The multi-item univariate delta check method: a new approach.

Area of Science:

• Clinical Laboratory Science
• Quality Control in Healthcare
• Medical Diagnostics

Background:

• Delta check methods are crucial for detecting random errors in clinical laboratory tests.
• Multivariate delta checks are superior but difficult to implement practically.
• There is a need for efficient and effective univariate delta check methods.

Purpose of the Study:

• To introduce the multi-item univariate delta check (MIUDC) method.
• To determine the optimal threshold (k) for triggering detailed investigations.
• To identify laboratory test items that warrant increased scrutiny.

Main Methods:

• The proposed MIUDC method involves performing univariate delta checks across multiple laboratory test items.
• Specimens with positive delta checks in at least 'k' items are flagged for further review.
• Real data and simulation studies were used to evaluate the method and determine 'k'.

Main Results:

• An appropriate value for 'k' was determined to be 4, balancing checking volume and efficiency.
• Total cholesterol, albumin, and total protein were identified as key test items with zero false positive rates in simulations.
• The MIUDC method demonstrated ease of implementation and high efficiency.

Conclusions:

• The MIUDC method offers a practical and efficient approach to quality control in clinical laboratories.
• Setting k=4 provides a good balance for error detection and workload management.
• Specific analytes like total cholesterol, albumin, and total protein are prime candidates for enhanced monitoring using MIUDC.