Does Supplemental Private Insurance Affect Care of Medicare Recipients Hospitalized for Myocardial Infarction?

Abstract

Objectives. We sought to determine whether supplemental private insurance coverage among Medicare recipients alters patterns of health care or outcomes associated with acute myocardial infarction.

Methods. Medicare patients hospitalized after a myocardial infarction were identified from New York City hospitalization records. Patients who had only Medicare coverage were compared with those who had supplemental private or public insurance coverage.

Results. Patients with supplemental private insurance exhibited increased rates of revascularization and decreased rates of in-hospital mortality relative to patients with either Medicare only or Medicare and public insurance. Moreover, Blacks and women were less likely to undergo revascularization and exhibited higher in-hospital mortality rates.

Conclusions. Despite Medicare, private insurance coverage appears to influence the likelihood of coronary revascularization among older patients hospitalized for acute myocardial infarction.

Coronary heart disease remains the leading cause of death and disability in the United States, despite a sustained decline in age-adjusted mortality over the past 4 decades.1,2 Improved hospital survival after acute myocardial infarction may have been the largest contributor to this decline. Many factors affect individuals’ short-term prognosis after a myocardial infarction. For example, health insurance status, while often signifying the presence or nonpresence of financial resources to purchase specific care, might be a marker of general socioeconomic status. Patients with different types of health insurance coverage but similar medical problems have been shown to receive treatment that differs in form and intensity, which may affect short- as well as long-term outcomes.3 More specifically, race, insurance status, and income have all been associated with unequal use of revascularization procedures among coronary heart disease patients.4–6

Measuring and monitoring access to health care is a central concern of public health and health service researchers, and it has been of great interest to the federal government since the establishment of Medicare and Medicaid benefits in 1966.7,8 Because Medicare provides nearly universal hospital insurance coverage after the age of 65 years, we hypothesized that differences in revascularization prominently associated with insurance status among younger persons9 might be ameliorated or disappear after age 65. Since some Medicare beneficiaries also have supplemental insurance coverage, we wondered whether this additional coverage influenced medical services or survival.

To address this issue, we examined data from the New York State Department of Health’s Statewide Planning and Research Cooperative System (SPARCS). We found, as reported subsequently, that race and gender continue to influence revascularization use and in-hospital mortality among Medicare recipients hospitalized for acute myocardial infarction. However, supplemental private insurance coverage favorably affects both in-hospital care and hospital survival.

METHODS

Data

In this study, we used SPARCS data from the years 1988 through 2001.10 SPARCS contains discharge data abstracted for at least 95% of all New York State acute care hospitalizations other than those involving psychiatric and federal hospitals. SPARCS data include patients’ disposition, age, gender, race, admission status, physician and hospital identifiers, principal diagnosis and up to 14 secondary diagnoses, principal procedure code and up to 14 other procedure codes, principal payment resource, and second payment resource. Data records in the system are abstracted from medical records by trained medical records personnel in each hospital, and the New York State Department of Health is responsible for verifying the accuracy of reported information. The present analysis included all records of hospitals located in the 5 boroughs of New York City.

Diagnostic coding was based on the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM).11 Myocardial infarction was defined according to its appearance as the principal diagnosis code (ICD-9 codes 410.0–410.9). Other diagnostic codes identified comorbid conditions and complications. Patients’ disposition signified their vital status at discharge.

In this study, the outcomes of interest were revascularization procedure use and in-hospital mortality. Revascularization was coded as percutaneous transluminal coronary angioplasty (PTCA) (ICD-9 codes 36.01, 36.02, and 36.05) or coronary artery bypass grafting (CABG) (ICD-9 codes 36.10–36.19). Discharge status was used in determining in-hospital mortality.

