Scandinavian Journal of Clinical & Laboratory Investigation最新文献

The aim of this study was to investigate the effect of vacuum urine collection system on urine sediment and chemical dipstick analysis results by comparing it with a non-vacuum urine collection system. Vacuum urine collection systems are commonly used due to their user-friendly design, standardization, and low contamination risk, but their impact on the accuracy of urinalysis is not well-documented. A total of 124 urine specimens were collected from patients using vacuum and non-vacuum (Argeron UriCollecT) urine collection systems. The samples were analyzed using Dirui FUS-200 sediment and Dirui H-800 dipstick analyzers within 1 h of collection. Microscopic sediment analyses were confirmed when needed. Microscopic urine sediment analyses showed no significant difference between vacuum and non-vacuum systems for leukocytes, squamous epithelium, non-squamous epithelium, and hyaline casts. However, significant differences (p < 0.05) were observed for red blood cells, crystals, and granular casts. In chemical dipstick analysis, no significant difference was found for hemoglobin, bilirubin, ketone, urobilinogen, protein, nitrite, leukocyte, glucose, and pH. However, significant difference was observed in specific gravity (p < 0.05). Vacuum urine collection systems can cause statistically significant variations in certain sediment and chemical dipstick parameters. These differences should be considered in clinical evaluations, and further research with larger and diverse patient samples is needed to assess the clinical implications and effects on patient outcomes.

NT-proBNP is crucial in diagnosing and monitoring heart failure. Data from external quality assessment (EQA) schemes reveal existing method differences among manufacturers. Median values of patient results can serve as a quality indicator. The aim was to study if differences between manufacturers seen in EQA data also are evident in patient medians. Over one million NT-proBNP results from adults during 2011-2021 were extracted from six Swedish counties. Results were grouped by method, county, and according to eGFR and diabetes diagnosis. Medians and consensus values were calculated. From 2011 to 2017, Roche accounted for 76% of NT-proBNP results. After 2017, other manufacturers entered and then methods yielded more diverse results. Findings for Roche and Siemens align with a previous EQA study, while Abbott's results differed. Between 2019 and 2021 Roche results were 2% above consensus, Siemens 16% above, Ortho 8% below, and Abbott 11% below. As the number of NT-proBNP requests increased, medians decreased, possibly due to updated guidelines and more widespread testing. Additionally, patients with eGFR < 60 mL/min/1.73m² BSA or diabetes had medians that were 2-3 times higher compared to the overall results. This study largely confirms varying levels noted in an EQA scheme between manufacturers of NT-proBNP methods, with a notable exception for unknown reasons for Abbott. The expected higher results for patients with eGFR < 60 mL/min/1.73m² BSA or diabetes reinforced the validity of calculated overall medians. Therefore, patient medians are a clinically relevant quality indicator, and a robust approach to monitor methods as a supplement to internal and external controls.

Guidelines on storage for samples intended for Prothrombin Time - International Normalized Ratio (PT-INR) analysis have changed over time, sometimes advising against cold storage due to presumed cold activation of the coagulation cascade. Previous studies on PT-INR storage have mainly been underpowered, performed in glass tubes, and not in a modern laboratory setting. In this study, we re-analyzed 1149 PT-INR samples, divided into low-level samples (PT-INR <1.3), and high-level samples (PT-INR 1.8-3.5) after 3 h of cold storage. We found no statistical difference for high-level samples but statistically higher PT-INR values in low-level samples. The differences were minor and not considered clinically relevant. No cold activation could be detected, as cold activation would have diminished PT-INR. These findings open the possibility of transporting and storing centrifuged PT-INR samples refrigerated. The higher PT-INR values in low-level samples after cold storage impede a mechanistic principle that needs to be further investigated.

Calprotectin is a 24 kD heterodimer of calcium-binding proteins S100A8 and S100A9. At present, there is a lack of knowledge about the specificity of various methods for calprotectin detection, whether they measure only dimers between S100A8 and S100A9, S100A8-S100A8 dimers, S100A9/S100A9 dimers, or free subunits. This study aimed to compare total calprotectin levels with those of its subunits, S100A8 and S100A9, in ICU patients. This prospective observational study includes 271 sepsis and non-sepsis patients. Inclusion criteria were admission to intensive care and the presence or need for an arterial catheter. Plasma total calprotectin was measured at ICU admission and the following two days by particle-enhanced turbidimetric (PETIA) calprotectin reagents from Gentian AS and a Mindray BS380 chemistry analyzer. S100A8 and S100A9 were analyzed by commercial sandwich ELISA DY4570-05, and DY5578, R&D Systems, respectively. Sepsis was defined according to Sepsis-3 as suspected infection and a Sequential organ failure assessment (SOFA) >2 on admission. Receiver operating characteristic (ROC) analysis showed that total calprotectin had a larger area under the curve (AUC) for distinguishing sepsis from non-sepsis patients (0.67) compared to S100A8 (0.59) and S100A9 (0.52). For predicting 30-day mortality, S100A9 had a higher AUC value (0.64) than S100A8 (0.59). However, weak correlations between total calprotectin and its subunits suggest no significant predictive relationship for 30-day mortality, while also highlighting potential assay harmonization challenges across manufacturers.

