As the Agency for Healthcare Research and Quality states, health IT must help clinicians by adapting to the way they work (their “workflow”). Neglecting workflow may result in higher expenses, inefficiency, and delays in all clinical activity.
Generally, we need to understand the way health care is currently performed in order to determine how to make cost-effective improvements. This holds true whether the intervention is new procedures, equipment, staff, or health IT, but it’s especially important for health IT projects. One key goal of health IT is to improve the efficiency and quality of care, which heavily depends on its workflow.
Much of health care is coordinated human work that is not about to go away. Health IT, like any important resource, will enable some workflows and constrain others. The workflows that a health IT application enhances must have better efficiency and quality, as judged by medical professionals who are responsible for the care it must support. Making workflow an integrated part of the new design, instead of an unpredictable response to it, simply makes sense. We can “build in” better workflows that add value, and at the same time avoid awkward or unwanted aspects of workflow that health IT might otherwise cause accidentally. A similar technique has been remarkably successful for complex, safety-critical systems in other industries, such as commercial aviation.
One of the most important goals of health IT is to make care more efficient and less costly, while preserving or even increasing care quality. Workflow models provide a representation of care that can be analyzed for gains in care efficiency and some important aspects of quality. The gains allow health IT functions that provide the information to be prioritized on the basis of the value they add to workflow. By comparing the gains of the new workflow over the current one, the customer can see how the new system is worth the investment in terms they readily understand.
SHARPC colleagues at the University of Washington created a workflow modeling tool called MATH. MATH is a standards-based visual diagramming tool. MATH allows professionals to model workflow quickly and accurately.
The tasks that make up a workflow in MATH, each record their own information requirements as one of their key properties. These individual task requirements are captured in the context of a more desirable workflow, which is provides a more accurate and detailed technique for information modeling. Each task’s information needs properties are recorded in MATH’s information dictionary, which assembles information usage across the tasks of the entire workflow.
One of the most disruptive problems in software projects is requirements drift. Often customers can’t really say what their technology needs are, and by the time they see how the new system will work it can be too late or expensive to correct. Medical professionals can readily critique their task information needs in the context of MATH workflows to verify requirements. The MATH information dictionary can then be queried for information usage reports. One key report computes the information architecture that will best meet the usage patterns for the entire workflow. The information architecture can be exported as Java classes. This model-based design method allows inexpensive, fast design iterations over workflow and health IT to come up with the best combination of value added within the constraints of technology and cost.
Contact our UW SHARPC colleagues.
