Testing the stability of internal quality controls stored in the automated Atellica Solution refrigerated storage module

Abstract Objectives Atellica Solution (AS) is a platform that incorporates immunoassay and chemistry modules. AS is fitted with a refrigerated storage module (RSM) for internal quality controls (QC). The objective of this study was to assess the maximum permissible storage time in AS for QCs. Methods A total of 48 analytes were tested using QC materials: Liquid Assayed Multiqual (MQ), Liquichek Immunology (LI), Liquichek Lipids (LL), and Liquichek Urine Chemistry (UC). The percentage of variation between results (Xt%) was calculated as the difference between the mean value of the triplicate performed at every time point of the study (Xt) and the average of the triplicate performed in the baseline time (Xo). Stability was assessed based on the total change limit (TCL), which combines analytical and biological variation: TCL=±√((1.65 * CVa)2 + (0.5 * CVb)2). Results A total of 40 of the 48 analytes tested remained stable at the end of the study. In relation to MQ and UC QCs, 32 of the 39 analytes remained stable for the whole study period (15 days) except for alkaline phosphatase, aspartate aminotransferase, calcium, lactate dehydrogenase, and total bilirubin in MQ, and chlorine and glucose in UC. In LI and LL QCs, eight of the nine analytes were stable throughout the 20 days of the study, except transferrin in LI. Conclusions The Atellica Solution refrigerated storage module is a reliable system for the storage of quality control materials.


Introduction
Clinical laboratories have to process a high volume of samples in an optimal time to help clinicians provide early diagnosis and monitor the course of patient disease.In addition, the quality of results should be guaranteed by the use of internal and external controls.
Atellica Solution is a platform developed by Siemens Healthineers (Tarrytown, NY) that combines immunoassay and chemistry modules.One of the advantages of this analyzer is its capacity to store inside internal quality controls (QCs) and calibrators through the refrigerated storage module (RSM) located in the sample handler module.This enables automation of QC processing.The RSM maintains temperature at 2-8 °C and has a maximum storage capacity of 60 QCs and/or calibrators.The containers used for them are 5 mL tubes that are loaded to the analyzer uncapped.The RSM has magnetic caps that cover the position where every tube is located to avoid evaporation.QCs can be stored for some days and storage time is set by the user.
The analyzer must guarantee the stability of QC materials while they are used.Loss of stability may lead to QC repetition and verification, which increases the workload, delays sample processing, and as a consequence there is a higher consumption of controls and reagents.QC supplier should establish recommendations about the maximum permissible storage time in the refrigeration module.However, it is always recommended that each laboratory performs its own stability test.ISO 15189:2012 recommends that internal quality control materials are assayed regularly, according to system stability [1].The objective of this study was to assess, for each analyte, the maximum permissible storage time of the QC materials stored inside the AS RSM.

Sample selection and processing
The study was conducted at the Core Laboratory in the Hospital Clínic of Barcelona in February 2019.During the study, the same AS analyzer and reagent lots were used.The number of calibrations was minimized to prioritize lot calibration over cartridge calibration, to the extent possible.
Four QC materials of each two levels were used: Liquid Assayed Multiqual (MQ) levels 2 and 3 of lot 45800; Liquichek Immunology (LI) levels 1 and 2 of lot 66370T; Liquichek Lipid (LL) levels 1 and 2 of lots 57560 and 57550, respectively; and Liquichek Urine Chemistry (UC) levels 1 and 2 of lot 66790T.All were manufactured by Bio-Rad Laboratories (Irvine, CA).During the study, the same QC lots were used.MQ and LL QC vials were transferred to 5 mL tubes, loading 3 mL of QC into each tube.In LI and UC QC, QC materials were ready for use in 3-mL and 4-mL tubes, respectively.Five tubes of each MQ QC level and three tubes of each UC, LI, and LL QC level were stored in the RSM.The number of tubes was established based on the panel of assays requested foreseeing a higher consumption in MQ QC.MQ and UC QC materials were stored in the RSM for 15 days, whereas LI and LL QC materials were stored for 20 days.Differences in storage time were due to the control material consumption as broader panels of analytes were assayed in MQ and UC.As a result, these control materials were consumed earlier than LI and LL.Each QC material was tested in triplicate once daily to avoid potential random pipetting errors and ensure the reliability of results.Before each run, QC materials were tested to ensure the comparability of results.The analytes that did not meet the quality specifications established by our laboratory (mean ± 2 * SD) were calibrated and controlled again.The zero time point was defined as the time at which QC materials were first loaded.The value obtained at this time point was established as the reference value.
To determine whether AS methods had an adequate analytical imprecision, CVa was calculated for each analyte using the QC materials tested to ensure the comparability of results over the study.Following SEQC ML Analytical Quality Assurance Workgroup recommendations, we used quality specifications based on biological variation [2].
Following each analytical run, QC aliquots were automatically capped and stored in the AS RSM.