Given its narrow therapeutic index and risk of nephrotoxicity, vancomycin requires accurate concentration monitoring. While immunoassays such as the Kinetic Interaction of Microparticles in Solution (KIMS) are widely used for their speed and simplicity, they are susceptible to interference by paraproteins such as monoclonal immunoglobulins. Here, we present the first case of falsely elevated vancomycin concentrations measured by KIMS due to monoclonal IgM interference, indicated by a 'kinetic flag' error, and evaluated pre-analytical approaches to mitigate this interference. We applied three pre-analytical techniques to overcome this issue: dithiothreitol (DTT) treatment to reduce disulfide bonds in IgM, ultrafiltration with a 100-kDa cutoff to physically remove IgM molecules, and heterophilic blocking tubes (HBT) to neutralize nonspecific binding by heterophilic antibodies. Vancomycin concentrations were measured using both KIMS (Cobas c702, Roche Diagnostics) and Chemiluminescent Microparticle Immunoassay (CMIA, Architect i2000SR, Abbott Laboratories). In pre-vancomycin-dose samples, both DTT treatment and ultrafiltration resolved the 'kinetic flag' error and corrected vancomycin concentrations to below the lower limit of quantification (LLOQ). In post-vancomycin-dose samples, DTT reduced vancomycin concentration to 16.82 μg/mL (93.8% of the naïve concentration), while ultrafiltration reduced it to 9.22 μg/mL (51.4%). HBT treatment did not resolve the error, and falsely elevated results persisted. This study reports the first case of IgM paraprotein interfering with vancomycin measurement by KIMS, underlining the importance of recognizing monoclonal IgM as a potential source of interference. DTT treatment effectively corrected the interference, while ultrafiltration reduced vancomycin concentration, highlighting the need for careful interpretation.

The haemoglobin variant Hb Tacoma [HBB:c.93G > T;CD30(Arg > Ser) was detected in approximately 1 in 1000 patients tested for HbA1c at Karolinska University Hospital, Stockholm. Incidentally, five homozygous individuals were found, none of them presenting with severe haematological abnormalities. When the cation exchange chromatographic method Bio-Rad Variant II was used, chromatograms representing Hb Tacoma heterozygotes were easily misinterpreted as normal, although the HbA1c results were about 40% lower than those obtained by immunoassays. This is explained by the fact that the glycated form of Hb Tacoma coelutes with the LA1c fraction, while the main fraction of Hb Tacoma is not separated from normal HbA. Automatic calculation of the ratio LA1c/HbA1c was soon introduced and samples, where this ratio exceeded certain limits, were further investigated. In case of Hb Tacoma, HbA1c results were obtained from immunoassays. Several other HbA1c methods were evaluated using samples from Hb Tacoma heterozygotes. Unlike the Variant II method, the HbA1c results from the Bio-Rad D-100 method did not significantly differ from immunoassay. However, the results for Hb Tacoma samples deviated in an unpredictable way. The cation exchange chromatographic method Tosoh G11 constantly showed a positive bias. Hb Tacoma did not significantly affect the HbA1c results using immunoassays like Tina-quant Gen. 3 (Roche) and DCA Vantage (Siemens), an affinity method (Afinion 2, Abbott) or an enzymatic method (Alinity c, Abbott). Although the capillary electrophoresis HbA1c method (Sebia Capillarys 3 Tera) did not demonstrate any significant bias, the imprecision for this method was unacceptably high for samples with Hb Tacoma.

In biological fluids the charge is dominated by strong ions with pH-independent complete dissociation. The requirement of electroneutrality in combination with the principle of mass conservation and rules of dissociation as understood from physical chemistry including auto-dissociation of water allows definite specification of charge in any water-based fluid with known composition. Herein, two important publications are revisited using charge-balance modeling. In the first case it is shown that knowing a fixed pH difference between two types of fibroblasts and knowing also the pH-dependent buffer-capacities in the two cells allows quantitative assessment of buffer concentrations and strong ion differences. In the second example it is shown that the difficult problem of finding by calculation or titration the titratable acidity in urine modeled as a phosphate solution with high ionic strength can be solved with just measuring urine SID and total phosphate concentration without knowing either the correct dissociation constant or pH in urine. It is suggested that charge-balance modeling based on basic physical chemistry may reveal non-trivial ontological details of the system under study.

