Once a team has a goal, it needs to figure out what measures, or metrics, are needed to track progress. Measurement is a critical part of testing and implementing changes. The right measures tell a team whether the changes it is making is actually leading to improvement.
The use of metrics answers the question, "How will we know that a change is an improvement?"
The more specific a goal is, the easier it will be to test with the Plan, Do, Study, Act cycle.
Here are some tips for measuring data:
- Plot data over time.
- Seek usefulness, not perfection.
- Use sampling.
- Keep it simple.
- Integrate collection, display and analysis into the daily routine.
- Use qualitative and quantitative data.
Three Types of Measures
It's useful to remember that different measures will reflect what is happening at different levels of the organization. "Outcome measures" reflect the experience of the member/patient. "Process measures" speak to how well or poorly a process is working. "Balancing measures" examine multiple parts of the larger system to see how different parts are affecting each other.
Teams can use these questions to help select an appropriate measure for a goal.
Outcome measures (voice of the member or patient)
- How is the system performing? What is the result?
- Tied directly to goal statements.
- May measure time, clinical outcome, financial performance or patient satisfaction.
Process measures (voice of how the process works)
- Are the parts/steps in the system performing as planned?
Balancing measures (viewing system from different directions/dimensions)
- Are changes designed to improve one part of the system causing new problems in other parts of the system?
- What happened to the system as we improved outcome and process measures?
Three Types of Data
Teams should also be aware that the data varies qualitatively. It's important to choose understand this variation—and they don't necessarily need research quality data to demonstrate the effectiveness of a change.
This is data generated for reporting purposes. Its characteristics are that it is:
- Reported to federal and state agencies.
Data for research must meet the "beyond doubt" test. By nature:
- There is lots of it.
- It proves hypotheses
- It is statistical.
This data provides just enough information for team learning. It tends to be:
- Limited in quantity.
- Small samples/tests of change.
- Incorporates changes as needed.
For an easy-to-print version of the information on this page, download the Establishing Measures tool.