Sunday, February 3, 2013

Quality and the Research Laboratory -one more time with vigour

Last week I had the opportunity to make a presentation to a group of laboratory sciences graduate students.  My topic was on Quality, a topic that I suspect was pretty marginal in their sphere of knowledge or interest. 

I had two agendas; first that they should be aware that Quality can be viewed both only as a subjective characteristic based on market influenced notions of value and craftsmanship, and specialness, and at the same time as an objective measurable based on specifications, and requirements, and commitment.  The second was to introduce to them not only that Quality in the objective sense has a role to play in every research laboratory, but to go further, the absence of Quality awareness makes everything that they think and do, null and void.

I started with the notion that there are tiers of Quality, starting from the base of Quality Control, then Quality Assessment, and finally Quality Management. 

To be fair, I acknowledge that not all laboratories will attain a level of achievement that includes a full Quality Management System.  Many clinical laboratories have at best a perfunctory Quality Manual and a pretty iffy Document Control system or Process Control and unfortunately most clinical laboratory directors do little Management Review.  But I am also aware that these are by-and-large completely absent in research laboratories unless their funding agency demands that they demonstrate they follow Good Manufacturing Practices (GMP).   

Many researchers scoff at the notion that they should participate in any form of Quality Assessment, thinking that exclusively means some form of proficiency testing or inter-laboratory comparison, and they forget about the simple basics of internal audit, and competency assessment, that let them know whether anything is being done the way that they think it is supposed to be done. 

But the sad reality is that even the most basic Quality Control is by-and-large absent.  All too often basic assessment of equipment and reagents and supplies through use of control materials barely occurs, and when it is done, it rarely is used critically.  There is little use of control charts (sometimes referred to as Levey-Jennings) and a near complete absence on its interpretation. 

I was not alone in the room being knowledgeable on the subject.  One of the seminar leaders lobbed be a really good softball question.  Is there not value in reproducibility as a reflection of accuracy?  If a value is tested multiple times and the same value is achieved, doesn't that give credibility to the value?   Tempting, but sadly, not.  What it did was open up the conversation on accuracy versus precision and bias.  It reminds me of an age-old description of surgeons, confident, quick, adept and wrong. 

Research is a critical part of health progress, and to be fair there has been great progress of the last couple hundred years.  But investigation is slow and erratic, and expensive beyond expensive.  For every step forward, there are dozens of steps back.  That is the nature of unraveling new knowledge.  But we make the situation worse and not better when graduate students don’t know of understand about the roles of Quality Control and Competency.  I would argue that for every laboratory that uses a standard piece of analytic equipment, there should be availability of programs that provide materials to ensure that the equipment is being used properly.  You could call that a variation of proficiency testing or competency assessment.  I would also call it common sense.  I would further argue that an aware funding agency would require the regular use of such challenges, perhaps even tying funding continuity to performance.

Our universities and training centres have an obligation to teach and guide graduate students to mold them into being excellent investigators.  When we leave the very basics of Quality out of the experience and equation, we are just perpetuating our folly, not fulfilling our obligations for investigation improvement. 

We have another session with the group next week.   


  1. Great Post. I work in QA in a research lab and have tried to institute competency testing regarding gas chromatographs. If testing was in force issues such as using LP N2 instead of HP N2 would not have happend.



Comments, thoughts...