Sunday, October 11, 2009
We all think of using data to tell us ‘what has happened’but managers in the behavioral health and social service field will increasingly need the information to predict ‘what will happen.’ Why? Because with a call to move payment for clinical services to payment per case or payment based on performance, the ability to predict the resource needs and outcomes of the consumers we serve will be critical to the financial health of our organizations.
My colleague, Joe Naughton-Travers, describes the ‘predictive phase’ of organizational information literacy as the period where an organization can stretch its use of technology and information to include modeling, forecasting, and prediction. Here data is used to predict and plan for possibilities to further enhance operations and care delivery.
Thomas W. Doub, Ph.D., Chief Operating Officer for the Centerstone Research Institute (CRI) emphasizes that using data for predictive modeling should be viewed as a clinical support tool to inform clinical practice—not to replace clinical judgment. He points out that any statistically derived model is based upon groups, and thus cannot be automatically applied to individuals. On the other hand, predictive models can provide clinicians an important resource to help better inform day-to-day clinical decision-making.
By utilizing data analytics and predictive modeling, behavioral health provider organizations can make important strides not only in day-to-day administration and in developing important cost-effective practices and operational efficiencies, but also in the use of evidence-based clinical care. A full spectrum of analytic and predictive modeling tools are available, from basic reporting to data queries and drill-down (identifying specific problems or events), to statistical analysis of why certain outcomes or results have occurred, as well as to trend forecasting and the development of optimization strategies for future success.
Dr. Doub will be a panelist in “Using Informatics to Gather Predictive Information and Select Optimal Clinical Practices,” a session at the upcoming Institute for Behavioral Health Informatics. If you are unable to join us this year, stay tuned on October 14, 15, and 16 for our new, live on-line coverage of the Institute!
John F. Talbot, Ph.D.
Executive Vice President
For more on data-driven modeling, access “Executive Leadership in a Data-Driven Organization: Using Metrics-Based Performance Improvement Models.” PREMIUM