The Arkansas Health Department is starting to receive patient discharge data from hospitals around the state, but the information will be of little or no use to patients or payers.
That's because the state Legislature attached some significant strings when it wrote the 1995 law authorizing the data collection. The measure prohibits the department from linking the names of specific hospitals to any information it releases.
Instead, the health department will produce general reports analyzing variations in access, diagnosis, procedures performed, patient demographics and rates of illness.
"Hospitals account for 39% of all healthcare expenditures in this country," said Douglas Murray, director of the Arkansas Health Department's Center for Health Statistics. "It's a gold mine of information with significant public health implications."
Arkansas lacks a state healthcare database, said Paul Cunningham, vice president of the Arkansas Hospital Association. As one example of the kind of information that will be accessed, he pointed to the record of every discharge from every hospital recorded on the federal UB92 form, which will enable the state to highlight clusters of cancers, heart disease and other health problems.
However, consumers who want cost or quality information derived from the state reports won't be able to get it. Murray said he is aware that many states now publish comparative data on hospital outcomes and costs. New York even publishes physician outcomes data by name.
But he said he didn't think that was a good idea or appropriate for Arkansas.
"Many hospitals in Arkansas are having a difficult time," Murray said. "They feel sensitive about release of information that could damage their ability to provide services to their communities. Hospitals are very often the largest single employer in their community, with the possible exception of the public schools."
Murray said the Joint Commission on Accreditation of Healthcare Organizations ensures quality of care. Cunningham said the hospital association was leery of publishing quality or outcomes data. "We go back to the mortality information that at one time was published by HCFA. We felt, as did all the other hospital organizations, that that information was subject to gross misinterpretation."
Arkansas has one of the lowest rates of managed-care penetration of any state, with HMOs covering just 6.4% of the population, according to the Arkansas Hospital Association. Arkansas Blue Cross and Blue Shield is a 50% shareholder in the state's largest HMO, HealthAdvantage, which insures 56,000 people. The other shareholder is Baptist Medical System, which operates the largest hospital in the state.
"We're certainly trying to work on cost control in the healthcare industry," said Blues spokesman Patrick O'Sullivan. But he said the Blues has no position on publishing quality or cost data from the state's hospitals.
Because the managed-care industry is so undeveloped in Arkansas, there is no managed-care association to counterbalance the influence of the hospital lobby in Little Rock, as happens in other states.
Cunningham said he wasn't sure whether the association had ever taken a position on publishing hospital charge data. "I can't see any good that would come of it," he said. "Charges in any individual institution are subject to a number of factors. You could do a dissertation on charge structures....It's based on patient mix, severity from one year to the next, cost structure-there are lots of things that go into it. You're mixing apples and oranges in many cases."
Murray, the state statistician, took much the same approach to the question of publishing charge data. "Are you suggesting that hospitals gouge the patients? I don't have any evidence to support that," he said.
The data collection law arose from a task force report compiled in 1993-1994. Roger Busfield, president emeritus of the Arkansas Hospital Association, served on the task force, which suggested issuing hospital report cards.
"We did recommend that report cards would be appropriate," Busfield said, "but you've got to learn to crawl before you learn to walk." The task force did not want individual hospital names attached to the measurements.
"We were concerned that they would come out with horrible distortions of mortality rates, which the federal government did, scaring the hell out of the people going into the hospital," he said.
Busfield said Arkansas is such a rural state that only three or four cities have more than one hospital. Therefore, a patient's identity could possibly be traced from hospital-specific discharge data.
Medicare cost reports reveal what hospitals are spending, Busfield said. What they do not want to reveal is their sources of revenues, especially nonpatient revenues. "If you've got money coming in from bequests and portfolios, you don't want that publicized very much," he said. "I don't blame administrators for that.
"What they're getting for rental, parking, cafes, cleaning establishments-anything operated for profit that has nothing to do with patient care, you can almost see regulators rubbing their hands and saying, `You don't need all this money from Medicaid; you can subsidize it,'*" Busfield said.
In neighboring Missouri, by contrast, the state health department has been publishing both outcomes and charge data. Missourians can get consumer guides on obstetrics, outpatient surgery and emergency care that list hospitals by name.
"The philosophy here is that consumers certainly have a right to know outcomes-generated quality data so they can make informed decisions to the extent they have ability to do that," said Leslie Speake, vice president of patient care at Ozarks Medical Center in West Plains, Mo.
"Second, the benchmarking aspect for hospitals, when we have uniformly collected data, is of value to the hospitals," she said.
The Show Me Buyer's Guide on obstetrics, published in April 1994, offers statistics on Missouri's length of stay, Caesarean-section rates, patient satisfaction, neonatal deaths expected and observed, and 11 other measures. It also lists patient charges.
The booklet includes comments by some of the hospitals in the survey. They point out discrepancies and errors of interpretation in the data, and they try to explain why certain scores may appear to be lower than they should be.
Ozarks, for one, had 19 neonatal deaths when 12 should have been expected. That prompted a letter of explanation from Speake.
She wrote that the way statistics were collected and adjusted made Ozarks look much worse than it really was. She argued that certain kinds of infant deaths, such as sudden infant death syndrome and "acts of God," should have been excluded and that socioeconomic status should have been included as a variable.
Ozarks, a 120-bed hospital just 17 miles north of the Arkansas line, serves a rural area with few physicians, high illiteracy and low incomes.
"On the OB data released to the public, which was good in its intent, the reason we had written is that nine of the 19 deaths were inappropriately reported," Speake said in an interview. "That may be some of the crux of what Arkansas is dealing with. It's difficult to have confidence in the data unless it is extremely clean and we're doing apples-to-apples (comparisons) across all hospitals."