Women, African-Americans and Medicare-covered patients hospitalized with liver-related conditions are less likely to be evaluated and referred for a liver transplant, according to a University of Pittsburgh School of Medicine study.

The study, published in the September issue of the American Journal of Transplantation, looked at 144,507 patients hospitalized in Pennsylvania with liver-related conditions between 1995 and 2001 and tried to determine if any potential barriers existed at the referral stage.

The study found that 4,361 of those patients underwent transplant evaluation. Of that group, 3,071 were wait-listed and 1,537 went on to have a transplant. Another 57,020 died during the study period.

According to the study, 61 percent of men were evaluated for transplantation compared to 39 percent of women; 73.8 percent of whites were evaluated compared to 8.6 percent African-Americans; and 62 percent of patients with commercial insurance were evaluated compared to 4.7 percent with Medicare only.

The lead author said the researchers "cast a wide net" to determine if inequities existed in the early stages of the transplantation process by including "transplant-potential" people in the research.

"What we found is that from the point of hospitalization, where people are diagnosed with liver-related conditions, gender, race and insurance status play a big role in who is evaluated and listed [for transplant], much more so than we'd see in UNOS data from listing to transplant," said Cindy L. Bryce, Ph.D., study lead author and associate professor of medicine at the University of Pittsburgh.

UNOS is the acronym for the United Network for Organ Sharing, which oversees the national database of clinical transplant information.

"What the good news is, in the latter stages, where UNOS provides oversight, we don't see disparities," Dr. Bryce added. "Organs go to needy candidates, and sociodemographics -- gender, race, insurance -- don't play a role in who gets an organ at that stage. But it still appears to play a role early on in who gets to see a transplant center for evaluation."

The study was not designed to identify causes for the disparities, Dr. Bryce said.

"We don't know all the reasons for the differences," she said. "Practices vary from hospital to hospital if there aren't guidelines, so that might be part of the problem. And we also don't know what role patient preferences play."

The study was done by linking several different data sources, the main one being hospitalizations in Pennsylvania, Dr. Bryce said. Data also came from transplant centers in the commonwealth, UNOS data, and death certificate data from the Department of Health.

"What we were trying to do was identify people through hospitalization data who might need a transplant at some point," she said. "We cast a wide net."

The researchers identified people who were hospitalized with not just liver disease as a primary diagnosis but liver-related conditions.

"That's why it goes from 144,000 patients to 4,000," Dr. Bryce said. "[It's] not because 144,000 should have been evaluated for liver transplant.

"The problem is if you don't cast a wide net and miss them early, you can never get them back. We called them 'transplant-potential.' They were the people we followed for a long time."

Dr. Bryce said researchers sought to gain the kind of information on liver patients that is already available regarding potential kidney transplant patients, gathered earlier in their care.

"It's different because patients who go on dialysis are covered by Medicare, and Medicare tracks them from that point on. So there's never going to be a Medicare system for liver transplant, but it made us curious as to what we would find in liver."