Patient-based real-time quality control management enhanced with biological variation data

Ensuring and improving quality control in medical laboratories is a primary goal for specialists in medical biochemistry. In this context, alongside internal quality control performed at the beginning of analyses and external quality control applied at wider intervals, the need for an additional system providing real-time assurance has gained significance. This study evaluates the capability of Patient-Based Real-Time Quality Control (PBRTQC) methods to detect allowable total error levels. Data from biochemical tests analyzed between 01.01.2023 – 31.12.2023 were examined, and 23 different biochemical tests were selected based on sigma levels and average daily result counts (albumin, alkaline phosphatase (ALP), alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), calcium (Ca), chloride (Cl), iron, phosphorus, gamma-glutamyl transferase (GGT), glucose, high-density lipoprotein (HDL), potassium (K), creatinine, lactate dehydrogenase (LDH), low-density lipoprotein (LDL), magnesium, sodium (Na), total bilirubin, total cholesterol, total protein, triglyceride, and uric acid). Bias simulations were applied at various rates to these tests. Subsequently, Exponentially Weighted Moving Average (EWMA) and Moving Average (MA) methods were used to conduct power function analysis, bias simulation analysis, and Youden index comparison. The results indicated that EWMA generally outperformed MA. Throughout the study, the most effective methods and sub-parameters were determined for each test parameter; graphical systems were developed to facilitate device-based and measurement method-based comparisons. It was concluded that, in routine practice, the use of PBRTQC methods with specific parameters tailored for each test, along with device-based and measurement method-based comparison graphics, would be beneficial. Keywords: patient based real time quality control, quality control, total allowable error, exponentially weighted moving average, moving average, power function analysis, bias simulation analysis, Youden index.