Construction and Interpretation
Quality Control charts have always been a handy way in which to analyze, synopsize and present process quality control. In the event of a problem with a product, all quality control information will need to be reviewed by any interested regulators. A company able to facilitate such review by immediately pulling out the relevant QC charts is a company that is also sending the message (to both its regulators and its customers) that it has, on its own initiative, reviewed the relevant QC information and found it satisfactory BEFORE the product in question was ever released. A company unable to demonstrate a competent and confident facility with QC Charts is likely to face additional questioning from regulators, who are, after all, just trying to do their jobs.
A QC Chart is not just a place in which to record different kinds of process information relating to sequential lots of product produced by a given plant. To take a simple example, let's assume that the pH of a finished product should be 4.3. It is expected that different pH readings taken on different, freshly formulated lots of a product will not all be 4.3 exactly. Some may be a little above that value. Others may be a little below. As a sequence of pH readings is entered into the QC Chart and the observed variation recorded several times a day for, say, a week a month or a quarter, the observed variation may be calculated statistically. One may ask whether the average (mean) value of the pH readings remains stable or whether it migrates. With sufficient accumulated points, one may evaluate the standard deviation of the observed variance and ask whether the amount of observed variation is stable or variable. It is also advisable to construct a frequency distribution of the pH readings involved and then ask whether the observed distribution is symmetrical. If it is, then ask whether it is a Poisson distribution. If the frequency distribution is not symmetrical, then one must ask whether it is skewed to the left or the right, or whether it is Log Normal. If the frequency distribution of a measurement changes its characteristics with time, then such an observation is significant.
The foregoing is, of course, something that many food manufacturers already implement. However, the foregoing paragraph may strike others as being just a little dense. Such thinking is an indication that your company may need a little help. Food Safety Analysis will take on the task of producing as many QC Charts as your production situation may require. If you wish, certain of your own employee(s) may be trained in this task, while being remotely trained and/or managed by Food Safety Analysis, LLC.. With sufficient training, your designated staff may eventually be able to take on the day-to-day functions of this job in-house, leaving Food Safety Analysis only an oversight role, as needed. The time-honored goal of consultants of all stripes is always to work themselves out of a job.
One should always keep in mind that, as discussed elsewhere on this site, HACCP is currently the paradigm for and definition of process quality and safety, according to regulatory agencies such as FDA. Any HACCP plan makes the implicit assumption that a Critical Control Point (CCP) such as the final product pH is a stable quantity. This is why QC Charts of CCP's are needed. It is in a company's own self interest to demonstrate the stability of its critical control points.
To extend this example just a bit further, any pH reading is a function not only of the sample being measured. It is also a function of the cumulative effect of the various matrices presented to the pH meter for measurement, the quality of the pH meter itself, the nature of the electrodes used, the manner of their maintenance as well as how recently they have been cleaned. It is one thing to stick a pH electrode set into a sample and read the number off the dial of a pH meter. It is quite another to be in competent control even of something as "simple" as a pH measurement. The QC Chart is designed to show exactly the status that any food producer wants their customers (and their regulators) to know, that food production is being done correctly. If QC Charting shows that your staff could use additional training, better for all concerned that they change their ways before a problem arises that requires regulatory scrutiny.
Keep in mind as well that QC Charts don't just show problems. When things are working as they should, they also show that all is well. Indeed, they can show that a potential issue was recognized early and rectified before it ever got the chance to become a problem. In a very real sense, QC Charts are the kind of advertising a food manufacturer cannot buy, no matter what happens on the production floor.
Copyright © 2012 by M. Mychajlonka, Ph. D.