Virtual Patient Education From Real Cases

Diagnostic errors and resulting misdiagnosis-related harms represent a major public health problem. Most diagnostic errors result from failures in bedside diagnostic reasoning. Gaps in expertise (ultimately linked to faulty knowledge, inadequate training, or lack of feedback) account for many of these failures. Current systems of medical education, residency training, and licensure requirements have proven insufficient to prevent harms from diagnostic errors. The National Academy of Medicine recommends using simulation training with early exposure to a variety of typical/atypical cases to improve diagnostic performance. A systematic review found strong positive associations between simulation training and improved outcomes of knowledge, skills, and behaviors. Training in bedside diagnosis could be dramatically enhanced through symptom-specific virtual patient (VP) curricula that expose learners to real-world cases in a deliberate practice framework - practice that is motivated, purposeful, and systematic. This approach allows the appropriate mix of cases and difficulty to be presented to learners sequentially, potentially enhancing clinical skills.

Clinical presentations with nonspecific symptoms and diagnoses with wide differentials are especially prone to diagnostic errors; dizziness may be the epitome of this conundrum. Dizziness is a symptom that is common, costly, and associated with missed stroke. Isolated dizziness is the most common clinical context for missed stroke. Stroke is a leading cause of major long-term disability in the United States and an enormous source of global disease burden. It is listed as the fourth most common diagnostic errors among those reported by physicians. In fact, closed-claims analyses focused on neurologic conditions found stroke as the most common misdiagnosis, and more than 20% occurred in the ED. A hospital records analyses indicated that deaths due to cerebrovascular events result from diagnostic error far more frequently than those due to myocardial infarction. The ED is a high-risk site for diagnostic errors and indiscriminate use of neuroimaging for diagnosis of dizziness is neither accurate nor cost-effective. Risk stratification using symptoms and signs at the bedside offer the potential to provide cost-effective reductions in misdiagnosis-related harms. There is evidence to suggest that providers are currently ill-equipped and harbor misconceptions about the best approach to dizzy patients. Therefore, the investigators chose dizziness as the "model symptom" for study.

Even at Johns Hopkins Medicine, where dizziness diagnosis has been studied extensively, this remains a problem. The Center for Diagnostic Excellence group recently showed that (a) <5% of dizziness charts have correct documentation of standard bedside examination techniques; (b) 40% of patients leave the ED with a symptom-only diagnosis (at least half of whom could have been correctly diagnosed and treated); and (c) 39% receive a CT (>90% of which are inappropriate). Emergency medicine residents overwhelmingly express the need for better training in dizziness diagnosis.

The investigators recently demonstrated that <10 hours' worth of simulation-based deliberate practice training using a dizziness-focused ViPER (Virtual Patient Education from Real Cases) curriculum made internal medicine interns twice as accurate as senior resident colleagues on VP cases. The investigators now seek to do the same for emergency medicine residents and demonstrate real-world impact.