VisualDecisionLinc: Real-Time Decision Support for Behavioral Health

The societal burden of psychiatric disorders is large. Improving the initial selection of treatments for these disorders has the potential to reduce the time to remission, leading to a reduction in the likelihood of medication errors and adverse events when medications are changed due to initial treatment failure. Guidelines usually lack treatment algorithms that are tailored to a patient’s symptom profile and disease history. Supplementing clinical guidelines with data on treatment response from patients sharing similar profiles would likely narrow the range of treatment options to those with the best available evidence. Incorporating evidence-based recommendations into decision support tools has enormous potential to improve psychiatric care, including initial treatment strategies.

This project designed and developed approaches to identify, aggregate, and present treatment-response information on individual patients and comparative populations as “data views.” These data views were then aggregated to build a visual analytics–based clinical decision support prototype called VisualDecisionLinc (VDL). VDL was designed to improve clinical decisionmaking through the use of integrated data and knowledge derived from electronic medical records (EMRs).

The specific aims of the project were to:

• Develop and validate expert-driven, guideline-driven, and data-driven attribute sets for the creation of comparative populations. 
• Develop a data visualization-based user interface to aid in the selection of treatment choices. 
• Conduct an exploratory effectiveness evaluation of VisualDecisionLinc in preparation for a larger scale, health information technology implementation research methods. 

The project began by aggregating de-identified patient data from MindLinc, the largest available warehouse of psychiatry data, for the design and development of the VDL. The project focused on patients with a primary diagnosis of major depressive disorder (MDD).

Usability testing was conducted with three participants who reviewed a video highlighting the VDL user interface (UI) features. Participants used VDL with simulated patient data and provided verbal feedback that was captured by a screen-capture tool. Information from this initial evaluation was used to make changes to the VDL UI. A second evaluation with six participants focused on how well the VDL UI aligned with the clinician’s workflow. Pre- and post-test results were analyzed. Overall, the results were encouraging. The participants liked the different data views and the ability to customize the evidence to meet their needs.