Some aspects of the day-to-day treatment of critically ill patients in hospitals have been more an art than a science. But techniques developed by LDS Hospital may change that.
The hospital is exporting a computerized system it developed to help treat adult respiratory distress syndrome, a life-threatening condition in which the patient has difficulty breathing.LDS Hospital started building its computer data base 30 years ago. Beginning in 1988, it put instructions on computers about how to treat ARDS.
Doctors and nurses no longer have to fish through many pages of flow diagrams to find recommendations for the next step of treatment.
Instead, instructions "come up on their bedside terminal," says Thomas D. East, a researcher who helped develop the protocols. Computers are stationed in every hospital room, where physicians use them regularly.
Health experts update the computers with the latest information on an ARDS patient's ventilator settings and other data as part of their regular chart notes.
The computer analyzes the information using a data base that the hospital has built up about ARDS, said hospital spokesman Jess Gomez.
Then up on the terminal in the patient's room pops a record of all decisions proposed for treatment, plus other information, he said.
For example, if a blood test shows that a patient with respiratory distress has a low amount of oxygen in the blood, the computer will suggest increasing the percentage of oxygen administered and increasing the amount of pressure to help keep the lungs open.
The protocols seem to work well in Utah. LDS Hospital used them to improve the chances of its ARDS patients, according to a study that ended in 1991.
For people with less severe cases, average survival rate is 50 percent, but using the protocols, LDS Hospital increased this to 63 percent. For more severe cases, the average survival rate was between 9 percent and 15 percent. Gomez said the new survival rate is 43 percent.
But doctors and administrators wondered if the protocols would work in a different environment than the relatively calm setting of a hospital in Salt Lake City - say, an inner-city emergency ward where ARDS is more likely to be caused by a gunshot wound than by pneumonia.
In an experiment funded by the federal Agency for Health Care Policy and Research, hospitals in Los Angeles, Houston, St. Louis and Pittsburgh are trying the protocols. The program is funded with a $775,000 grant for two years, and a request is pending for $2 million for an additional three years.
"We don't know for sure" if the improved survival is because of the protocols or some other factor, East said.
"The feeling among the clinicians here is that they really like it. . . . They're convinced here that it has a value."
But until now, no scientific study has been completed to show whether the protocols actually do work, or how they can be adapted to other settings.
One of the hospitals in the trial - Los Angeles County-King/Drew Medical Center, which joined the study in February - "treats many more traumas, such as gunshot wounds and stabbings, than we do," East said.
To expand the study, programmers had to adapt the protocols to different computer systems. At LDS Hospital, the terminals are tied in to a mainframe computer, but some of the hospitals in the study had no such overall system.
"We had to get the protocols onto a computer that could be used in these centers," East said. PCs were put at every bedside in the special intensive care units at King/Drew that are used in the study.
Siemens Medical, a Swedish corporation, loaned 12 ventilators to the hospital. The ventilators are moved on carts, which have extra shelves.
"The computers sit on the shelves" by the patients' bedsides, he said.
So far, too few patients have been enrolled in the study to answer whether the protocols save lives in other settings. The study aims at gathering information on 400 patients before drawing conclusions. So far, it has only about 33 patients.
But the study is already answering some tough questions, such as how other hospitals would accept the computerized protocols.
"The acceptance was great," East said. "We've had over 95 percent of the instructions generated followed at the bedside."
A related question is whether it's possible to standardize critical care. And to date, the results indicate it's "possible to take the same standards of care and transfer them to very different hospital settings," he said.
Dr. Alan Tonnesen, professor of anesthesiology at the University of Texas Medical School, Houston, said the school's Hermann Hospital has been enrolling patients in the study since August.
"The first steps were just getting ourselves and all of the nurses and respiratory therapists familiar with how to operate the computer - that it was OK to follow the instructions the computer was putting out," he said.
Tonnesen said some at Hermann Hospital were skeptical or had what could be called a sort of bias, "in the sense of different groups of physicians evolving into different ways of managing patients with ARDS." It was somewhat different than the way patients are managed at LDS, he said.
"But certainly the data that they had indicated that their method was probably as good as our method, so we thought it was a good first step here to help out and see whether the computerized control mechanism would actually work."
Following the patient closely, the clinicians found "that the system nearly always comes up with a reasonable next step. So we've been able to follow nearly all of the computer's instructions."
When the computer generated an instruction that didn't seem right to doctors at Hermann, "most of the time . . . we'd go back and find that the computer really didn't have the most current information available" about the case.
The system will help to improve management, Tonnesen said. Until now, treatment hasn't been as consistent as it will be. Regardless of the details of the protocols, the new system will give the advantage of systematized treatment.