A priori risk factors selected for this study included general risk factors (e.g., age, gender, race, length of hospital stay, and admission status), comorbid conditions (e.g., diabetes [ICD-9 code 250], hypertension [ICD-9 codes 401–405]), and complications of myocardial infarction, including congestive heart failure (ICD-9 code 428). We established whether patients had experienced a previous myocardial infarction by reviewing their medical records for mention of such an event (ICD-9 code 412). The location of the myocardial infarction was categorized as anterior (ICD-9 codes 410.0–410.1), lateral/inferior (ICD-9 codes 410.2–410.6), or subendocardial (ICD-9 code 410.7).

Insurance status was based on primary and secondary coverage. Patients were categorized in the following groups: (1) Medicare only; (2) Medicare with supplemental private insurance, including both Medicare and any comprehensive private insurance plan; and (3) Medicare with other public insurance, including both Medicare and Medicaid coverage.

Study patients were limited to those older than 65 years who reported having Medicare insurance coverage (88% of all patients 65 years or older reported having such coverage) and who were discharged from the hospital with a principal diagnosis of acute myocardial infarction. Because 93% of the patients were non-Hispanic Whites, non-Hispanic Blacks, or Hispanics, patients in other race/ethnicity groups were eliminated. Length of hospital stay categories were less than 3 days, 3 to 7 days, and more than 7 days. Admission status was categorized as emergency, urgent, or elective. Patients who were discharged from the hospital alive and who had been in the hospital fewer than 3 days were eliminated from our analyses because acute myocardial infarction was likely to have been ruled out among these patients.

Statistical Analysis

Associations between insurance status and sociodemographic characteristics, revascularization use, and hospital outcome were established. Multiple logistic regression models were used to estimate odds ratios and 95% confidence intervals for factors related to revascularization, as well as in-hospital mortality, controlling for other characteristics that differed significantly in the univariate analysis. Logistic regression models also were used in examining subgroups established through stratification according to race/ethnicity.

RESULTS

Patient Characteristics

During the study period, 124 599 Medicare patients hospitalized as a result of myocardial infarction were available for analysis. More than half of these patients (51.3%) were women; 11.5% were non-Hispanic Blacks, and 9.4% were Hispanics. Overall, 42.3% had Medicare as their only health insurance, 39.1% had supplemental private insurance, and 18.6% had both Medicare and Medicaid. Patients’ average age was 76.9 years; 57.5% were older than 75 years. There were no differences in age among patients in the different insurance status categories.

In comparison with patients who had Medicare coverage only, patients with supplemental private insurance were more likely to be White and male and to have had a previous myocardial infarction. Those with Medicare and other public insurance coverage were more likely to be female, Black, or Hispanic and more likely to have hypertension, diabetes, and congestive heart failure. Blacks and Hispanics were more likely than Whites to have Medicare in combination with other public insurance coverage and to have only Medicare coverage. Other characteristics, including length of hospital stay and admission status, differed significantly among the 3 insurance groups as well (Table 1 ▶).

TABLE 1—

Sample Characteristics, by Insurance Group

				Group Comparisons (P)
	Medicare Only (Group 1)	Medicare With Private Insurance (Group 2)	Medicare With Public Insurance (Group 3)	1 vs 2	1 vs 3	2 vs 3
Age, y, mean (SD)	77.4 (7.7)	77.5 (7.5)	77.5 (8.1)	>.05
Men, %	48.7	52.2	38.2	<.001
Race/ethnicity, %
Black	12.5	5.6	19.3	<.001
Hispanic	10.5	2.4	21.3
Emergency admission, %	87.3	80.1	83.8	<.001
Hospital stay >7 d, %	59.6	51.8	52.9	<.001
Hypertension, %	40.4	46.9	56.2	<.001
Diabetes, %	26.7	26.0	38.8	.14	<.001	<.001
Congestive heart failure, %	42.6	42.5	46.1	.52	<.001	<.001
Previous MI, %	3.9	5.4	4.3	<.001
Anterior MI, %	25.9	23.3	20.5	<.001

Note. MI = myocardial infarction.

