Recommendation to treat continuous variable errors like attribute errors

Area of Science:

• Clinical Chemistry
• Laboratory Medicine
• Medical Diagnostics

Background:

• Clinical laboratory errors are classified as attribute (pre-analytical, post-analytical) or continuous (analytical) variables.
• Current error goals for continuous variables (e.g., 95% confidence intervals) differ significantly from attribute errors, potentially masking a high rate of medically unacceptable analytical errors.
• This discrepancy highlights a need for improved methods to assess and control analytical errors.

Purpose of the Study:

• To propose a method for reclassifying continuous analytical errors into attribute errors based on severity.
• To improve the assessment and management of clinical laboratory errors, particularly analytical ones.
• To enhance patient safety by addressing the most critical laboratory result deviations.

Main Methods:

• Classifying analytical error rates into distinct severity categories, similar to existing glucose error grids.
• Counting the number of results falling into each defined error grid zone.
• Transforming continuous variable errors into attribute errors for more effective evaluation, following FDA recommendations.

Main Results:

• The proposed method effectively converts continuous analytical errors into attribute errors by categorizing them by severity.
• This approach allows for a more accurate representation of medically unacceptable errors compared to traditional uncertainty intervals.
• The method aligns with FDA recommendations for error assessment, focusing on the most severe outcome zones.

Conclusions:

• Classifying analytical errors into severity categories, akin to error grids, is a superior method for managing laboratory quality.
• This transformation of continuous errors into attribute errors provides a clearer picture of potential patient harm.
• Implementing severity-based classification enhances the ability to control and reduce the impact of critical laboratory errors.