Laboratories should participate in interlaboratory proficiency testing at a sufficient frequency to enable them to demonstrate confidence in their results. Although participation in proficiency testing is only one of a number of options in ISO 17025 clause 5.9, which deals with the actions laboratory’s need to take to ensure confidence in results, in practice accreditation bodies will regard it as a ‘compulsory option’. Indeed some accreditation bodies regard proficiency testing as vital and will refuse to accredit tests for which no proficiency testing is available. In any event an argument that proficiency testing is difficult to access or expensive will not be accepted by any accreditation body. Where schemes exist which are relevant the laboratory seeking accreditation will be required to participate. Where there is no organised proficiency scheme then some direct interlaboratory comparisons organised by the laboratory will be expected where at all possible.

The reason that accreditation bodies accord PT such preeminent importance is that it is the most searching test of a laboratory’s capability. It is as near as possible to a simulation of the real situation in which the laboratory normally functions.

In selecting a proficiency testing scheme the laboratory should take into account the following factors.

Frequency of Participation

It is difficult to give definitive guidance on this topic but the following points need to be considered.

Internal Administration and Operation of PT

Sample Handling and Administration

In the ideal world the laboratory should ensure that PT samples are treated in exactly the same way as client samples. They should, for example, be submitted blind to the laboratory staff by using a dummy customer. In reality few laboratories will do this especially if they are accredited. Any PT failure will be scrutinised by assessors and there is great temptation to give the samples special and careful treatment. Natural though this reaction is it does reduce the effectiveness of PT as a test of the laboratory’s capability and should be resisted.

If the nature and quality of sample allows it is an extremely valuable exercise to allow as many staff as possible independently to test the PT samples. This then makes the data useful as a demonstration of the ongoing competence of individual operators. The results can be incorporated in training records for example.

PT is not only a test of the technical competence of the laboratory but also of the overall sample handling and quality system. For example you will not be able to dismiss an error in reporting results or problems with sample identification as irrelevant. Just as a technical failure in PT will need to be dealt with as will a failure in other areas which would have resulted in misreporting to a real client.

Evaluation of PT Performance

The report from the proficiency scheme will basically provide the laboratory with a measure of its performance. Parameters of interest are the laboratory’s result, the assigned correct result, from the schemes’ point of view, and an estimate of the acceptable variation. This information is then used to generate a score, usually a z-score or similar. The laboratory should take the score provided by the scheme as an initial indication of performance but should also evaluate its performance in respect of its own internal criteria of quality.

In addition to evaluation of the individual round the laboratory should also examine any trends in the data from proficiency schemes to date.

There must be a documented procedure for evaluating and responding to PT results. The general approach should be as follows.

• A brief report should be prepared either noting that the results are satisfactory or identifying any failures or unsatisfactory trends.

• Any failures should be treated as non-conforming work, investigated and corrective action planned and recorded.

• The results of the investigation and the corrective action taken need to be recorded. There should also be evidence, in the form of satisfactory results from quality control checks, that the corrective action has been successful.

Accredited laboratories will have their PT results and reports examined by assessors at surveillance visits. What should be presented is the report form the PT scheme, details of the evaluation by the laboratory and reports on successful corrective actions. The accreditation body will not be concerned by the occasional problem with PT provided the laboratory is shown to be taking effective and prompt corrective action. Persistent problems with a particular test which are not successfully addressed will, however, result in eventual suspension of accreditation for that test.

The worst scenario is where a laboratory fails take seriously any problems revealed by PT. In this case accreditation will be immediately jeopardised.

Some Proficiency Testing Schemes in Textiles and Leather Testing