Revascularization and Mortality

Among the study group as a whole, as well as in the case of Whites, those with supplemental private insurance coverage exhibited the highest rate of PTCA use, followed by those with supplemental public insurance; patients who had only Medicare coverage exhibited the lowest rate of PTCA use (Table 2 ▶). Among Blacks and Hispanics, those with supplemental private insurance exhibited the highest frequency of PTCA, while there were no differences between those with Medicare coverage only and those with Medicare and supplemental public insurance coverage. This was also the case for overall rates of CABG use, as well as race/ethnicity- and gender-specific rates (Table 2 ▶). Altogether, revascularization was most frequent among Hispanics and least frequent among Blacks, and this was particularly true in the case of PTCA.

TABLE 2—

Percentages of Patients Undergoing PTCA or CABG, by Race, Gender, and Insurance Status

				Group Comparisons (P)
	Medicare Only (Group 1)	Medicare With Private Insurance (Group 2)	Medicare With Public Insurance (Group 3)	1 vs 2	1 vs 3	2 vs 3
PTCA
Overall	7.7	13.3	9.0	<.001
White men	7.6	12.9	9.9	<.001
Black men	4.6	10.2	4.7	<.001	.43	<.001
Hispanic men	8.8	15.4	8.9	<.001	.53	<.001
White women	5.2	10.2	5.9	<.001	.01	<.001
Black women	4.8	7.9	4.6	<.001	.29	<.001
Hispanic women	6.4	18.3	6.6	<.001	.61	<.001
CABG
Overall	7.1	10.2	7.2	<.001	.37	<.001
White men	8.4	11.6	8.6	<.001	.12	<.001
Black men	4.4	7.4	4.3	<.001	.45	<.001
Hispanic men	6.0	11.6	6.3	<.001	.29	<.001
White women	4.9	7.3	4.8	<.001	.29	<.001
Black women	4.1	6.4	4.1	<.001	.67	<.001
Hispanic women	5.4	9.5	5.0	<.001	.13	<.001

Note. PTCA = percutaneous transluminal coronary angioplasty; CABG = coronary artery bypass grafting.

In-hospital mortality was highest among patients with only Medicare insurance coverage, followed by those with supplemental public insurance; those with supplemental private insurance exhibited the lowest in-hospital mortality. When stratified according to race/ethnicity and gender, in-hospital mortality was highest among Whites, lowest among Hispanics, and higher among women than among men (Table 3 ▶).

TABLE 3—

In-Hospital Mortality Percentages, by Race, Gender, and Insurance Status

				Comparisons (P)
	Medicare Only (Group 1)	Medicare With Private Insurance (Group 2)	Medicare With Public Insurance (Group 3)	1 vs 2	1 vs 3	2 vs 3
Overall	17.0	14.0	16.1	<.001	.02	.002
White men	16.9	13.7	14.5	<.001	<.001	.04
Black men	15.6	11.7	13.8	<.001	.003	.01
Hispanic men	13.4	8.7	11.3	<.001	.01	.004
White women	20.6	15.6	18.4	<.001	.01	<.001
Black women	19.6	13.2	17.5	<.001
Hispanic women	16.6	13.4	15.0	.001	.03	.004

Associations between insurance status, gender, race/ethnicity, and revascularization procedures (PTCA and CABG), as well as in-hospital mortality, were explored through logistic regression analyses controlling for other characteristics such as age, comorbid conditions (diabetes, hypertension), length of hospital stay, presence of congestive heart failure, admission status, previous myocardial infarction, and location of myocardial infarction (Table 4 ▶). Overall, both gender and race were significantly associated with procedure use and in-hospital mortality; Whites and men were more likely to undergo procedures than Blacks and women. However, there was no difference between Whites and Hispanics, although univariate analyses showed that Hispanics had higher revascularization rates and lower in-hospital mortality rates.

