Evidence-based decision support for neurological diagnosis reduces errors and unnecessary workup

Area of Science:

• Neurology
• Medical Informatics
• Clinical Decision Support

Background:

• Diagnostic errors pose a significant challenge in clinical practice, particularly in complex fields like neurology.
• Existing diagnostic processes may be suboptimal, even with access to narrative resources and web searching.
• Decision support tools have the potential to aid clinicians in complex diagnostic tasks.

Purpose of the Study:

• To evaluate the impact of the SimulConsult diagnostic decision support software on the accuracy of differential diagnosis and workup in neurology.
• To identify factors influencing the effectiveness of diagnostic decision support.
• To assess the impact on the efficiency and relevance of diagnostic workup.

Main Methods:

• Neurologists used real-case vignettes to generate differential diagnoses and workup plans.
• Performance was assessed before and after utilizing the SimulConsult software.
• Time spent entering clinical findings into the software was recorded and analyzed.
• Changes in diagnostic accuracy, workup errors, and workup relevance were quantified.

Main Results:

• Significant error reduction observed: up to 75% for diagnosis and 56% for workup.
• Diagnostic accuracy improved in 96% of cases where the software influenced the diagnosis.
• Increased time (>2 minutes) for data entry correlated with improved performance.
• A 6% decrease in workup items and a 34% increase in workup relevance were noted.

Conclusions:

• Diagnostic decision support software, like SimulConsult, can substantially reduce diagnostic and workup errors in neurology.
• Accurate and timely input of clinical findings is crucial for maximizing the benefits of such software.
• Implementing decision support tools can lead to more accurate diagnoses and more efficient, relevant patient workup, potentially reducing healthcare costs.