Measuring Process Adherence
One of my clients has a project to improve the control of diabetes, as measured by patient’s blood test of hemoglobin A1c. The outcome measurement: the fraction of patients in a reporting period whose A1c levels are deemed “out of control”.
The project team has focused on following a process that looks like this flow chart:
A wise and experienced colleague, Beth Waterman, advised the team to measure adherence to their process as a sure-fire way to improve.
One way to interpret Beth’s advice: query the electronic record system to run a report at the end of the week or month that counts the patients who did not make it to the end of the process out of all the patients eligible. Calculate the per cent of patients, week by week or month by month, as a summary of process performance. Plot the per cent dots in a run chart and you have a picture of process performance.
However, another wise and experienced colleague Kelly McCuthcheon-Adams quotes this adage: Weighing a cow doesn’t fatten it. Measuring doesn’t improve anything.
How you generate the per cent dots matters. There’s a different way to generate the per cent dots that directly sets up improvement.
Lean coaches recommend a focus on the process itself as the path to better process outcomes.
There’s another fundamental principle in Lean: make value flow.
If you are making parts in a sequence of operations, the flow principle implies you should not make large batches of parts because batching disrupts flow. Batch production typically leads to all kinds of wasted motion and effort. The ultimate reduction is to a batch size of a one part, which enables “one-piece flow.”
When we marry “focus on the process” to “one-piece flow”, we get a useful implication for process measurement in the clinic:
Track how well the clinic staff can follow their process, patient by patient on a white-board or tally sheet. Note where the process fails, patient by patient; note the circumstances for failure. This patient-by-patient tracking represents the fundamental meaning in Beth's advice to “measure adherence.”
At the end of each week, tally up the patients who make it through all the steps; compare to the number of patients served that week. As before, you can calculate a weekly per cent and track the process performance over time. (If it's easier to get the per cent dot for each week by query of some electronic data file, then you can use the automated report. But only use the automated report once you've convinced yourself that it matches the direct observation/tally method.)
With patient-by-patient tracking on a board or tally sheet, you have much more information than just the per cent dots: Which steps have the most failures? What do the observations tell you about the nature of those failures?
This additional process information provides clues and provokes change ideas that deserve tests. The patient-by-patient approach to process measurement automatically focuses team attention on process adherence and sets up process improvement in ways that an electronic query once a month simply can’t match.