Abstract Objectives Knowledge of the stability of analytes in clinical specimens is a prerequisite for proper transport and preservation of samples to avoid laboratory errors. The new version of ISO 15189:2022 and the European directive 2017/746 increase the requirements on this topic for manufacturers and laboratories. Within the project to generate a stability database of European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) Working Group Preanalytical Phase (WG-PRE), the need to standardise and improve the quality of published stability studies has been detected, being a manifest deficit the absence of international guidelines for the performance of stability studies on clinical specimens. Methods These recommendations have been developed and summarised by consensus of the WG-PRE and are intended primarily to improve the quality of sample stability claims included in information for users provided by assay supplier companies, according to the requirements of the new European regulations and standards for accreditation. Results This document provides general recommendations for the performance of stability studies, oriented to the estimation of instability equations in the usual working conditions, allowing flexible adaptation of the maximum permissible error specifications to obtain stability limits adapted to the intended use. Conclusions We present this recommendation based on the opinions of the EFLM WG-PRE group for the standardisation and improvement of stability studies, with the intention to improve the quality of the studies and the transferability of their results to laboratories.
We suggest here a pragmatic approach for managing results of clinical chemistry testing in hemolyzed samples collected from adults/older children, attempting to balance the need to produce quality laboratory data with clinical urgency of releasing test results. Automatic measurement of the hemolysis index (H-index) in serum or plasma is highly advisable, whilst low-quality assessment of this test remains less good than a visual inspection. Regarding its practical use, when the H-index value does not generate an analytically significant bias, results can be released, whilst when the value is associated with analyte variation in a range between analytically and clinically significant bias (i.e. variation does not exceed the reference change value [RCV]), results of hemolysis-sensitive tests can be released in association with a comment describing the direction in which data are potentially altered, suggesting the need to collect another sample. When the H-index is associated with analyte variation exceeding clinically significant bias (i.e. variation exceeds the RCV), results of hemolysis-sensitive tests should be suppressed and replaced with a comment that biased results cannot be released because the sample is preanalytically compromised and advising the recollection of another sample. If H-index values reach an even higher critical cut-off (i.e. H-index corresponding to a cell-free hemoglobin concentration ≥10 g/L), all laboratory data may be unreliable and should hence be suppressed and replaced with a comment that all data cannot be released because the sample is grossly hemolyzed, also suggesting the recollection of another sample. Due to inaccuracy and imprecision, the use of corrective formulas for adjusting data of hemolysis-sensitive tests is discouraged.
