What is change analysis?
Change analysis is a comparative root cause analysis technique. You describe the situation where the problem occurred and a comparable situation where it did not, then look at every difference between them. Because a stable process only starts failing when something changes, the differences, and particularly the changes over time, are the most likely causes.
The method was formalized in the MORT (Management Oversight and Risk Tree) program and is used in incident investigation, deviation investigations in GMP environments and troubleshooting in engineering and IT.
The six steps of change analysis
- Describe the situation with the problem.
- Describe a comparable situation without the problem.
- Compare the two situations factor by factor.
- Write down every difference.
- Analyze the effect of each difference on the problem.
- Integrate the relevant differences into the root cause analysis and verify them.
Choosing a good comparison
The comparison makes or breaks the analysis. The closer it is to the problem situation, the fewer irrelevant differences you will have to sift through. Good choices include:
- The same process before the problem started (last week, last batch, before the shutdown).
- A parallel process without the problem (another line, shift, site or product variant).
- The ideal or validated state described in procedures and validation documents.
Change analysis example
The example investigates a bioburden result above the action limit at a purified water point-of-use. Compared with the ten previous compliant samples, several differences stand out: water was stagnant after a three-day shutdown, the weekly sanitization was skipped, a new ball-type sampling valve with a dead leg had just been installed and the valve was flushed for a shorter time by a newly qualified sampler. Three differences are marked as likely causes and four as possible. Location, loop temperature and supervision are ruled out.
Pair change analysis with Is/Is Not analysis to sharpen the problem definition, and with a root cause tree to document how the changes combined.
Change analysis and change control
Change analysis frequently reveals changes that were never formally assessed: a new valve, a modified guard, a supplier switch. That is a signal to strengthen change control. In Cloudtheapp, change requests are risk-assessed, routed for approval and linked to the documents, training and validation they affect, so the next investigation starts with a complete change history.
How to use this Change Analysis tool
- Describe the problem situation. State the event, batch, run or period in which the problem occurred.
- Choose a comparable situation. Pick the closest similar situation without the problem: the previous run, another line or shift, or the same process before the problem started.
- Compare factor by factor. For each factor (what, when, where, how, who), describe both situations side by side.
- Record the differences. Note every difference, even small ones, and describe how it could have contributed to the problem.
- Assess each difference. Mark differences as a likely cause, possible cause or not relevant. Candidate causes are listed automatically.
- Verify and export. Verify the likely causes with data or tests, then export the analysis for your investigation or CAPA.
Frequently asked questions
What is change analysis?
Change analysis is a root cause analysis technique that compares a situation where a problem occurred with a similar situation where it did not. Every difference between the two is examined to see whether it could have caused the problem. It assumes that if something used to work and now does not, something changed.
When should I use change analysis?
Use it when a process, product or system that performed well suddenly starts failing, when the cause is not obvious, or when you have a good comparison: the same process last month, another line, another shift or another site without the problem.
What is the difference between change analysis and Is/Is Not analysis?
Both use contrast. Is/Is Not describes the problem itself along what, where, when and extent dimensions to narrow down causes. Change analysis compares two whole situations, one with and one without the problem, factor by factor. They are often used together.
What factors should I compare?
A common checklist is what (conditions, equipment, materials), when (timing, schedule, status), where (location, environment), how (practices, procedures, steps) and who (people, training, supervision). Add anything else relevant to your process.
Does every difference cause the problem?
No. Many differences are irrelevant. Assess each one: could it plausibly produce the problem? Mark it as a likely cause, possible cause or not relevant, then verify the likely causes with evidence or testing.
Is this change analysis template free?
Yes, it is free with no sign-up. Export your analysis as a PDF, PNG, Excel or CSV file, or save it as a .json file to continue later. Your data stays in your browser.