A-116 Mixing it up: Salvaging frozen specimens

Abstract Background Proper specimen collection is necessary to minimize pre-analytical errors. Analyte stability after freezing is often provided with FDA approvals but is done on an aliquot instead of in the specimen collection container. The aim of this study is to examine the effect of freezing due to improper storage and/or transport on serum and plasma analytes from centrifuged and gel-separated samples. We investigate whether mixing thawed specimens can salvage the sample, preventing unnecessary patient re-draws and reducing laboratory costs. Methods Thirty fresh plasma and serum specimens that were not required for clinical care were randomly chosen. Baseline values for TSH, vitamin B12, vitamin D3, total PSA and Comprehensive Metabolic Panel (CMP) were collected. Specimens were frozen at -20? C for 24 hours, mimicking the winter conditions in outpatient pickup boxes. Specimens were then thawed at room temperature for 90-120 minutes and re-tested. Then, they were recapped and gently mixed for 10 minutes and re-tested a third time. CLIA, CAP and RCPA total allowable error (TAE) guidelines were utilized to assess whether or not a significant change occurred, in at least one sample, between the baseline and mixed values. Additionally, the data were analyzed using both male and female reference intervals to see if the clinical interpretations changed significantly within our laboratory (10% of samples affected). Results Both serum and plasma specimens were affected significantly when assessed by either TAE and/or reference range standards. For the frozen, non-mixed plasma samples, the majority of the analytes (14/18, 78%)—except TPSA, Vitamin B12, TSH, and Alkaline Phosphatase—were outside TAE standards. However, upon mixing, eight of these analytes corrected themselves, leaving only six analytes (Potassium, Glucose, Total Bilirubin, Bicarbonate, AST and ALT) in the out-of-range category. Three analytes (AST, Bicarbonate, and Potassium, 18%) were altered when using the female reference intervals. Whereas when using the male reference intervals, Bicarbonate, Potassium, and TPSA had alterations in interpretation (17%). Regarding the serum samples, all of the analytes in the frozen, non-mixed specimens (18/18, 100%) were out of range per TAE standards. After mixing, only eight of the analytes (8/18, 44%) had fallen back within range. Regarding a change in clinical interpretation, five analytes (Bicarbonate, Chloride, Potassium, Sodium, and Vitamin D, 29%) were altered when using the female reference intervals, whereas when using the male values, a total of 6 analytes (aforementioned plus Creatinine, 33% of specimens) had changes in interpretation. Conclusion Real-world conditions may create pre-analytical errors that impact analyte results and must be considered. Importantly, this study showed that despite an analyte being stable per the FDA approval on an aliquot, this was not always the case in the specimen collection container. Plasma samples were generally more stable than serum samples, but both had clinically significant changes. These findings have the potential to impact patient care significantly and suggest that manufacturers should be required to provide freeze-thaw stability in approved specimen containers.