Sidney N. Thornton^1,2,3, and Jonathan P. Saltern³ 

Department of Biomedical Informatics, University of Utah¹, Intermountain Healthcare,

Salt Lake City, Utah², and Utah PWSA³ 

Introduction/Background:  The preponderance of individuals affiliated with Utah PWSA having a diagnosis of Prader-Willi syndrome do not receive the basic and available interventions known to improve their condition despite repeated education and training. When questioned, care givers provide variable reasoning for their noncompliance including factors of information overload, contradictory advisories, unsupportive physicians, and unarticulated transition plans.  Physicians attribute their cautionary approach to a lack of authoritative review amid an insurmountable accumulation of data and opinions.  Through exclusionary language and policy, publicly funded agencies display a general lack of understanding regarding Prader-Willi syndrome progression and presentation.  To break through the informational barriers impeding the delivery of basic care to our population, we envisioned and created a knowledge management strategy. 

Methods:  We followed knowledge engineering methods for clinical care process modeling using a problem-based organization. We modeled domains that are comprehensive for the variety of issues facing the care of an individual with Prader-Willi syndrome (medical, family dynamics, community integration, education, etc.).  Within each domain we harmonized the available published resources, and extracted actionable tasks that could be marked with a binary compliance status. We preserved the link between the extracted recommendation and the source reference. 

To facilitate individualization, we modeled categorical dimensions (age, gender, genetic subtype, educational setting, residential setting, etc.) with mutually exclusive stages within each dimension that would reasonably discriminate along the natural progression of issues affecting Prader-Willi syndrome. Each specific task/recommendation within the care domain was mapped according to the applicable stages for each dimension of care.  We took a representative sample of cases known to Utah PWSA to test for accuracy and completeness of the modeled domains and care dimensions, and factored the care giver workflows into our design. 

Results/Discussion:  From our analysis we constructed a problem-oriented database and a task-based computer application with three tiers of hierarchy (Care Area > Issue > Task), two-tiered referencing (Reference > Citation), authority scoring and life cycle tagging.  In essence, the application manages a filtered list of actionable tasks that are linked to reference citations, progress notes, and compliance logs.  We added workflow features for tracking body mass index, tantrum severity, medication history, and multi-specialty clinic physician notes, in order to facilitate both the individual care giver and the system-wide care manager charged with promoting and tracking guideline compliance. 

Conclusion:  Through the implementation of our knowledge management strategy we have been able to isolate areas of uncertainty and focus resources toward resolution.  As we track progress over the coming years, we anticipate quicker adoption of emerging best practices and improved compliance to a dynamic standard of care.

Edited: 02/09/2012