Optimizing moving average control procedures for small-volume laboratories: can it be done?

Summary: Wider implementation of laboratory information systems (LIS) in clinical laboratories in Serbia has been initiated ten years ago. The first LIS in the Railway Health Care Institute has been implemented nine years ago. Before the LIS was initiated, manual admission procedures limited daily output of patients. Moreover, manual entering of patients data and ordering tests on analyzers was problematic and time consuming. After completing tests, laboratory personnel had to write results in patient register (with potential errors) and provide invoices for health insurance organisation. First LIS brought forward some advantages with regards to these obstacles, but it also showed various weaknesses. These can be summarised in rigidity of system and inability to fulfil user expectation. After 4 years of use, we replaced this system with another LIS. Hence, the main aim of this paper is to evaluate advant ages of using LIS in laboratory of the Railway Health Care Institute and also to discuss further possibilities for its application. After implementing LIS, admission procedure has proven to be much faster. LIS enabled electronic requests, barcoded specimens prevent identification errors, bidirectional interface replaces redundant data entry steps, QC data are transferred automatically, results are electronically validated and automatically archived in data base, billing information is transferred electronically, and more. We also use some advanced options, like delta check, HIL feature, quality indicators and various types of reports. All steps in total testing process are drastically improved after the implementation of LIS, which had a positive impact on the quality of issued laboratory results. However, we expect development of some new features in the future, for example auto-verification and inventory management. On the example of the laboratory of the Railway Health Care Institute, we show that it is crucial that laboratory specialists have the main role in defining desirable characteristics of LIS which institution aims to buy. This paper suggests that the main feature of LIS should be the flexibility of system and capability of adjustment to user needs and requests.

Integrating moving average control procedures into the risk-based quality control plan in small-volume medical laboratories

Moving average (MA) is one possible way to use patient results for analytical quality control in medical laboratories. The aims of this study were to: (1) implement previously optimized MA procedures for 10 clinical chemistry analytes into the laboratory information system (LIS); (2) monitor their performance as a real-time quality control tool, and (3) define an algorithm for MA alarm management in a small-volume laboratory to suit the specific laboratory. Moving average alarms were monitored and analysed over a period of 6 months on all patient results (total of 73,059) obtained for 10 clinical chemistry parameters. The optimal MA procedures were selected previously using an already described technique called the bias detection simulation method, considering the ability of bias detection the size of total allowable error as the key parameter for optimization. During 6 months, 17 MA alarms were registered, which is 0.023% of the total number of generated MA values. In 65% of cases, their cause was of pre-analytical origin, in 12% of analytical origin, and in 23% the cause was not found. The highest alarm rate was determined on sodium (0.10%), and the lowest on calcium and chloride. This paper showed that even in a small-volume laboratory, previously optimized MA procedures could be successfully implemented in the LIS and used for continuous quality control. Review of patient results, re-analysis of samples from the stable period, analysis of internal quality control samples and assessment of the analyser malfunctions and maintenance log have been proposed for the algorithm for managing MA alarms.

Background/Objectives: Low-density lipoprotein cholesterol (LDL-C) is a primary therapeutic target in cardiovascular risk assessment. While direct assays are available, LDL-C is frequently estimated using various formulas, whose performance can vary by population and lipid concentration. This study evaluates the analytical agreement and clinical classification performance of four formulas (Spasić–Kotur–Vujović, Friedewald, De Long, and Martin–Hopkins) against directly measured LDL-C. Methods: Lipid profiles from 3935 fasting adult patients were analyzed. LDL-C was directly measured via a homogeneous assay and compared with calculated values using linear regression and Bland–Altman analyses. Analytical bias and classification accuracy were assessed across stratified lipid levels. Results: Among the evaluated equations, the Spasić–Kotur–Vujović formula showed the closest agreement with directly measured LDL-C. It had the lowest mean absolute bias and mean percentage differences across most lipid strata, as well as narrower limits of agreement compared with Friedewald, De Long, and Martin–Hopkins formulas. Although all formulas showed strong correlations with directly measured LDL-C (Pearson’s r correlation coefficient = 0.956–0.958; intraclass correlation coefficient > 0.974), systematic underestimation of LDL-C was observed for all equations and became more pronounced with increasing total cholesterol, LDL-C, triglycerides, and non-high-density lipoprotein cholesterol concentrations. The Spasić–Kotur–Vujović formula achieved the highest overall classification accuracy (70%), followed by De Long (68%), Martin–Hopkins (61%), and Friedewald (58%). Conclusions: Despite strong correlations, clinically relevant differences in bias and classification exist among LDL-C formulas. The Spasić–Kotur–Vujović formula demonstrated superior agreement and accuracy in this population, suggesting it is a more reliable tool for routine laboratory practice. It should be noted that these findings reflect agreement with the specific homogeneous LDL-C assay used on the Alinity c platform (Abbott Laboratories,, Chicago, IL, USA) and may not be directly generalizable to other analytical systems.

Glycated albumin (GA) is a promising glycemic biomarker that offers advantages over glycated hemoglobin (HbA1c), especially under conditions that affect red blood cell turnover. In type 2 diabetes (T2DM), disturbances in glucose and lipid metabolism often coexist. This study aimed at evalute the diagnostic utility of GA, HbA1c, and their composite indices incorporating high-density lipoprotein cholesterol (HDL-C) in a Serbian cohort. A total of 124 adults (including both T2DM patients and healthy controls) were analyzed for GA, HbA1c, glucose, and lipid profiles. Composite indices, namely, GA/HDL-C and HbA1c/HDL-C, were calculated. Logistic regression and ROC analyses were used to assess diagnostic performance. GA and HbA1c levels were significantly higher in T2DM patients, while HDL-C levels were lower. Both GA/HDL-C and HbA1c/HDL-C showed strong associations with T2DM. HbA1c exhibited the highest AUC (0.966), followed by HbA1c/HDL-C (0.883), GA/HDL-C (0.859), and GA (0.855). The GA/HDL-C index exhibited particularly high specificity (92.06%). GA levels in the control group were consistent with reference values reported in other populations, and this is the first study to provide GA data for a Serbian cohort. The findings highlight GA as a useful complementary marker for T2DM. Composite indices, particularly GA/HDL-C and HbA1c/HDL-C, offer enhanced specificity and may serve as simple, cost-effective tools for integrated metabolic risk evaluation and early T2DM detection.