Andrews RM, Elixhauser A.
Use of major therapeutic procedures: are Hispanics treated differently than non-Hispanic whites?
Ethn Dis 2000;10(3):384-94.

Previous research suggests that disparities in medical care of Hispanics compared with non-Hispanic whites may be similar to those disparities found between blacks and whites. This study aimed to determine whether there is evidence to suggest such differences, examining a wide variety of medical conditions and patients treated in three U.S. states (California, Florida, and New York). Data for this study were drawn from the 1993 Healthcare Cost and Utilization Project State Inpatient Databases, which contain all discharges from acute care hospitals in selected states. The study included patients 17 years of age and older whose ethnicity was listed as Hispanic or non-Hispanic white. (It is important to note that the Hispanic populations across the target states are highly heterogeneous.)

Notable differences in sociodemographic characteristics between Hispanic and non-Hispanic whites included expected primary pay source (non-Hispanic whites were more likely than Hispanics to use Medicare - 55.7% versus 31.4% - and less likely to use Medicaid - 4.8% versus 24.8%), household poverty rates (non-Hispanic whites were less likely than whites to live in an area with a high poverty rate - 8.7% versus 37.9%) and type of hospital used (non-Hispanic whites were less likely to be treated in government-owned or teaching hospitals than Hispanics).

The authors then compared non-Hispanic whites with Hispanics with regard to being discharged with one of 63 different diagnoses and receiving a major therapeutic procedure for the diagnosis. All comparisons included adjustment for sociodemographic characteristics. Among diagnosed patients, for 38% of these conditions, Hispanics were less likely than non-Hispanic whites to receive a major therapeutic procedure. The conditions include coronary artery disease and diseases of the aortic valve; breast, colon, cervical or lung cancer or metastatic disease; some traumatic conditions (herniated intervertebral disk; fracture of the tibia; ankle injury; fracture, dislocation or sprain of the shoulder or head of humerus; bursitis; and open wound, blunt trauma of upper extremity); some gastrointestinal conditions; epilepsy; cerebrovascular disease; pelvic inflammatory disease; cirrhosis of the liver; osteoarthritis; peripheral vascular disease; and abdominal aneurysm.

For 6.3% of these conditions, Hispanics were more likely than non-Hispanic whites to receive a major therapeutic procedure. These conditions included renal failure, trauma to chest wall or breast, pancreatitis or diabetes mellitus.

The authors note that the difference in receiving major therapeutic procedures for coronary conditions between Hispanics and non-Hispanic whites mimicked the previously reported black-white pattern, whereas the difference in receiving procedures for renal disease contrasted with the black-white pattern. Although the finding that Hispanics were less likely to receive procedures for certain cancers might be partially explained by differences in severity of disease at presentation, the fact that this pattern persisted even in patients with metastatic disease makes this explanation unlikely. However, with regard to other illness, especially traumatic illnesses, the authors note that these differences might likely be explained by differences in the nature and severity of illness that were not controlled in the analysis.

The authors conclude that this study confirms previous research reporting disparities between Hispanics and non-Hispanics for coronary heart disease and conditions affecting the gastrointestinal and musculoskeletal systems. It was not possible from this study to determine whether Hispanics were receiving too few procedures or non-Hispanic whites were receiving too many procedures, but these findings do "raise concerns about appropriateness of care and access to needed services." The authors note that previous researchers postulated a number of possible explanations for race/ethnic patterns in major therapeutic procedures, including unmeasured clinical differences, differences in risk or treatment efficacy, differences in the practice patterns and types of providers, sociocultural differences that may affect patient preferences for treatment and communication between patient and provider, and ethnic stereotypes or physician bias. Again, it was not possible from this study to determine which, if any, of these explanations were operating..