TABLE 4—

Procedure Use and Mortality Odds Ratios, by Race, Gender, and Insurance Status: Logistic Regression Models

	Odds Ratio (95% Confidence Interval)
	Model 1: PTCA	Model 2: CABG	Model 3: Mortality
Overall
Race
Black vs White	0.56 (0.50, 0.62)	0.44 (0.40, 0.50)	1.19 (1.11, 1.27)
Hispanic vs White	1.01 (0.92, 1.10)	0.69 (0.62, 0.77)	0.96 (0.89, 1.03)
Gender (men vs women)	1.14 (1.08, 1.20)	1.55 (1.46, 1.64)	0.89 (0.85, 0.93)
Insurance
Private plus vs Medicare only	1.69 (1.60, 1.79)	1.53 (1.44, 1.63)	0.77 (0.73, 1.80)
Public plus vs Medicare only	1.05 (0.97, 1.14)	1.05 (0.95, 1.16)	0.95 (0.90, 1.01)
Non-Hispanic Whites
Gender (men vs women)	1.15 (1.09, 1.22)	1.60 (1.51, 1.70)	0.90 (0.86, 0.94)
Insurance
Private plus vs Medicare only	1.85 (1.75, 1.97)	1.78 (1.67, 1.89)	0.75 (0.72, 0.79)
Public plus vs Medicare only	1.18 (1.07, 1.30)	1.09 (0.98, 1.21)	0.95 (0.89, 1.02)
Non-Hispanic Blacks
Gender (men vs women)	0.96 (0.80, 1.16)	1.06 (0.86, 1.30)	0.87 (0.76, 0.99)
Insurance
Private plus vs Medicare only	1.74 (1.41, 2.15)	2.00 (1.58, 2.53)	0.78 (0.66, 0.92)
Public plus vs Medicare only	0.92 (0.73, 1.16)	0.83 (0.64, 1.08)	1.07 (0.93, 1.23)
Hispanics
Gender (men vs women)	1.05 (0.88, 1.25)	1.36 (1.10, 1.68)	0.80 (0.69, 0.93)
Insurance
Private plus vs Medicare only	1.88 (1.48, 2.38)	2.03 (1.51, 2.72)	0.72 (0.55, 0.93)
Public plus vs Medicare only	1.01 (0.84, 1.23)	0.90 (0.71, 1.14)	0.87 (0.75, 1.02)

Note. PTCA = percutaneous transluminal coronary angioplasty; CABG = coronary artery bypass grafting. In model 1, PTCA was the dependent variable; in model 2, CABG was the dependent variable; and in model 3, in-hospital death was the dependent variable. Other model variables were as follows: age, diabetes comorbidity, hypertension comorbidity, length of hospital stay, complications of congestive heart failure, previous myocardial infarction, admission status, and location of myocardial infarction.

The univariate analyses showed that Whites had the highest in-hospital mortality rates. After adjustment for other characteristics, there was no difference in mortality rates between Whites and Hispanics, and the rates among both of these groups were lower than those among Blacks. In comparison with Medicare-only patients, those with supplemental private insurance exhibited significantly more frequent revascularization and lower in-hospital mortality rates; those with supplemental public insurance exhibited revascularization and in-hospital mortality rates similar to those of the Medicare-only group. Stratification according to race/ethnicity revealed similar within-group associations of insurance status with revascularization use and in-hospital mortality (Table 4 ▶).

DISCUSSION

The principal finding here was that most (57.7%) Medicare recipients (at least those hospitalized for acute myocardial infarction) have supplemental insurance (either private or public). However, only supplemental private insurance seems to affect type of hospital care and short-term survival. The exception to this generalization is that, among Whites, those with supplemental public insurance were more likely to undergo PTCA than those with only Medicare coverage. Furthermore, after adjustment for insurance status and risk factors, racial differences in cardiac procedure use persisted. In view of the wide disparities in insurance coverage among younger Americans,9,12 these data indicate that universal availability of health insurance does not eliminate socioeconomic disparities in health service use and outcomes for older Americans.

Previous studies have repeatedly shown that use of medical care in the United States is related to health insurance status.13,14 In the case of the elderly, Medicare provides a basic level of access to health care. Still, one study showed that individuals who have Medicare and supplemental public insurance coverage and individuals who have Medicare coverage only are twice as likely as those who have supplemental private insurance to have unmet medical needs (including needs related to immunization, dental care, prescription medicine, eyeglasses, and mental health).15 However, this investigation did not address in-hospital service.