Data analysis
The percentage of variation between results (Xt%) was calculated as the difference between the mean of the triplicate performed at every time point of the study (Xt) and the mean of the triplicate performed in the baseline time (Xo).Difference was expressed as a percentage difference: Xt%=(Xt/Xo) * 100 [3].Stability was assessed based on the Total Change Limit (TCL), which combines analytical and biological variation: TCL=±√((1.65 * CVa) 2 + (0.5 * CVb) 2 ).The TCL criterion has been used in previous stability studies conducted by our working group [3,4].CVa is the coefficient of analytical variation for each analyte and it was calculated from internal QC data obtained in the previous 5 months from the date of the study.CVb is defined as within-subject biological variation for each analyte obtained from the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) database [5].In the case of analytes not listed by the EFLM, CVbs were obtained from desirable specifications from Dr. Carmen Ricos' biological variation database [6] (Table 1).For CL and Gluc in UC control, since no database provides CVb for these magnitudes in urine, only the analytical variation (CVa) criterion was used for TCL calculation.In these cases, the following formula was used: TCL=±√(1.65 * CVa) 2 .
Loss of stability was confirmed when Xt% exceeded the TCL.Maximum permissible storage time was defined for each magnitude as the time point from which Xt%, for one or both control levels, exceeded TCL during more than two consecutive days.From that time point, QC was not stable for that specific analyte.

Results
Analytical imprecision (CVa) was acceptable for the QC materials tested for 15 (in the case of MQ and UC) and 20 days (in the case of LI and LL) and in the previous 5 months.The acceptance criterion relied on compliance with quality specifications based on biological variation (BV).All analytes met optimal quality specifications, except for ALP, Crea, LDH, and preAlb, which met desirable specifications; ALB, APO A1, APO B, C3, cHDL, and Trf, which complied with minimal specifications; and Ca, CL, Mg, and Na, which did not meet specifications based on biological variation but had a CVa below that of their peer group.Peer group CVa was extracted from the 2019 FPCQLC External Quality Program.
Tables 2 and 3 show the Xt% for each analyte by day, along with the TCL.The time point at which Xt% exceeded the TCL, thereby resulting in loss of stability, is also provided.A total of 40 of the 48 analytes tested remained stable at the end of the study.In the case of MQ and UC QC, 32 of the 39 analytes remained stable for the 15 study days.In MQ QC, ALP, AST, Ca, LDH, and TBil lost stability at day 8 (ALP), 10 (AST), and 13 (Ca, LDH, and TBil).Figure 1 depicts the behavior of analytes over time.ALP and Ca show a tendency to increase, whereas AST, LDH, and TBil tend to decrease.In relation to UC QC, CL and Glu lost stability at day 8, without showing a clear tendency (Figure 2).
In LI, LL QCs, eight of the nine analytes were stable throughout the 20 days of study, except for Trf in LI, which lost stability at day 18, with a tendency to decrease (Figure 3).
In the case of ALT assayed in MQ and C3 tested in LI, a single Xt% value exceeded the TCL in only a QC level; therefore, it was not considered significant.

