Best Practices for Pipette Calibration Uncertainty Budgets and CMC Determination

Abstract

With unit sales in excess of one million pieces per year, pipettes are ubiquitous devices and found in a wide variety of laboratories. Applications vary from non-quantitative uses (such as mixing or decanting) to highly quantitative uses such as dispensing of standards and preparation of accurate serial dilutions. Because pipettes are used extensively in critical laboratory tests such as quality control assay of injectable drugs and DNA analysis which results in conviction or exoneration, proper calibration of pipettes is important and the consequences of poor calibrations can be severe. There is a definite upward trend in the number of pipette calibration laboratories and service organizations which have obtained accreditation to ISO/IEC 17025:2005. Each of these laboratories will have performed an uncertainty analysis and calculated their calibration measurement capability (CMC).In this paper, the available scopes of accreditation from forty different pipette calibration laboratories are compiled, compared and contrasted. The various formats were translated into a common basis and then plotted to show how CMCs vary with pipette volume and vary between laboratories. One of the most notable differences is the significant variation in the CMCs among laboratories, which can vary by factors of 100 or more. There appears to be little correlation between a published pipette CMC and other laboratory variables such as experience in the discipline, equipment, reference standards, environmental controls, and capabilities in related disciplines such as mass or volumetrics. To understand the reasons behind these differences in CMC, pipette uncertainty budgets from some leading laboratories were compared. Based on this evaluation, the authors conclude that laboratory practices for establishing pipette calibration CMCs are not well standardized, and the largest source of variation seems to be practice in how the repeatability contribution from the pipette unit under test (UUT) is evaluated and considered. ILAC policy P-14 [1] defines CMC and establishes general policy regarding inclusion of repeatability and reproducibility of the “best existing device”. In practice, there is currently no consensus on which sources of variation in the pipette calibration process should be included in the CMC evaluation, nor is there agreement on how to apply the concept of a best existing pipette. This paper attempts to begin filling this gap by providing recommendations for a best practice in evaluating and communicating the uncertainty of a pipette calibration and for evaluating the CMC of a pipette calibration laboratory. The important questions of reporting measurement uncertainty and the impact it has on evaluating inter-laboratory comparisons and determining compliance with tolerances are also discussed.