Transcript Training

Even More SQA:
CAPA
Corrective and
Preventive Actions
What is CAPA?

activities not intended to correct
defects, but to eliminate the cause
of those defects.

corrective - fix past problems
preventive - predict potential problems

Reality Check...

How can we see into the future and fix
problems that have not yet occurred?
How to conduct CAPA
1. information collection
2. analysis of information
3. development of improved methods
4. implementation of improved methods
5. follow-up
What info gets analyzed?

Roundup the usual suspects…








design review reports
code walkthrough reports
unit test and integration test reports
project progress reports
customer complaint records
software change requests and maintenance reports
training follow-up reports
But don't forget…


special reports on quality / quality audits
suggestions / observations by staff
Ways to sift through that
huge pile of records

asks the reporter to indicate
a priority for their report

random sampling

combination of the two
techniques above
Statistical SQA
1.
Info about defects are collected and
categorized.
2.
Defects are traced to their root cause.
3.
Isolate the vital 20%
4.
Move to correct problems that caused
defects.
SSQA is similar to
"Six Sigma"
Pareto Principle:
80% of defects can be traced
to 20% of causes.
When do you analyze that
huge pile of records?

Answer 1 : After a plane crash.

Answer 2 : As the plane is flying.
Background:
Common Cause and Special Cause
" 'A riot occurs in a certain prison. Officials and sociologists turn out a
detailed report about the prison, with a full explanation of why and how it
happened here, ignoring the fact that the causes were common to a
majority of prisons, and that the riot could have happened anywhere.'
The quote recognizes that there is a temptation to react to an extreme
outcome and to see it as significant, even where its causes are common
to many situations and the distinctive circumstances surrounding its
occurrence, the results of mere chance. Such behavior has many
implications within management, often leading to interventions in
processes that merely increase the level of variation and frequency of
undesirable outcomes."
Harry Alpert
Control Charts

Useful and Very Common Analysis Tool
–

detect and predict problems in the process
Lines:
–
–
Center Line = mean
Upper and Lower Control Limits (UCL and LCL)

–
Warning Limits


typically 4 standard errors
typically 3 standard errors
If the process is "in control", all points will plot
within the control limits.
adapted from: http://en.wikipedia.org/wiki/Control_chart
Scatter Charts

(example one)
Eruptions generally fall into two categories:
1) short wait, short duration
2) long wait, long duration

Hence, some simple data
analysis yields insights
into this process.
stolen from http://en.wikipedia.org/wiki/Scatter_diagram
Scatter Charts

(example two)
Cyclomatic Complexity v. Module Length
CAPA Outcome :
Improved Methods




updating processes
updating relevant work procedures
changing tools
improvement of reporting methods




report contents
report frequency
changes in training
probably, several of the above
And, of course, the obvious question

Why bother with CAPA?