| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
1 Department of Clinical Chemistry, Stichting KCL, Borniastraat 34, 8934 AD Leeuwarden, The Netherlands
aAuthor for correspondence. Fax 31-58-2882227; e-mail a.wolthuis{at}kcl.znb.nl.
To the Editor:
We would like to report an anecdotal quality assurance incident that looked like a complex problem at first but turned out to be the result of a trivial but significant event.
Our national immunochemistry external quality assessment scheme (EQUAS) initiates six distribution rounds per year, with two freeze-dried samples provided in each round. Although we have previously found good agreement between our results and the national consensus values, during this past year we were faced with a 200–300% increase in testosterone results relative to the national consensus (and our method group, Roche Modular E170) in rounds 4 and 5. This upward shift was not observed with our internal quality control (QC) program. The monthly averaged patient results, however, also appeared to be increasing during the last two EQUAS rounds (Fig. 1
). We ruled out the possibility that the averaged patient results were biased by a disproportionate male/female ratio (Fig. 1
). Because all other hormones gave results that were in line with the expected values, the problem appeared to be restricted to testosterone.
|
Reanalysis of stored EQUAS samples gave completely reproducible results. The first three rounds gave results comparable to the consensus, but the samples from rounds 4 and 5 deviated by 200–300%. The problem was therefore traced to these particular samples. A colleague from another laboratory kindly provided us with the round 3, 4, and 5 samples that had been reconstituted in his laboratory. Analysis of these samples produced results that were all within 10% of the consensus value.
Further inquiry into the process of reconstitution of the freeze-dried material did not reveal any extraordinary events that may have contributed to the strange results. However, the laboratory technician responsible for these immunochemistry QC samples reported that he used testosterone supplements for more than a year. Although no problems had arisen during the last year, he realized that during the summer he switched from Sustanon® intramuscular injections to Androgel® gel for skin application. Further temporal analysis revealed a link between the time he switched to the gel and the problems with the testosterone assay. We have to assume that there was some cross-contamination between the Androgel (170 000 nmol/sachet) and the EQUAS sample. Although we know that this particular technician is very hygienic, it appears that it is virtually impossible to completely remove the gel from the hands. This assumption was supported by the observations of Montgomery (1), who showed that five hand washes could not prevent progesterone cream from contaminating a sample through a pipet tip that was picked up bare-handed. In our situation, we believe that during decapping of the EQUAS sample, the gel was transferred to the part of the cap that comes into contact with the sample.
This case clearly shows how an apparently major problem can be traced back to a unforeseen incident and that reconstitution of QC samples is a critical step requiring adequate precautionary measures. In addition, although the incident involved a QC sample, such an event could also affect patient specimens, introducing major interpretive problems. The consistent use of disposable gloves is a prerequisite but not a guarantee because they can be contaminated when, for example, the glove finger is pulled from the box, as was observed by Montgomery (1). Avoiding contact between the hands and critical parts of the gloves should be common practice.
Acknowledgments
We acknowledge Dr. JGJ Pouwels for kindly providing us with EQUAS samples reconstituted in his laboratory (Scheper Ziekenhuis, Emmen, The Netherlands).
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
90/month) are also shown relative to the result for the January samples (solid line and horizontal bars). The percentages of male specimens in the monthly patient populations are indicated by columns. 
and 
represent the results of the testosterone analysis in the immunochemistry EQUAS samples reconstituted in our laboratory, relative to the national consensus (100%). • and 
represent results obtained in the same EQUAS samples but reconstituted in a different laboratory (also depicted relative to the national consensus).