Stand-Alone Decision Support System Components

The definition of a decision support system (DSS) is hard to pin down since there are so many varying ways people use the term decision support system. According the Wager, Lee and Glaser, a stand-alone decision support system in a health care setting generally has three main, distinct components or modules: data management, model management and the user dialog or user interface.

The data management module can take a variety of forms. A stand-alone system may have a data warehouse element where it extracts, transforms and loads data from a transactional system. Alternatively, It could be utilizing an existing clinical data repository system for its data source, or it could be employing a large institutional data warehouse where datastores from a variety of systems reside. What is needed is data. Data to model, view and analyze, regardless of where it resides; locally or external, in an optimized data warehouse, or in a transactional system.

The model management module is the most important and complex part of a DSS. The model management module provides the necessary transformations, functions, and visualizations to give the data context and meaning. The models can be statistical, mathematical, or based upon expert knowledge. For instance, a chart describing the prescribing habits of physicians at a clinic may have specific functions or mathematical models applied (e.g. normalization); to help streamline the data analysis process and make the data easier to interpret. The model management component is critical to the decision making process. It is the model managment system that ensures the correct models are applied to the appropriate data.

The actual user interface or system dialog is where the end user interacts with the decision support system. The user should have the ability the “ask the system questions” by locating the desired data and applying the appropriate model to the data. They may bring up an interactive chart of the physician’s prescribing habits. They may drill down into a specific physician and note details, anomalies or outliers. The dialog module should be simple, easy to use, and get out of the way of the user. But most importantly the dialog model needs to allow the user to obtain, visualize and drill into the data. This usability focus can be expensive as it can increase the coupling required between the dialog and model management components.

The definition of a DSS can be as broad as the industries that utilize them. In a stand-alone DDS system, within in a health care setting, there are generally three main modules or components consisting of data management, model management and dialog. The data management component deals with the data source, which may be local or external to the system. The model management component is the most complex and important part, as it is the piece that applies math, stats and/or expert knowledge to the data. The dialog module is the user interface where the user interacts with the data. Each of these components are required regardless of domain; whether a lab is looking a quality data, or an executive is looking at staffing statistics.

Wager KA, Lee FW, Glaser JP, Wager KA. Health care information systems: a practical approach for health care management. San Francisco, CA: Jossey-Bass; 2009.