Interview: Gender, Race Gaps in ER Care

By CHRISTINE DELL’AMORE

Many studies have focused on racial and gender differences in healthcare, but few have investigated whether these disparities exist in the emergency room. That’s why Liliana Pezzin, now of the Medical College of Wisconsin in Milwaukee, and Penelope Keyl and Dr. Gary Green, both of Johns Hopkins University, decided to research how race, gender and insurance differences factored into the care of a person in emergency care for chest pain. The research team drew data from the U.S. National Hospital Ambulatory Healthcare Survey of emergency departments between 1995 and 2000.

The study, published in February’s Academic Emergency Medicine, found disparities in race, gender and insurance in how four standard diagnostic tests were administered. More surprising to the researchers, such differences worsened during the five years of the study.

But the numbers may not reflect a healthcare worker’s conscious decision to discriminate — rather, it’s a very complex interplay of communication and perception between healthcare workers and patients, and how they describe symptoms, Green said.

United Press International also discussed the research with lead author Pezzin:

Q. Why did you begin investigating this topic?

A. At the time when we were doing this research, (Keyl and Green and I) came across racial differences in cardiac treatments for patients in hospitals. This is not a new area of study, (but) we started discussing whether this happens in emergency settings. We decided to pursue this as a joint topic to look for data to answer that question.

The question is more unique than in other studies, where everyone has been screened, diagnosed and referred. What happens to people who are not in the system — patients who are showing up for the first time with a symptom of chest pain? How are they are treated? The four tests we chose (electrocardiography, cardiac monitoring, oxygen saturation measurement with pulse oximetry and chest radiography) are non-invasive and standardized, and (thus) should be offered to anyone.

Q. Why focus on those four tests?

A. Guidelines recommend those tests for anyone over the age of 30 who presents with chest pain. They are relatively inexpensive and easy to ascertain diagnostic tests. They are also commonly used — in most places they are in standing orders, (which means) at the front desk this test can be ordered (before) a physician’s first contact with a patient.

Q. Have other studies looked at this topic before?

A. On emergency departments, (not much). The novelty is this really applied to people who are presenting with chest pain in nationally representative databases, as opposed to data from small, single hospitals. (Those studies are interesting), but they’re only representative of the experience of that hospital. (Our) data corresponds to 32 million emergency department visits throughout the United States. There are drawbacks to using secondary data, but we’re able to correct for a wide range of potential confounders. We controlled for age; the type of hospital, such as urban or rural; and the region of country, (among others).

Q. What are the most striking findings?

A. Consistently, we found African-American males are less likely to receive any of the tests we considered. The more striking result is we analyzed both one point in time and time trends, and with the exception of pulse oximetry, all the other tests showed a widening racial and gender disparity over time. So, it not only exists, it’s getting worse over time. It’s the combination of finding the disparities and that they are widening over time. There are a lot of arguments used in the past about why (there would be) lower rates of testing among African-Americans; (for instance, some say) they present atypically or their symptoms are different.

This argument is not used for insurance status, (however), and there is an enormous disparity in the rate of ordering these four tests among those who are covered by Medicaid. If insurance is a proxy for (showing symptoms), that’s an important finding — it goes a step beyond the focus on race and gender, and (it’s) more of a focus on socioeconomic status and underinsurance as a potential source of missed diagnosis for cardiac problems.

Q. What are potential explanations for your findings?

A. Some people say it’s not under-utilization (of healthcare) by African-Americans, it’s over-utilization by whites. But African-Americans have worse outcomes later on, so that argument doesn’t work. It’s increasingly difficult to justify differences based on presentation (of symptoms).

African-Americans and non-Africans have no difference in way they present. (Furthermore), what you would attribute as justification for not doing the tests don’t make sense for insurance status. The reimbursement rate (for Medicaid) is poor relative to commercial insurance. Specifically, with all the emergency department overcrowding, maybe people are getting more selective upfront as to who gets the test based on a perceived need, or where a perceived reimbursement will come from.

Q. Why is this study important?

A. We are not making the case for provider bias — we cannot make that assessment — but we have observed that a (large) difference exists at a much earlier point in contact with the health system that has been observed up till now. It’s something that requires attention, and it cannot be ignored. If (people are) not getting diagnostic tests, they cannot be diagnosed.