Analyte stability in whole blood using experimental and datamining approaches

Area of Science:

• Clinical Chemistry
• Laboratory Medicine
• Data Mining Applications

Background:

• Analytical stability of laboratory tests is crucial and typically relies on quality control procedures.
• Summarized patient data offers a supplementary method for monitoring analytical stability.
• Assessing analyte stability in whole blood is essential for accurate diagnostic testing.

Purpose of the Study:

• To evaluate a data mining method for the retrospective assessment of analyte stability in whole blood.
• To compare the data mining approach with traditional incubation-based experiments.
• To demonstrate the method's efficacy using representative laboratory tests.

Main Methods:

• Utilized results from a laboratory information system for a data mining approach.
• Collected blood samples from general practitioners' offices and hospital outpatient clinics.
• Compared data mining results with single-temperature and variable-temperature incubation experiments for potassium, phosphate, and lactate dehydrogenase.

Main Results:

• The data mining method demonstrated comparable results to the reference incubation experiments.
• Analyte stability was assessed by grouping data based on sample collection location and time to analysis.
• Potassium and phosphate analytes showed instability after short-term storage at reduced temperatures.

Conclusions:

• The data mining approach is a viable method for retrospective evaluation of analyte stability in whole blood.
• The study identified specific analytes (potassium, phosphate) susceptible to instability under certain storage conditions.
• Findings highlight the importance of considering sample handling and storage conditions for laboratory test accuracy.