Microbiology, Error and Risk

About 12-14 years ago I was involved in a university affiliated hospital laboratory that got dragged into the consolidation fever of the time.  The plan was to move most of the laboratory down the road with a few exceptions to provide a rapid service for patients in the Emergency Department.  It was proposed (NOT by me) that one of the tests that should remain would be gram stain analysis of cerebral spinal fluid (CSF).  Fortunately we were able to convince some still rational minds that this would be a very bad idea; a highly critical test being performed by under-trained, under-experienced cross-discipline technologists would have far too great a risk for poor staining, poor reading and inappropriate interpretation.

Jump forward to today; the subject came up again.  With our proficiency testing program we have a fair number of very small laboratories, many in rural parts of northern Canada in a number of provinces.  Most of the work done in these laboratories is chemistry or haematology using modern analyzers.  Rarely or intermittently they are asked to do a gram stain, sometimes on swabs collected during surgery, sometimes on inflamed knees, and sometimes on CSF to help confirm a possible diagnosis of bacterial meningitis.  These laboratories would do one or two stains every couple months.  Some do as few as one or two stains a year. 

As part of our annual review process we have been contacting these laboratories to ask some specific questions about what we as PT providers can do to enhance our educational, technical, and quality oversight value.  It was interesting how many asked if we could act as an advocate on their behalf to stop doing the stains. 

As a microbiologist and a qualitologist and an advocate for patient safety, this seems to be as poor an idea today as it was 15 years ago.  The opportunity for harm is high.  Gram stain reagents that are under-used or improperly used get contaminate with bacteria or things taking on fungal forms.    Inexperienced eyes can confuse residue with gram positive cocci.  Are those dots probably grunge, or are they maybe Staphylococci ? Inexperienced eyes miss fragile organisms such as Haemophilus influenzae or Neisseria species.

Can I put a risk value number to this?  Not absolutely, but there is lots of evidence in a variety of laboratory skills that if a skill is not kept active errors occur frequently.  (The same is true when workers are too busy or too tired).  But if a critical error occurs only once a year in a laboratory performing 10 gram stains a year that is a 10 per cent error rate.

This seems immensely unfair to the technologist who is put in a position of stress and blame, to the physician who has to decide on the quality of information, and especially to the patient.  Pretty much lose-lose-lose.

As a PT provider there are a number of things that we can do.  One is a variant on traditional PT.  In this situation we send common slides to the laboratory for staining and have them return them to us for evaluation of the quality of the stain.  Another is to provide knowlingly negative slides.  Any positive report is a critical error.  The other is to ensure that we provide the maximum number of quality samples that we can so that their experience level can be grown and  maintained. 

While we may not have a legal obligation to take the extra step to monitor and improve performance, we at least have an ethical obligation.  After all, none of the other quality partners are in quite the same situation to provide the additional monitoring. 

More on this subject later.


PS: Registration for the Certificate Course in Laboratory Quality Management has begun with a lot of interest.  A number of changes this year addressing new ISO standards and a look at CLIA.  Those interested in more information can go to www.POLQM.ca or contact ubcpolqm@gmail.com