External quality assessment of serum indices: Spanish SEQC-ML program.

OBJECTIVES: Serum indices included in clinical chemistry instruments are widely used by laboratories to assess the quality of samples. Instruments that report quantitative results allow an evaluation of their diagnostic performance in a similar way to other biochemical tests. The Spanish Society of Laboratory Medicine (SEQC-ML) launched a monthly External Quality program of serum indices in 2018 using three lyophilized materials of simultaneous annual distribution. We present the results of the first three years of the program. METHODS: The use of four different quality control materials with different concentrations in three alternate months allows an annual evaluation of the participant's accuracy. Assigned values are established by consensus among homogeneous groups, considering necessary at least 10 participants for a comparison at instrument level. The average percentage difference results per instrument allow the assessment of bias among groups. RESULTS: The imprecision of the three indices ranges between 3 and 9%, with no major differences among instruments. Significant differences were observed in all indices among instruments with more than 10 participants (Roche Cobas, Abbott Architect, Abbott Alinity and Siemens Advia). The 90th percentile of the distribution of percentage differences was used as the analytical performance specification (APS). An improvement in performance was observed in the first three years of the program, probably due to the learning curve effect. In 2020, APS of 7.8, 12.2 and 9.7% were proposed for hemolytic, icteric and lipemic indices, respectively. CONCLUSIONS: Serum indices have a great impact on the quality and the reliability of laboratory test results. Participation in proficiency testing programs for serum indices is helpful to encourage harmonization among providers and laboratories.

Laboratory sample stability. Is it possible to define a consensus stability function? An example of five blood magnitudes.

BACKGROUND: The stability limit of an analyte in a biological sample can be defined as the time required until a measured property acquires a bias higher than a defined specification. Many studies assessing stability and presenting recommendations of stability limits are available, but differences among them are frequent. The aim of this study was to classify and to grade a set of bibliographic studies on the stability of five common blood measurands and subsequently generate a consensus stability function. METHODS: First, a bibliographic search was made for stability studies for five analytes in blood: alanine aminotransferase (ALT), glucose, phosphorus, potassium and prostate specific antigen (PSA). The quality of every study was evaluated using an in-house grading tool. Second, the different conditions of stability were uniformly defined and the percent deviation (PD%) over time for each analyte and condition were scattered while unifying studies with similar conditions. RESULTS: From the 37 articles considered as valid, up to 130 experiments were evaluated and 629 PD% data were included (106 for ALT, 180 for glucose, 113 for phosphorus, 145 for potassium and 85 for PSA). Consensus stability equations were established for glucose, potassium, phosphorus and PSA, but not for ALT. CONCLUSIONS: Time is the main variable affecting stability in medical laboratory samples. Bibliographic studies differ in recommedations of stability limits mainly because of different specifications for maximum allowable error. Definition of a consensus stability function in specific conditions can help laboratories define stability limits using their own quality specifications.