Evaluation of a Six Sigma-Based Dynamic Quality Control Strategy for Hematology Analysis: A Multicenter Study.

Area of Science:

• Clinical diagnostics
• Laboratory medicine
• Quality management systems

Background:

• Traditional quality control (QC) in hematology testing faces limitations in timely error detection and adaptability to real-world lab conditions.
• Ensuring accuracy and reliability in hematology diagnostics is crucial for patient care.

Purpose of the Study:

• To develop and validate a dynamic QC strategy for hematology testing.
• To enhance analytical error detection and improve the reliability of hematology results.

Main Methods:

• Application of the Six Sigma quality management framework to evaluate five key hematology parameters (Hb, WBC, RBC, HCT, PLT).
• Implementation of a dynamic QC strategy integrating moving average (MA) monitoring with a long short-term memory (LSTM) predictive model.
• Validation using patient sample data alongside routine QC data for clinical adaptability assessment.

Main Results:

• Sigma metrics indicated variable performance, with Hemoglobin (Hb) and White Blood Cell (WBC) counts showing excellent (σ ≥ 6) performance, while Platelet (PLT) counts exhibited lower stability.
• The combined MA-LSTM approach demonstrated superior sensitivity in detecting analytical errors and reduced false positives compared to conventional rule-based QC.
• The dynamic model provided robust predictive capabilities for real-time QC monitoring across multiple sites.

Conclusions:

• A novel dynamic QC strategy combining Six Sigma, MA, and LSTM modeling addresses limitations of traditional methods.
• This intelligent, proactive, and clinically adaptable solution enhances hematology testing reliability.
• The proposed strategy contributes to improved quality patient care through more dependable diagnostic results.