

Testing changes is an iterative process: the completion of each Plan-Do-Study-Act (PDSA) cycle leads directly into the start of the next cycle.
A team learns from the test — What worked and what didn't work? What should be kept, changed, or abandoned? — and uses the new knowledge to plan the next test. The team continues linking tests in this way, refining the change until it is ready for broader implementation.
Note: People are far more willing to test a change when they know that changes can and will be modified as needed. Linking small tests of change helps overcome an organization's natural resistance to change and ensure physician buy-in.

 |
 |
 |
Tips for Successful Linked Tests of Change
- Plan multiple cycles for a test of a change.
- Think a couple of cycles ahead.
- Scale down the size of the test (the number of patients or location).
- Test with volunteers.
- Do not try to get consensus, "buy-in," etc.
- Be innovative to make the test feasible.
- Collect useful data during each test.
- Test over a wide range of conditions.Try a test quickly; ask, "What change can we test by next Tuesday?"
|
 | |
 |
 |
 |
 |
 |
Example 1: Improve management of the diabetic population blood sugar levels.
| Cycle 1: |
Develop a system to track hemoglobin A1c levels for the diabetic population. |
| Cycle 2: |
Establish a protocol for hemoglobin A1c routine measurements. |
| Cycle 3: |
Undertake collaborative planning for control levels. |
| Cycle 4: |
Set targets for hemoglobin A1c levels. |
| Cycle 5: |
Implement the protocol with all staff. | Example 2: Decrease length of stay (LOS) for Emergency Department (ED) patients with x-rays.
| Cycle 1a: |
Pilot quick-look for extremity x-rays on one shift. Monitor LOS for patients with x-rays and error rate. Review results with Radiology. |
| Cycle 1b: |
Revise documentation process and try quick-look for two days. |
| Cycle 1c: |
Redesign viewing area and continue quick-look for two weeks. |
| Cycle 1d: |
Make quick-look standard practice and monitor. |
|
 | |
 |
|  |
 |
|  |
|