Knowledge Discovery: Managing Beyond Simple Analytic Processes In Healthcare
Trends in healthcare information suggest that there is an explosion in volume and complexity on the horizon. The migration from ICD-9-CM diagnostic coding to ICD-10 is one well-known contributor to this explosion of data. Other factors include the advance of research in which new diseases, new treatments and new data relationships also add to volume and complexity. Finally, the widespread adoption of electronic health records and health information exchange soon will produce vast new sources of transactional data related to clinical encounters. Today it is nearly impossible to make knowledge-guided decisions using simple analytic processes, and the above-noted trends will compound the challenges.
The Knowledge Discovery Viewpoint Paper, authored by Kit Gorton, M.D. Vice President, Medical Management and Franklin Din, M.D., Director, HP Medical Informatics Center of Excellence, addresses how organizations spend valuable resources collecting and managing terabytes of data, yet struggle to turn the data into the knowledge required to make decisions about benefit design and program investments.
Drs. Gorton and Din emphasize that clinical information management processes need to be viewed as part of a global medical management approach that bridges the information technology, medical policy, clinical operations and quality management functions of most healthcare payer organizations. This multidisciplinary team approach ensures the knowledge discovery processes and methodology are valid and robust, the data used as input for the analysis are clean, precise and standardized, and the outputs are accurate and meaningful to the end-user’s needs. The authors list key questions the team must address in the areas of assessment, design, execution, validation and documentation to ensure the end result creates optimal value.
Three broad elements must be addressed in order to implement knowledge discover-based decision-making:
Governance – An organization’s knowledge governance approach must assess and manage ad hoc queries that can drain time, energy and resources. The paper provides an example of an ad hoc query process that ensures end products are clinically valid, scientifically robust and provide a solid foundation for decision-making.
Technical capabilities – Because knowledge discovery cannot be fully automated; human capabilities still are needed to design, executive and present knowledge outputs. The key is to choose the right tools for each job and get them into the hands of the right people.
Management of human resources – The paper outlines two knowledge discovery management models – self-service and business process outsourcing – and how each can address key challenges to knowledge discovery.
Failing to attend to all three of these elements probably will result in waste, duplication and an inability to achieve the primary benefit of knowledge discovery: fully informed, insightful decision-making. The paper outlines how a well-balanced multidisciplinary team and strong, disciplined knowledge discovery processes within these three areas can yield maximum value.
Drs. Gorton and Din also examine some of the analytic processes that transform data into knowledge, leading to informed decision-making. Also included are key factors public and private healthcare payer organization decision-makers should consider when planning their approach to information management and business intelligence.
HP believes every healthcare organization needs to actively engage in transforming data into actionable knowledge. Doing so provides the means to improve access and quality and control costs in an evidence-based fashion. Failure to do so will lead to gaps in understanding and uninformed decision-making.