Proficiency Testing is an assessment that provides calibration laboratories with an objective means to demonstrate competence to perform specific calibration processes and achieve the proper measurement results within the claimed measurement uncertainty. Proficiency testing is also referred to as ‘Interlaboratory Comparisons’ as the measurement results of an individual laboratory are compared with the results obtained by other participating laboratories.
The primary objective of proficiency testing is to provide the laboratories with the information to identify issues and implement corrective actions to improve the quality of their measurements and to provide greater confidence and reliability for the laboratory’s customers. Participation in proficiency testing can also validate that the laboratory’s calibration procedures, technical competence, traceability, and uncertainty budgets yield results that are within the range of other laboratories.
Proficiency Testing Requirements
In the 2005 revision of the ISO/IEC 17025 standard, proficiency testing was one of the recommended items for ensuring the validity of the laboratory results, but it was not required. The new 2017 revision of the standard now specifically requires laboratories to participate in proficiency testing or other inter-laboratory comparisons.
The requirements for proficiency testing vary between each accrediting body however, most of them use the recommendations published by ILAC by requiring laboratories to participate in a minimum of two proficiency tests per year and cover all major sub-disciplines on the laboratory’s scope every four years. Each laboratory is also required to maintain a Proficiency Testing Plan that documents the schedule of the proficiency tests that will be performed to cover the disciplines on their scope during the four-year period. Documented evidence must also be kept of all proficiency testing results, including the procedure used, data obtained, and submitted uncertainty budgets.
Proficiency Testing Process
A Proficiency Testing provider first sends a test artifact for a specific calibration discipline to an accredited reference laboratory to determine the reference value for the artifact.
The Proficiency Testing provider will then send the artifact to a participating laboratory. The laboratory will measure the artifact according to a given set of instructions and report its measurement results and associated uncertainty to the provider.
The Proficiency Testing provider will compare the results reported by the laboratory to the reference value for the artifact and issue a preliminary report of the comparison to the laboratory. This initial report is sent for the laboratory to review the results and to ensure that the correct data has been recorded. Any discrepancies can be corrected at this time.
If a sufficient number of laboratories have submitted their results to the PT provider, the laboratory can request the final report where the results are compared with the other participants’ results. The laboratory can see their results, but the other participants are only identified by a number, to protect their identity.
The laboratory will then forward the artifact to the next participating laboratory. After each participating laboratory has completed testing, the artifact is returned to the provider.
A technical advisor for the specific proficiency test will monitor the results of each laboratory throughout the course of the proficiency test and make any adjustments to the reference value by using statistical evaluations of the data if needed.
Proficiency Testing Results
Proficiency testing results are commonly evaluated using two performance statistics, En and Z-score.
En, or Normalized Error, is a statistical evaluation used to compare proficiency testing results and the reported measurement uncertainty between the participant and the reference laboratory. It is the primary evaluation used to determine whether a participant’s results are satisfactory or unsatisfactory. If the value of En is between -1 and +1 the results are considered satisfactory. If the value is outside this range, the results are considered unsatisfactory.
Z-score is a statistical measurement of a participant’s testing results as compared to the reference artifacts assigned value and a performance statistic of the results of all participants, such as standard deviation and population mean. The Z-score is used to identify outliers in the measurement results of all participants and that data may be removed from the proficiency testing results. If the value of the Z-score is less than 2, the results are considered satisfactory. If the value is greater than 2, the results are considered questionable or unsatisfactory.
If a laboratory has any unsatisfactory results, the laboratory is required to investigate the discrepancy and document the investigation by implementing the corrective action process. The investigation should determine the cause of the discrepancy and determine if the results may have affected any customer calibrations. If a laboratory fails to do these tests or they continually receive unsatisfactory results, that specific calibration discipline could be removed from their accredited scope or their measurement uncertainty may be adjusted to encompass the results achieved.