Background: Although misdiagnosis represents a major patient safety issue, there remains debate as to the exact prevalence of diagnostic error. Several studies have employed retrospective review and physician surveys to identify diagnostic errors, yet these methods may not provide accurate estimates of the prevalence and effect of such errors. Real-time physician-based error reporting systems, however, may allow substantial insight regarding the scope of the problem. Previous reports have shown that a significant number of diagnostic errors can been identified using this method. These systems, however, were not specific to diagnostic error and may have been hampered by poor physician participation. We thus completed a pilot study of a physician-based error reporting system combined with an aggressive educational campaign as a means of determining the prevalence and severity of diagnostic errors on a medical service.
Methods: An anonymous point-of-care reporting system was developed for use by physicians on an adult inpatient medical service. A desktop icon with a link to a SAS-based database was placed on all clinical workstations. Upon clicking the icon, a clinician was prompted to enter four pieces of information: 1) patient identifier, 2) type of error observed (delayed, missed or wrong diagnosis), 3) a brief error description, and 4) the degree of resultant patient harm. Error reports were then reviewed by physician volunteers verifying an error had occurred and recording the causes of the error. An extensive physician-education program was concurrently implemented regarding the importance of diagnostic error and the error reporting system itself.
Results: Over a six month period, a total of 46 reports were entered. Six were deemed not to be diagnostic errors and three represented duplicate entries, yielding a total of 37 unique error reports. The most commonly reported errors included the diagnoses of stroke (3), pneumonia (3), and cervical epidural abscess or hematoma (3). Multiple cases of sub-dural hematoma, acute coronary syndrome, congestive heart failure and pancreatitis were also reported. A majority of the errors were judged by reporting physicians to have resulted in moderate (51%) or severe (22%) harm to patients.
Conclusion: Developed in conjunction with an aggressive educational campaign, a physician-based point-of-care diagnostic error reporting system can be successfully employed to identify diagnostic errors. Many of the reported errors resulted in significant harm to patients. Further detailed analysis of the identified errors may lead to specific interventions to prevent recurrent error.
See more of: The 33rd Annual Meeting of the Society for Medical Decision Making