Determination of hemolysis index thresholds for biochemical tests on Siemens Advia 2400 chemistry analyzer.

ABSTRACT:

Background

In vitro hemolysis is still the most common source of pre-analytical nonconformities. This study aimed to investigate the hemolytic effects on commonly used biochemical tests as well as to determine the hemolysis index (HI) thresholds on Siemens Advia 2400 chemistry analyzer.

Methods

Peripheral blood samples were collected from forty healthy volunteers. Hemolysis was achieved using syringes. Five hemolysis levels were produced including the no hemolysis group, slight hemolysis group, mild hemolysis group, moderate hemolysis group, and heavy hemolysis group. We then used the bias from baseline (no hemolysis) and HI to construct regression functions. The HI corresponding to the bias limits was considered as HI thresholds. We chose the total allowable error (TAE) as the bias limit.

Results

Of the twenty-eight analytes, ten analytes had clinical significance. Creatine kinase-MB, creatine kinase, potassium, aspartate aminotransferase, and hydroxybutyrate dehydrogenase were all positively affected; the corresponding HI threshold was 45.2, 99.96, 4.07, 10.16, and 7.94, respectively. Lactate dehydrogenase was also positively interfered, but we failed to calculate the HI threshold. Total bile acid, uric acid, and sodium were all negatively affected, and the HI threshold was 42.23, 500 and 501.8, respectively. Glucose was also negatively interfered, but it failed to achieve the HI threshold.

Conclusions

When the HI value was higher than its threshold, the corresponding analyte was considered inappropriate for reporting. The implementation of the assay-specific HI thresholds could provide an accurate method to identify analytes interfered by hemolysis, which would improve clinical interpretations and further boost laboratory quality by reducing errors associated with hemolysis.

SUBMITTER: Du Z 

PROVIDER: S-EPMC6589729 | biostudies-literature | 2019 May

REPOSITORIES: biostudies-literature