Discussion
This study assessed the stability of a range of biochemical analytes in the QC materials stored in the AS RSM.The 48 analytes remained stable for a minimum of 7 days, based on the stability criterion of TCL.This criterion has been used in similar studies assessing the postanalytical stability of patient's serum samples stored in automated refrigerated modules and connected to an automated system [3,4].In contrast, we tested stability in internal quality control materials stored in the RSM of the analyzer, which influences the analytical phase directly.
To the best of our knowledge, there are only two chemistry analyzers available in the market that incorporate an automated refrigerated module for storage of QC materials that allows automated scheduling.These systems are Siemens AS and Abbott Alinity (Germany).Unlike Alinity, AS has a dedicated RSM for QC and calibrators equipped with magnetic caps.Conversely, in Alinity, QC material and calibrators are stored uncapped in the reagent compartment.
The RSM module saves time to operators and facilitates laboratory work, since QCs can be scheduled to be carried out when the workload is low.Understanding the stability of QC materials stored in the RSM is essential to ensure analytical quality.To date, and to the best of our knowledge, no studies have been published to assess the stability of internal quality control materials stored in the RSM of AS.
Bio-Rad InteliQ control materials were not used in this study because they were not available at the time of the study.However, in our opinion, the results obtained are applicable to current InteliQ controls, since they use the same control materials.Only the container is different, as InteliQ uses 5 mL tube and in the study were used QCs vials TCL, total change limit; "-" result not available, LL level  was not estimated for APO-A due to inadequate concentration and because it is not used as internal control of this analyte.Bold values, TCL values and values out of TCL. that were transferred in 5 mL tubes.The time of stability provided in the instructions for use (IFU) for the QC materials used in this study is consistent with the time indicated in the IFU for InteliQ QC materials.According to Bio-Rad, in MQ QC, all analytes remain stable for 14 days, except for ALP, AST, and TBil, which are stable for 9 days, and DBIL, cHDL, CK, IP, and Trig, which are stable for 7 days.In contrast, the results of this study demonstrate that all analytes remained stable for a minimum of 15 days, except for ALP, which lost stability since day 8, AST since day 10, and Ca, LDH, and TBil since day 13.With respect to UC QC, Bio-Rad IFUs indicate that all analytes are stable for 30 days.In our study, however, UC QCs were assayed for 15 days, and we observed that CL and Glu lost stability since day 8, based on the TLC criterion.Finally, in LI and LL QCs, Bio-Rad IFU indicates that all analytes of the study are stable for 30 and 14 days, respectively.However, the results of this study demonstrate that all analytes remained stable for a minimum of 20 days, except for LI Trf, which lost stability since day 18.These results highlight the necessity to test QC stability based on the criteria established by each laboratory.
In conclusion, the AS refrigerated storage module preserved the QC materials studied appropriately.Awareness of the time of stability of each analyte makes it possible to set the maximum time of use of a QC by analyte  and not by control material.Finally, the results of this study enable adequate QC testing by laboratory professionals, thereby reducing the number of repetitions and changes of control material in the presence of biased results.

Limitations
This study has an analytical limitation due to a possible bias caused by the analysis of control materials in different analytical runs.According to the paper published by Fernández-Calle et al. [7], aliquots should have been used and frozen for analysis in the same run.However, these recommendations were published in 2019, after this study was carried out.
al.: Stability of internal quality controls in Atellica Solution Table : (continued) TLC, total change limit; "-" result not available due to an error of the analyzer.Bold values, TCL values and values out of TCL.Rayo Hidalga et al.: Stability of internal quality controls in Atellica Solution Table : LL and LI controls, percentage change for every analyte throughout the study and TCL value.

Table  :
Analytical and biological variation coefficient of the analytes included in the study.

Table  :
MQ and UC controls, percentage change for every analyte throughout the study and TCL value.