Download Is the process improved? Correct use of statistical process behavior

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts
no text concepts found
Transcript
Is the process improved? Correct use of statistical process behavior charting for beginners.
On the whole I found the paper informative and readable.
This paper aims to introduce readers (assumed to be mostly family physicians with no background in
process charting for QI), to process behavior charts, how they may be used for quality improvement
(QI), and some issues with misuse of process behavior charts.
The aims and scope of this paper are useful and reasonable, presenting neither too little nor too much
information for its stated purpose.
There is a variety of terms used “process behavior charts”, “quality statistics”, “process behavior
charting techniques”, “process behavior techniques” is potentially confusing on a first read of the
article.
The argument on page 4 lines 6 through 11, very compellingly (with the figure) argue that annual
averages miss important details shown by monthly averages, but this leaves open the question of
exactly what timescale to use. Would we have seen even more looking at weekly or daily data?
On page 5 line 22, after introducing several terms unique to process behavior charting, you use the
expression “process mean (or median)”, it is not immediately clear if “process mean” is behavior
charting equivalent of “median” or if there are two distinct guides to target specification (although it
becomes clear later in the reading by inference.)
Also the term “Plan-Do-Check-Act cycles” is used with the assumption the reader will know the term.
I have no idea if the average family physician knows that terminology, but as I didn't know it, I thought
it was worth mentioning.
I found the descriptions of the key questions by Oakland (page 6 lines 3-7) slightly ironic; on a literal
reading it is possible to “be doing the job correctly” (process has very little variation while
simultaneously “not capable of doing the job correctly” (the mean is not close to the target
specification). I don't think inclusion of the reference undermines the paper, but the irony did catch my
notice.
The wording of line 9 page 8 and line 11 page 6 lead to some confusion about terminology.
The explanation of the use of run charts and control charts is clear and readable, except on one point
(that may be beyond the scope of an introduction to the subject): the mean, median and standard
deviation don't seem to be updated as each new observation is taken, so how do we choose what initial
population is used to determine these statistics? Or if they (mean, median, and standard deviation) are
chosen rather than observed, how is that process done?
The figures were informative, but I would have liked one more table; table two addresses number of
runs in run charts, but there is no table addressing size of trends for control charts.
On the whole I found the paper informative and readable.
John Wauters
[email protected]