A recent Institute of Medicine report that states that medical errors had a major impact on death rate is overstated, according to a new analysis of the data which formed the basis of the report. The IOM report relies heavily on an observational study without a control group to make exaggerated claims about large numbers of preventable deaths according to an article published in the July 5 issue of the Journal of the American Medical Association. Analyzing available comparison data, the authors of the letter found that adverse events in the original study of 31,429 patients had no estimable effect on the death rate. Adverse events are defined in the IOM study as injuries caused by medical management.
The JAMA letter, written by Clement McDonald, M.D., Michael Weiner, M.D., and Siu Hui, Ph.D. of the Regenstrief Institute for Health Care and the Indiana University School of Medicine, says that the IOM report overemphasized and overdramatized data from a 1984 study of New York hospital admissions. They write that this report did not isolate a critical determinant of death the fact that the patients studied were already quite ill.
"Patients admitted to hospitals have high risks before they even enter the hospital. Although some hospital deaths are clearly preventable, most will occur no matter how many 'accidents' we avoid. This base-line death risk has to be known and factored out before we draw any conclusions about the real effect of adverse reactions on death rates preventable or otherwise," say McDonald, Weiner and Hui.
The authors point out that the IOM report's assertion that the 13.6 percent death rate is caused by medical errors is tantamount to saying that the death rate would be zero among equally sick hospital patients who had a similar baseline death risk but avoided an adverse event. "Common experience tells us that this could not be true," they say.
"There certainly could be a relationship between adverse events (preventable or otherwise) and death," the authors say. Noting that they suspect such a relationship exists, they encourage the development of mechanisms for reducing adverse effects such as computer-based medical records systems that include automated reminders and forms of error detection. Such a system --the Regenstrief Medical Records System has been developed during the past 25 years by Dr. McDonald and colleagues and is used to collect and provide patient data to many health care providers in the Indianapolis area.
Dr. McDonald is director of the Regenstrief Institute for Health Care. He is distinguished professor of medicine at Indiana University and holds the Regenstrief chair at the IU School of Medicine. He was elected to the Institute of Medicine in 1994.
In addition to his research scientist position at the Regenstrief Institute, Dr. Weiner is an assistant professor of medicine at the IU School of Medicine and a center scientist at the IU Center for Aging Research.
Dr. Hui is a senior research scientist at the Regenstrief Institute, professor of medicine at the IU School of Medicine and senior biostatistician at the IU Center for Aging Research.
Information on the Regenstrief Medical Records System, the LOINC data base and the Regenstrief Institute for Health Care
The Regenstrief Medical Records System (RMRS) is a physician-designed integrated inpatient and outpatient information system that contains over 25 years of data. Physicians and other health care providers enter information derived from physical examinations, diagnostic images, clinical laboratory tests and other patient treatment data at terminals located at the point of care.
The RMRS is unique because of the standardized manner in which lab results, in addition to medical procedures, prescriptions, discharge summaries and other information, are coded and entered for rapid reference. Unlike paper hospital records that can be misplaced or read by non-authorized individuals, the RMRS always is available and maintains confidentiality. Access to records is password controlled and monitored, and transmitted data is encrypted. Regenstrief-designed clinical workstations are user friendly and user smart. Physicians write all of their orders and, increasingly, their notes at these workstations.
The RMRS has registered over 1.55 million patients since 1972 and contains more than 19 million prescriptions, 200 million numeric or coded patient observations, 3.25 million dictated reports and 350,000 EKG tracings. It is accessed more than 800,000 times a month.
The RMRS is used at more than 40 inpatient and outpatient facilities in Indiana and is the largest coded, continuously operated medical records system in the country.
A commercial version of the system's software is being successfully marketed internationally. Medical facilities, which purchase the software, develop their own patient data bases using the RMRS established parameters.
The Logical Observation Identifier Names and Codes (LOINC) database consists of 26,000 standardized names and codes for medical tests and clinical measurements These standards are being widely adopted by major health delivery systems for use in electronic medical records and outcomes research. It is distributed for public use on the Web at http://www.mcis.duke.edu/standards/guide.htm
The Regenstrief Institute for Health Care is an international leader in the fields of medical informatics and health services research. The mission of the Institute is to conduct research to advance health care by improving the capture, analysis, content and delivery of the information needed by patients, their health care providers and policy makers and to conduct interventional studies designed to measure the effect of the application of this research on the efficiency and quality of health care. Information on the Institute can be found at http://www.regenstrief.org
Journal
JAMA