VisualDecisionLinc: Real-Time Decision Support for Behavioral Health 1. The determination of best treatment options for patients at the point of care is a critical component of providing quality care. Suboptimal choices can lead to more frequent clinical visits, higher health care costs, patient downturns that require emergency intervention, poorer quality of life for patients, and adverse reactions to treatment. The ability of the clinician to provide optimal treatment strategies is being strained by reduced time per patient, increased reliance on mid-level practitioners, and information overload related to treatment options and their effectiveness. Clinical guidelines provide some help, but are often out of date and not readily available as part of a clinician's workflow. The use of Health IT applications and databases of aggregated electronic medical records offer promise that decision support systems can augment the judgment of clinicians in determining treatment options at the point of care by aggregating evidence of historical treatment outcomes. The proposed work hypothesizes that selection of best treatment options at the point of care can be improved by providing clinicians with expert and evidence derived knowledge of how similar patients have fared under different treatment approaches. The application will develop a software prototype termed 'VisualDecisionLinc', which will improve decision making through the use of integrated data and knowledge. VisualDecisionLinc will provide a novel user interface that allows clinicians to quickly assess the outcomes of similar patients across treatment options by integrating historical patient databases and published clinical guidelines into health IT systems. The proposed work will leverage the existing MindLinc electronic medical record (EMR) system developed by the Duke University Health System. MindLinc-EMR is a codified electronic medical record used in behavioral health, widely distributed throughout the United States in academic, private, and public setting and contains a data warehouse of over 2,100,000 patient encounters making it the largest de-identified psychiatry data warehouse in the United States. The proposed work will focus on: 1) developing new approaches to selecting comparative patient populations based on expert-driven, guidelines-driven, and data-driven approaches, 2) development of software user interfaces to quickly allow clinicians to determine which treatment approaches have been effective for patients similar to the presenting patient, and 3) provide an initial evaluation of approaches in preparation of a larger scale deployment and test of clinical effectiveness. As a result of this work, novel ways to leverage historical patient databases and apply Health IT will be demonstrated and tested in the critical area of optimizing treatment choices for behavioral health care.
The project will use a Health IT approach to identify novel ways for clinician to determine optimal treatment options in behavioral health care. An ability to determine optimal treatment options is critical to improve patient care, and reduce the costs.
| Mane, Ketan K; Bizon, Chris; Schmitt, Charles et al. (2012) VisualDecisionLinc: a visual analytics approach for comparative effectiveness-based clinical decision support in psychiatry. J Biomed Inform 45:101-6 |
