With the introduction of ISO 17025 in 1999 all accredited laboratories have
started to implement a procedure for estimating measurement uncertainties for
the analytical methods that are used by those laboratories. These procedures and
uncertainties are integral part of the quality system of the accredited
laboratory and thus will have been reviewed during the audits of the
accreditation bodies.
In January/February 2004 the status of the measurement uncertainties that are
available at the accredited laboratories was evaluated. To investigate the
actual status a large number of laboratories (> 380 in 60 countries) was
contacted via email and/or telephone in January 2004. They were asked for their
accreditation status and they were requested to provide the measurement
uncertainty of some accredited tests as estimated by their internal procedure.
Even for very simple tests like pH or density measurement very large differences
were observed between the various laboratories. Obviously, the current lack of
standardization has caused the estimated measurement uncertainties to vary over
a large range even for the most simple determinations.
It was also found that most laboratory customers do not make use nor want to
make use of the available measurement uncertainties.
Many laboratories have made costs to fulfil the ISO 17025 requirements on
measurement uncertainty to give the users of the test results a doubtful sense
of reliability, while most laboratory customers do not make use nor want to make
use of the available measurement uncertainties. And the availability of
measurement uncertainties does not improve the quality (reliability of the test
results) of the accredited laboratories.
This together with the different approaches of the accreditation bodies and the
different reporting rules and practices causes great confusion in international
trade.
The full article was recently published in the October issue of the journal
Accreditation and quality Assurance, (2004) 9:717-723 and is also electronically
available via the Internet: DOI 10.1007/s00769-004-0888-9. Alternatively you may
send a request for a copy to the author of the article ( rob.visser@sgs.com
)