Diabetes is a growing global health concern affecting millions worldwide. Glucose measurement remains a cornerstone in the diagnosis, monitoring, and management of this disease. Considering that numerous point-of-care (POC) devices for glucose measurements are available, systematic evaluation of their analytical performance and user-friendliness is essential. This study presents the findings of an independent evaluation of the cobas pulse glucose meter (Roche Diagnostics GmbH) conducted by the Scandinavian evaluation of laboratory equipment for point of care testing (SKUP). The study, which assessed both the analytical performance and user-friendliness, included 217 participants with diabetes and was conducted in both a hospital laboratory and primary healthcare settings. Capillary whole blood measurements were compared to a plasma glucose hexokinase method at Vejle hospital, Denmark. Certified reference materials from the National Institute of Standards and Technology (NIST) were used to correct all sample results. Repeatability and accuracy were assessed against predefined analytical performance specifications, and user-friendliness was evaluated using a structured questionnaire developed by SKUP. The cobas pulse demonstrated good repeatability (CV: 1.8-4.0%) and fulfilled the specification for accuracy, with minimal bias (maximum +0.16 mmol/L in the 7-10 mmol/L range). Haematocrit had no significant effect within the tested range (5-70%). User-friendliness was rated satisfactory, attributed to intuitive operation, rapid analysis times, and effective quality control features. In conclusion, the cobas pulse glucose meter met the performance criteria for professional use in clinical and primary care settings, supporting its suitability in near-patient glucose monitoring.

To establish reference intervals (RIs) for the maximal aggregation rate (MA%) mainly of platelet aggregation tests induced by Collagen (COL), epinephrine (EPI), and ristocetin (RIS) using two automated haemostasis analyzers. Platelet aggregation tests were performed on 130 volunteers using CN6000 and CS5100 haemostasis analyzers. RIs were determined by a non-parametric method according to Clinical and Laboratory Standards Institute (CLSI) document EP28-A3c. The coefficient of variation (CV%) of MA% (excluding 0.6 mg/mL RIS) ranged from 3.23% to 6.29%. No significant differences in MA% were observed across gender and age groups. Significant differences were noted between the two analyzers. The RIs induced by COL at 2 µg/mL and 5 µg/mL ranged from 83.4% to 97.5% and 82.0% to 96.9% on the CN6000, and 76.3%-94.4% and 77.1%-95.7% on the CS5100, respectively. For EPI at 5 µmol/L the RIs were 76.0%-94.6% and 70.2%-93.5%. For RIS at 1.2 mg/mL, the RIs were 77.8%-95.0% and 76.4%-91.3%; for RIS at 0.6 mg/mL, were from 0%-3.0% and 0%-3.8% for CN6000 and CS5100, respectively. RIs for MA% induced by COL, EPI and RIS were established using two automated haemostasis analyzers, providing a valuable resource for clinical laboratories.

Knowledge of the pre-analytical stability is crucial when using frozen biobank samples for e.g. reference intervals studies. Such samples are sometimes subjected to multiple freeze-thaw cycles before analysis. There is a lack of comprehensive data on how repeated freezing and thawing affects the stability of many analytes in serum and plasma, particularly when compared to fresh samples. We studied the systematic effects of one to four -80 °C freeze-thaw cycles on alanine transaminase, albumin, alkaline phosphatase (ALP), alpha-1-antitrypsin, amylase, apolipoprotein A1 and B, aspartate transaminase, C-reactive protein (CRP), calcium, ceruloplasmin, chloride, creatine kinase (CK), creatinine, cystatin C, ferritin, gamma-glutamyl transferase, glucose, haptoglobin, high-density lipoprotein (HDL) cholesterol, immunoglobulin A, G and M, interleukin-6 (IL-6), iron, lipase, lipoprotein (a), low-density lipoprotein (LDL) cholesterol, magnesium, orosomucoid (alpha-1-acid glycoprotein), phosphate, potassium, sodium, total cholesterol, total protein, transferrin, transthyretin (prealbumin), triglyceride, urate and urea (carbamide) in serum pools and cardiac troponin T and N-terminal pro-brain natriuretic peptide (NT-proBNP) in serum and plasma pools. Mean relative changes after repeated freezing and thawing compared to samples analyzed fresh and samples frozen and thawed once were estimated. Most analytes were stable for up to four freeze-thaw cycles compared to fresh samples according to at least the minimum allowable bias (AB) based on biological variation. Analytes found to be unstable according to at least the desirable AB criterion after at least one of the freeze-thaw cycles were HDL cholesterol, chloride, and sodium.