Escarce JJ, Epstein KR, Colby DC, Schwartz JS.
Racial differences in the elderly's use of medical procedures and diagnostic tests.
Am J Public Health 1993:83(7):948-54.

Although several studies have reported race differences in the use of certain procedures among the elderly, it is still uncertain whether these findings are isolated examples or whether they represent a systematic pattern of lower use of specialized or high-technology medical services by elderly blacks compared with whites. The purpose of this study is to examine differences between elderly blacks and whites in the use of a wide range of medical procedures and diagnostic tests in a Medicare population. The authors further examined whether differences persist among elderly persons who have Medicaid in addition to Medicare and whether similar differences exist for rural versus urban residents.

Data for this study were derived from the Health Care Financing Administration's databases (1986), which include patient information for a random sample of 5% of Medicare enrollees in the United States. Only persons aged 65 years and older were included in this sample. Thirty-two medical procedures and diagnostic tests were selected for the analysis.

Whites were more likely than blacks to receive care from a physician during the study year. Whites were also more likely than blacks to receive 23 of the 32 study services. The largest disparities were found in coronary bypass, coronary angioplasty, and carotid endartectomy, for which the white-black RR was greater than 3. In contrast, for 7 of the 32 study services, blacks were more likely than whites to receive the service. Large differences were noted only in the three ophthalmologic procedures that are used to treat primary open-angle glaucoma or proliferative diabetic retinopathy. (There was no racial difference in the remaining two study services.)

The authors selected seven pairs of services to contrast higher- versus lower-technology services (e.g. radionuclide stress test versus exercise stress test). In three of the seven pairs, the white-black RR was higher for the high-technology service than for the low-technology service. This finding suggests that whites might have a particular advantage in terms of access to the higher-technology services.

For most of the study services, white-black relative risks were similar among urban and rural elders. However, for 12 services, the relative risks were higher among rural than urban elders. The differences were largest in the South. Additionally, when the study sample was restricted to those who also received Medicaid (the least affluent sub-sample who also have the lowest out-of-pocket responsibilities), whites were more likely to receive 20 of the 32 services. The magnitude of effect was slightly lower. In contrast, blacks were more likely to receive 4 of the study services (among Medicare-Medicaid recipients).

Although severity of illness could not be evaluated, the authors argue that the study services are for conditions that do not vary in prevalence between blacks and whites, meaning that racial patterns in illness characteristics are unlikely to explain the current study findings. (However, this is not true for glaucoma, as blacks have higher rates of this illness, and one might expect differences in procedures perhaps even higher than those observed under conditions of equal access to treatment for blacks and whites.) Additionally, since this study demonstrated that the racial difference in the proportion of elders who received care from a physician was small compared with the racial difference in the use of the services studied, it is unlikely that access to care completely explains the observed disparities in procedure use. The authors also suggest that financial barriers not controlled in the subgroup evaluation of Medicare-Medicaid recipients might explain these results. Finally, bias (physician and institutional decision making), and patient preferences (preferences, health beliefs and expectations) were suggested as possible explanations.

Limitations of this study include lack of information on illness and income characteristics, quality of care, unmet needs, and patient outcomes.