects of changes in the health care system on the utilization of these procedures. For example, lower initial cost, as well as recent publication of results of clinical trials showing rates of death or acute myocardial infarction after 5 years that did not differ significantly with CABG and PTCA in selected patients ([31, 32]), might lead to a greater relative rate of increase in PTCA rates in blacks than in whites, similar to that reported here for cardiac catheterization. Studies of first PTCA rates by race are needed. Long-term trends in coronary procedures should be examined in Medicare data for subjects ≥65 years old. Data from multiple years of the National Survey of Ambulatory Surgery and the NHDS combined and state or regional data should be examined for all ages to replicate trend analyses reported here. Studies using demographic, clinical, noninvasive and arteriographic databases are needed to establish whether the lower utilization of CABG and PTCA in blacks compared with whites is due to a lower prevalence of coronary stenosis in blacks than in whites, less frequent operations in black than white patients with an equivalent degree of stenosis, lesser ability to pay, less acceptance of surgical treatment by blacks or less frequent recommendation of surgical treatment for blacks or all of these. Meta-analysis of data for blacks from published surgical trials and series and data from new trials with larger numbers of randomized black patients are needed to confirm the efficacy of PTCA and CABG compared with medical management and between PTCA and CABG in blacks (). Continued professional and public education efforts are needed to ensure medically appropriate levels of procedure access and utilization by blacks.
- coronary artery bypass grafting
- confidence interval
- International Classification of Diseases, 9th Revision, Clinical Modification
- National Center for Health Statistics
- National Hospital Discharge Survey
- percutaneous transluminal coronary angioplasty
- Received October 15, 1996.
- Revision received February 19, 1997.
- Accepted February 28, 1997.
- The American College of Cardiology
- Gillum RF
- ↵Gillum BS, Graves EJ, Kozak LJ. Trends in hospital utilization: United States 1988–92. National Center for Health Statistics. Vital Health Stat , no. 124. Washington (DC): U.S. Government Printing Office.
- ↵Graves EJ. National Hospital Discharge Survey: annual summary, 1993. National Center for Health Statistics. Vital Health Stat , no. 121. DHHS Publication No. (PHS) 95-1782. Washington (DC): U.S. Government Printing Office, 1995.
- Graves EJ. Detailed diagnoses and procedures for patients discharged from short-stay hospitals, United States, 1993, National Center for Health Statistics. Vital Health Stat , no. 122. DHHS Publication No. (PHS) 95-1783. Public Health Service, Washington (DC): U.S. Government Printing Office, 1995.
- ↵Haupt BJ, Kozak LJ. Estimates from two survey designs: National Hospital Discharge Survey. National Center for Health Statistics. Vital Health Stat  no. 111. DHHS Publication No. (PHS) 92-1772. Washington (DC): U.S. Government Printing Office, 1992.
- The International Classification of Diseases, 9th rev., Clinical Modification. 3rd ed. DHHS Publication No. (PHS) 89-1260. Washington (DC): U.S. Government Printing Office, 1989.
- Curtin LR, Klein RJ. Direct standardization (age-adjusted rates). Healthy People 2000 statistical notes. Hyattsville (MD): National Center for Health Statistics, No. 6, 1995:1–9.
- Peniston RL,
- Swarup J,
- Barnwell JM,
- Crittenden MD
- Cheitlin M
- Zarate AO
- ↵Kozak LJ, Hall MJ, Pokras R, Lawrence L. Ambulatory surgery in the United States, 1994. Advance data from vital and health statistics; no. 283. Hyattsville (MD): National Center for Health Statistics, 1997.
- ↵Kozak LJ. Underreporting of race in the National Hospital Discharge Survey: advance data from vital and health statistics, no. 265. Hyattsville (MD): National Center for Health Statistics, 1995.
- ↵National Heart, Lung, and Blood Institute. Report of the Working Group on Research in Coronary Heart Disease in Blacks. Bethesda (MD): National Institutes of Health, 1994:35–67.