A statistical model for restoration of serum potassium level disturbed by hemolysis

Assessing the influence of true hemolysis occurring in patient samples on emergency clinical biochemistry tests results using the VITROS® 5600 Integrated system

Hemolysis is one of the most frequent causes of pre-analytical errors in the emergency department (ED), and it can lead to inaccurate blood results and often requires repeat testing. The purpose of the present study was to evaluate the effects of true hemolysis occurring in ED blood samples on routine clinical biochemistry tests using the VITROS^® 5600 Integrated system. A total of 92 pairs of blood samples were collected from 92 ED patients. Each pair of samples included one hemolyzed sample and one successful (non-hemolyzed) redraw from the same patient. A total of 21 common laboratory analytes and the hemolytic index (HI) were examined. The degree of hemolysis (slight, mild, moderate and heavy) was determined based on the HI. A clinically significant difference in one analyte was defined as a difference greater than its Clinical Laboratory Improvement Amendments of 1988 (CLIA'88) total allowable error (TAE) limits. The results demonstrated that the mean differences in 11 serum analytes (unconjugated bilirubin, Ca²⁺, equivalent CO₂, Cl^-, creatinine, glucose, Mg²⁺, phosphorus, Na⁺, urea nitrogen and uric acid) in hemolyzed and non-hemolyzed samples were within their CLIA'88 TAE limits, while the differences in the other 10 analytes [alanine aminotransferase (ALT), albumin (ALB), amylase (AMYL), aspartate aminotransferase (AST), total bilirubin (TBIL), creatine kinase (CK), CK-myocardial band isoenzyme (CK-MB), lactate dehydrogenase (LDH), K⁺ and total protein (TP)] in paired samples in at least one of the four groups were greater than their CLIA'88 TAE limits. These results suggest that hemolysis had a notable impact on ALT, ALB, AMYL, AST, TBIL, CK, CK-MB, LDH, K⁺ and TP levels. Furthermore, for ALT, AMYL, TBIL and TP, wet chemistry methods displayed superior anti-hemolytic ability compared with dry chemistry methods. Notably, a high concentration of AST was less affected by hemolysis.

Demonstrating the feasibility of accurately and reliably correcting potassium results for mildly hemolytic samples using a new experimental design

BACKGROUND AND AIMS

For several decades, there has been an ongoing debate about the appropriateness and reliability of correcting potassium concentration results for hemolyzed samples. As part of implementing a new Roche Cobas Pro analyzer system the possibility of correcting potassium results in hemolytic samples using a new, thorough experimental design, was investigated.

MATERIALS AND METHODS

The relationship between hemolytic index (HI) and increases in potassium concentration was studied by performing a linear regression on hemolysate dilution series (HI 0-160 mg/dL) from 20 left-over patient samples. The obtained correction procedure was validated using another 20 left-over patient samples. Corrections were accepted according to a correction concordance of 100% within the total allowable error criterion of 4.85%.

RESULTS

The obtained reporting procedure was: HI 0-17 quantitative potassium reporting, HI 18-100 correct potassium for HI, and report as text including a disclaimer for in vivo hemolysis; samples were rejected for HI > 100. In the validation cohort, 70/70 samples eligible for correction were within the TEa criterion. The maximum negative and positive errors were -2.8% and 2.9%, respectively.

CONCLUSION

Correcting potassium concentration results in a designated HI range is feasible and increases the accuracy the potassium results in samples with mild in vitro hemolysis.