Researchers have found that as many as 28% of adult patients in intensive care units die each year with a misdiagnosis,
and up to 8% die with a potentially fatal "major missed diagnosis," such as pulmonary embolism or myocardial infarction.
These findings come from a meta-analysis of 31 autopsy-based studies (BMJ Qual Saf 2012; doi:10.1136/bmjqs-2012-000803), and might even understate the rate of missed diagnoses, said Bradford Winters, MD, PhD, associate professor of anesthesiology and critical care medicine at Johns Hopkins University School of Medicine, in Baltimore, who led the research.
"Since we did not include non-autopsy-based studies in our analysis, we did not evaluate misdiagnoses that did not result in death, but that are likely associated with increased morbidity health care costs," Dr. Winters told General Surgery News.
The 31 studies--which were observational, mostly retrospective studies and largely based in the United States--included information from 5,863 autopsied adults who had died in an intensive care unit (ICU). The papers were published between 1966 and 2011. The analysis excluded publications that examined the rate of disease-specific misdiagnoses and studies
that did not include original data. A median of 43% of ICU deaths that occurred during the study period were autopsied.
The investigators turned to the Goldman Classification, widely used for autopsy findings, to group the misdiagnoses they identified. The criteria define class I errors as "missed major misdiagnoses with potential adverse impact on survival and that would have changed management"; class II errors as missed major diagnoses that would not have affected survival or altered the course of care; and class III and class IV errors as misdiagnoses related to the terminal disease but not related to death or unrelated to both disease and death, respectively.
The rate of misdiagnoses detected during autopsy ranged from 5.5% to 100%, with a 28% overall rate (1,632 of 5,863), the researchers found. Class I and class II errors accounted for 8% and 15% of misdiagnoses, respectively; class III and class IV errors accounted for 15% and 21%, respectively. Dr. Winters noted that some studies reported only the total number
of misdiagnoses and class I or class II errors, leaving the specific misdiagnoses of the remaining 41% unclear.
The most common class I and class II misdiagnoses reported in the studies were vascular events and infections.
The 8% rate of major and potentially lethal ICU misdiagnoses is higher than the 5% rate of lethal misdiagnoses documented in the general hospital population in a previous study (JAMA 2003;289:2849-2856). The difference, Dr. Winters explained, can be attributed to ICU-specific factors such as the inability of patients to communicate their medical history during the workup process and limited staff resources leading to "competition for care." Factors not specific to the ICU, including an overload of information and cognitive errors that lead to a biased interpretation of patient data, also may play a role.
Richard Dutton, MD, executive director of the Anesthesia Quality Institute, in Park Ridge, Ill., who specializes in trauma, said several limitations may undermine the generalizability of the findings. "Most autopsied patients have some level of diagnostic uncertainty to begin with, which makes the population in this meta-analysis not completely representative of the general ICU population," said Dr. Dutton, who was not involved in the research.
Some of the studies included in the meta-analysis were conducted before the introduction of more accurate and advanced imaging-based diagnostics, Dr. Dutton noted. And he questioned the effect that missed class I or class II diagnoses would have had on patient outcomes had they been identified. "If a patient is dying of septic shock, secondary events like myocardial infarction and pulmonary embolism, which are common during the immediate premortem period, may not have affected their survival."