The objective of this study was to examine variation in angiography procedure use for a U.S.-wide cohort of patients with regard to sociodemographics, geographic region (West, Midwest, South and Northeast) and procedure availability, as well as the interrelations among these factors.

Data for this study were derived from claims and administrative data from the Health Care Financing Administration. The cohort consisted of a Medicare population of patients at least 65 years of age who were hospitalized for acute myocardial infarction in 1987. Patients were excluded if they belonged to an HMO, were discharged from the hospital after less than 5 days, or had an acute myocardial infarction (AMI) during the previous year. This study evaluated the number of patients receiving angiography during the 90 days following the AMI. Variables of interest included age, gender, race, comorbidity (an index), region of residence, and regional availability of coronary angiography.

The propensity for cardiac catheterization, accounting for national averages, was 0.09 (95% confidence interval=0.08 to 0.11) to 0.28 (95% confidence interval=0.26 to 0.30). The national median was 0.17. Those in the Northeast states had the lowest probability of cardiac catheterization.

The probability of angiography was influenced by patient age, gender, race and comorbidity in each state. The magnitude of these effects varied across states, as did the underlying rate of angiography after adjusting for patient characteristics. Interstate differences displayed clear geographic patterns and were also related to the availability of angiography. That is, states with high-onsite availability had more angiographies performed. However, the level of procedure availability did not seem to be related to age, gender or race differentials within states.

Angiography rates were lower for blacks than non-blacks. The percentage of blacks receiving the procedure ranged from a minimum of zero in several states to a maximum of 62% in Washington DC, with a median of 2.8%. There were important geographic differences in the effect of race on the probability of angiography, with greater equality between the races in the northern and midwestern states. Race differences were more pronounced in the South. The adjusted black to non-black odds ratios for angiography ranged from a low of 0.41 (95% confidence interval=0.30 to 0.54) in Mississippi and to a high of 0.94 (95% confidence interval=0.55 to 1.59) in Kansas. Gender and age differences, in contrast, showed small variation around the national trend.

The authors conclude that “the existence of large variation in procedure use for a patient cohort with relatively uniform insurance coverage is more evidence that insurance coverage is only one of the determinants of access to medical technologies. Sociodemographic factors, medical practice styles and other regional factors [also] play an important role in determining access to care…”