The present study showed that Medicare cannot universally provide the health care needed by hospitalized myocardial infarction patients. Therefore, our initial hypothesis—that Medicare coverage supplemented by private insurance would not result in improvements in regard to either revascularization or in-hospital mortality rates among patients hospitalized after a myocardial infarction—was not confirmed. Moreover, the fact that the groups who, for the most part, lacked private supplemental insurance coverage were Blacks, Hispanics, and women suggests that other sociodemographic differences existed among patients with private insurance coverage.

Previous research on treatment of myocardial infarction has shown that patients with private insurance are more likely to undergo revascularization than those with Medicare, Medicaid, or no insurance.13,16 Such analyses are often confounded by age because, by definition, Medicare beneficiaries are older than 65 years. It has been reported that older patients are less likely to be reperfused than younger patients.17 A report of the National Registry of Myocardial Infarction showed that Medicare/Medicaid patients, in comparison with privately insured patients, undergo fewer reperfusion procedures, undergo fewer invasive cardiac procedures, and have longer hospitalizations. However, in this comparison, which included all age groups, Medicare patients were older and of more advanced clinical status.18

The fact that differences in PTCA and CABG rates use between Whites and Blacks persisted in each insurance group and among both men and women suggests that even with universal medical insurance in place, individuals who are socioeconomically disadvantaged are less likely to receive the services they need. In fact, the differences in revascularization use among White and Black patients observed here were generally similar to those reported in a previous study involving the overall adult population (i.e., individuals 35 years or older); in that study, White/Black ratios in regard to PTCA and CABG use were 1.63 and 1.55, respectively.19

The finding that patients with Medicare-only coverage were more likely than patients in the other insurance groups to remain in the hospital more than 7 days was consistent with an earlier report indicating that, in comparison with patients with good access to health care, those with poor access exhibit significantly longer hospital stays and poorer health outcomes in regard to the condition requiring hospitalization.20 Moreover, Medicare patients with supplemental private insurance were more likely than patients with Medicare only and those with supplemental public insurance to have had a previous myocardial infarction, suggesting in turn that these patients were more likely to survive the first event. Overall, women underwent fewer revascularization procedures than men, and Blacks underwent fewer procedures than White or Hispanic patients, regardless of insurance status. These associations persisted after adjustment for sociodemographic characteristics, comorbid conditions, complications, length of hospital stay, previous myocardial infarction, admission status, and location of myocardial infarction.

While in-hospital mortality rates were highest among Whites and lowest among Hispanics, adjustment for other characteristics revealed that Whites and Hispanics exhibited similar in-hospital mortality, and the rates for both of these groups were lower than those among Blacks. However, gender-specific mortality differences persisted after adjustment for other characteristics.

The strengths of SPARCS are its large size, standardized data collection methodology, and individualized database. However, this study was limited to the administrative database, which provides minimal clinical details and does not include diagnostic study results or information on use of thrombolytic therapy or particular medications. No long-term follow-up information is available. Also, our study included only patients admitted to New York City hospitals, limiting extrapolation to the overall US population. In addition, federal, military, Department of Veterans Affairs, and institutional hospitals were not included. Therefore, our results cannot be generalized to all hospitals in the United States.

In summary, the results of this study suggest that purchase of supplemental private health insurance leads to increased use of specialized medical care and hospital survival among Medicare patients hospitalized with acute myocardial infarction. In view of the marked insurance-linked disparities in services and outcomes observed among younger Americans, it seems likely that universal health insurance coverage involving Medicare alone will not reduce health disparities among older Americans. Moreover, since sociodemographic characteristics, including gender and race, continue to be associated with receipt of health services as well as overall in-hospital mortality, it is clear that equalizing insurance coverage for hospital services through Medicare does not, in itself, eliminate all health disparities.