A diagnostic error is any mistake or failure in the diagnostic process leading to a misdiagnosis, a missed diagnosis, or a delayed diagnosis. This is an operational definition that includes any failures in the process of care, including timely access in eliciting or interpreting symptoms, signs, or laboratory results; formulating and weighing differential diagnosis; or lack of timely follow-up and specialty referral and evaluation. A diagnostic error is a construct that is usually based on reference to a subsequent test, clinical outcome, consultant's diagnosis, or autopsy—gold standards that are themselves often imperfect or unavailable. Errors in diagnosis-related processes are ubiquitous, ranging from a trivial failure to ask an “insignificant” historical question to overlooking minor lab abnormalities, to switching specimens between two patients, to more serious errors in interpretation of data, which may or may not have adverse clinical consequences in terms of labeling a patient with an erroneous diagnosis or impacting clinical actions or outcomes. Detecting diagnostic errors is critical to correction of the ongoing care for a current patient, as well as for learning how to avoid similar errors in the future.
Although there is a paucity of data on the prevalence of diagnostic errors in everyday practice, studies using a wide range of approaches suggest that the error rate is not small, conservatively 10–15% for many diagnoses. Selected examples and rates from these studies are summarized in Table 8-1. These studies, however, have serious limitations. To better quantify the frequency and types of diagnosis errors and their clinical outcomes we need research to supplement these indirect and retrospective data with more direct, more encompassing (ie, looking at more than just one diagnosis or patients who die), prospective studies, similar to those that have been done with medication errors. This is necessary not only to determine the magnitude of the problem in various settings but also to gauge the effectiveness of interventions. Unfortunately, we lack reliable, validated, and efficient methods for carrying out such studies. We believe one future role for hospitalists will be to contribute to ongoing surveillance to help characterize the incidence and types of such errors.
Table 8-1 Estimated Incidence of Diagnostic Error ||Download (.pdf)
Table 8-1 Estimated Incidence of Diagnostic Error
|Type of Study||Study Example|
|Autopsy||Major unexpected discrepancies that would have changed the management are found in 10%.|
|Patient surveys||One-third of patients reported experience with a diagnostic error involving themselves, a family member, or close friends.|
|Second reviews—radiology||10–30% of breast cancers are missed on mammography.|
|Second reviews—pathology||6171 pathology specimens at Johns Hopkins were reread: major changes in prognosis or treatment were found in 1.4%.|
|Lab errors||9% overall error rate, including pre- and posttest errors.|
|Standardized patients||Internists misdiagnosed 13% of patients presenting with common conditions in clinic.|
|Error databases||Of voluntary reports by Australian physicians, 34% were diagnostic errors, and these were judged to be the serious and least preventable.|