Effective detection of bias and variability using R-value statistics in the even-check method for real-time quality...

Patient-based real-time quality control (PBRTQC) is expected to enhance patient safety by improving accuracy of clinical testing, but its adoption remains limited. The Even Check Method (ECM) converts Δ values into simple positive or negative signs, enabling detection of systematic errors through R-values. The potential benefits of incorporating mean and standard deviation (SD) of R-values as auxiliary indicators for more sensitive monitoring have not been fully explored.

Clinical data for 20 different laboratory test items were collected at Tenri Hospital before and after analytical equipment replacement. The R-values, R-scores, and the mean and SD of R-values were calculated in real time using patients' historical data. Internal quality control (iQC) samples were measured daily, and SDs and coefficients of variation were compared before and after equipment replacement.

Analytical stability improved after equipment replacement, with decreased SDs and coefficients of variation in both iQC and R-values. In sodium measurements, ECM detected a + 3 mmol/L bias earlier and more sensitively than conventional iQC, enabling prompt corrective action via electrode replacement. Monitoring mean and SD of R-values allowed detection of both acute and gradual systematic errors and supported real-time evaluation of analytical performance and maintenance timing across multiple test items. This approach also facilitates interpretation by laboratory staff and practical integration into routine testing.

Incorporating the mean and SD of R-values into ECM enhances PBRTQC by improving systematic error detection, supporting timely corrective actions, and enabling routine monitoring, thereby strengthening laboratory reliability and contributing to patient safety.