Study Links 9 Common Medical Errors to Deaths
Nine common medical errors--ranging from prescribing the wrong drugs to misdiagnosis of stroke--appear to have caused as many as 27% of all deaths from heart attack, stroke and pneumonia among hospital patients in a 12-hospital study released today.
The study by two RAND Corp. researchers working with the cooperation and financial support of American Medical International of Beverly Hills is believed to be the first systematic examination of preventable hospital deaths from common conditions.
“It’s an indictment of the medical care system if it’s generalizable,” said Dr. William R. Fifer, a consultant to AMI, one of the nation’s largest hospital management companies. “What we don’t know is if this is atypical. We have no way of knowing.”
Cases of 182 Patients
The researchers, using AMI hospitals in the Southern and Western United States, studied the cases of 182 patients who had died of heart attack, stroke or pneumonia. Panels of experienced specialists then examined medical records to determine which deaths were preventable.
A majority on each panel concluded that 27% of the deaths could have been prevented--that is, that poor care at that hospital led to the patient’s death. The panels voted unanimously in the case of 14% of the deaths.
“I think what we found was that (the percentage) was a surprisingly high number,” said Dr. Robert W. Dubois, a co-author of the report. “I guess if you asked the typical physician, they might say something in the neighborhood of 1% to 4%.”
The researchers say their work can serve as a guide for hospital officials wishing to reduce rates of preventable death. Eventually, their approach might also help in providing consumers with information on which to base decisions about hospital care.
However, the researchers cautioned that their findings are preliminary and need to be explored through further study. They said it is impossible to know whether the 12 facilities they examined are representative of other hospitals nationwide.
AMI officials said in a statement Friday that “similar findings would almost certainly result from a sampling of any group of . . . hospitals.” However, they said the study had proved useful in identifying areas of their operations that “demanded immediate evaluation.”
The statement said AMI underwrote the RAND study in recognition of the need for a reliable system of monitoring patient care. AMI also began using in 1986 a computer-based patient screening program, which AMI said is unique in the hospital industry.
The two-year RAND study, published in Annals of Internal Medicine, was intended to determine the frequency of preventable deaths and their causes. The researchers also explored whether hospitals could identify in advance patients at high risk of preventable death.
The 12 hospitals, which were not named in the report, were selected from an earlier study by Dubois and his co-author, Dr. Robert H. Brook of UCLA and RAND. Six of the hospitals had death rates significantly higher than would be expected; six had death rates significantly lower than expected.
The researchers found a small number of causes accounted for most of the preventable deaths.
- According to the study, the 23 patients who died unnecessarily from heart attacks were victims of “significant errors in management.” For example, some patients were treated inadequately for chest pain, permitting it to progress to heart attack and death.
- The panel examining the cases diagnosed as strokes found that many of those patients died because they were misdiagnosed or not fully diagnosed. The panel concluded that some may have died of other conditions that went unidentified by hospital staff because insufficient or inappropriate tests were done.
- Patients who died of pneumonia suffered from mistakes in management as well as diagnosis, the study found. Some had received inadequate initial workup while others were given the wrong, or insufficiently powerful, antibiotics.
While Dubois and Brook concluded that it is possible for a hospital to determine after the fact which types of deaths are most likely to have been preventable, they found no way of identifying upon admission those patients at high risk of preventable death.
“I think the hospital medical system is very complex,” Brook, a professor of medicine and public health at UCLA, said in an interview. “We haven’t devoted sufficient resources to the kind of quality-control activities that maybe the best industries have identified.
“My hope would be that the constructive outcome (of this study) would be to try to seriously . . . develop better quality-control systems in hospitals.”
Dubois is a consultant at RAND and vice president of Value Health Sciences, a medical consulting group.
