PT Frequency: Is Once a Year Better for the Labs?

by Tom Coyner, Business Development Manager

The TNI Laboratory Accreditation Systems Committee has prepared a white paper proposing that the TNI PT frequency be dropped to one per year from the current twice per year requirement.  While this appears to be a simple cost savings proposal, which on the surface would save the laboratories half the cost of PT participation, adoption and implementation of the proposal has greater impact.

It is certainly true that cutting the frequency of PT participation to once per year will decrease the laboratories cost for "routine" participation.  Similarly, such a change would also assist the states that that do not operate an effective PT database and struggle to process PT data on the six month interval.

However, as we learned very early in NELAC the issue with requirements is often not with the cost but with the effect.

The very first NELAC PT Committee struggled with the frequency question.  This Committee felt very strongly that both the States and the laboratories placed too much importance on the PT program.  States used, and still do use, the PT program as a way to “remove” accreditation efficiently.  As a result of the States emphasis on PT samples, laboratories tended to “over work” the analysis of PT samples.  The original PT Committee actually suggested increasing the frequency to quarterly to lessen the impact of any set of PT samples on the laboratories and the States.  Needless to say, the recommendation didn’t fly and the NELAC Board of Directors compromised to twice per year.

It is important to understand the logic of that decision.  The most important question raised during that original discussion was: “How long should a laboratory be allowed to generate bad data before action was taken by the State?”  With a frequency of twice per year and a two-out-of-three passing rule, a laboratory could generate bad data for a year before action could be taken by a State.  The twice per year PT program is a result of this decision and had nothing to do with PT sample cost.

This same question of how long is bad data going to be acceptable is the key to the new Laboratory Accreditation Systems proposal.  If the proposal is accepted and the passing criteria stay the same the answer is that laboratories could generate bad data for two years.  That seems unreasonable and certainly could place significant environmental programs at risk.  If the proposal was accepted and the passing criteria was changed to 100% passing each year, then the laboratories would be running continuous quick turnaround PT samples to make up for failures and their PT cost, confusion, and frustration would increase not decrease.  The PT frequency and the acceptance criteria must be considered together but they work in opposite directions.  Adjusting PT frequency is truly going to be a complex risk-benefit analysis that can have an important impact on the quality of environmental data.

For more information please contact Tom Coyner at t.coyner@apgqa.com.