When bias becomes part of imprecision: how to use analytical performance specifications to determine acceptability of lot-lot variation and other sources of possibly unacceptable bias.

ISO 15189 requires laboratories to estimate the uncertainty of their quantitative measurements and to maintain them within relevant performance specifications. Furthermore, it refers to ISO TS 20914 for instructions on how to estimate the uncertainty and what to take into consideration when communicating uncertainty of measurement with requesting clinicians. These instructions include the responsibility of laboratories to verify that bias is not larger than medically significant. If estimated to be larger than acceptable, such bias first needs to be eliminated or (temporarily) corrected for. In the latter case, the uncertainty of such correction becomes part of the estimation of the total measurement uncertainty. If small enough to be acceptable, bias becomes part of the long term within laboratory random variation. Sources of possible bias are (not limited to) changes in reagent or calibrator lot variation or calibration itself. In this paper we clarify how the rationale and mathematics from an EFLM WG ISO/A position paper on allowable between reagent lot variation can be applied to calculate whether bias can be accepted to become part of long-term imprecision. The central point of this rationale is to prevent the risk that requesting clinicians confuse changes in bias with changes in the steady state of their patients.

Cost-effectiveness of home-based screening of the general population for albuminuria to prevent progression of cardiovascular and kidney disease.

Background: Chronic kidney disease (CKD) is often detected late, leading to substantial health loss and high treatment costs. Screening the general population for albuminuria identifies individuals at high risk of kidney events and cardiovascular disease (CVD) who may benefit from early start of preventive interventions. Previous studies on the cost-effectiveness of albuminuria population screening were inconclusive, but were based on survey or cohort data rather than an implementation study, modelled screening as performed by general practitioners rather than home-based screening, and often included only benefits with respect to kidney events. We evaluated the cost-effectiveness of home-based general population screening for increased albuminuria based on real-world data obtained from a prospective implementation study taking into account prevention of CKD as well as CVD events. Methods: We developed an individual-level simulation model to compare home-based screening using a urine collection device with usual care (no home-based screening) in individuals of the general population aged 45-80, based on the THOMAS study (Towards HOMe-based Albuminuria Screening). Cost-effectiveness was assessed from the Dutch healthcare perspective with a lifetime horizon. The costs of the screening process and benefits of preventing CKD progression (dialysis and kidney transplantation) and CVD events (non-fatal myocardial infarction, non-fatal stroke, fatal CVD event) were reflected. Albuminuria detection led to treatment of identified risk factors. The model subsequently simulated CKD progression, the occurrence of CVD events, and death. The risks of experiencing CVD events were calculated using the SCORE2 CKD risk prediction model and individual-level data from the THOMAS study. Relative treatment effectiveness, quality of life scores, resource use, and cost inputs were obtained from literature. Model outcomes were the number of CKD and CVD-related events, total costs, quality-adjusted life years (QALYs), and the incremental cost-effectiveness ratio (ICER) per QALY gained by screening versus usual care. All results were obtained through probabilistic analysis. Findings: The absolute difference between screening versus usual care in lifetime probability of dialysis, kidney transplantation, non-fatal myocardial infarction, non-fatal stroke, and fatal CVD events were 0.2%, 0.05%, 0.6%, 0.6%, and 0.2%, respectively. This led to relative decreases compared to usual care in lifetime incidence of these events of 10.7%, 11.1%, 5.1%, 4.1%, and 1.6%, respectively. The incremental costs and QALYs of screening were €1607 and 0.17 QALY, respectively, which led to a corresponding ICER of €9225/QALY. The probability of screening being cost-effective for the Dutch willingness-to-pay threshold for preventive population screening of €20,000/QALY was 95.0%. Implementing the screening in the subgroup of 45-64 years old reduced the ICER (€7946/QALY), whereas implementing screening in the subgroup of 65-80 years old increased the ICER (€10,310/QALY). A scenario analysis assuming treatment optimization in all individuals with newly diagnosed risk factors or known risk factors not within target range reduced the ICER to €7083/QALY, resulting from the incremental costs and QALY gain of €2145 and 0.30, respectively. Interpretation: Home-based screening for increased albuminuria to prevent CVD and CKD events is likely cost-effective. More health benefits can be obtained by screening younger individuals and better optimization of care in individuals identified with newly diagnosed or known risk factors outside target range. Funding: Dutch Kidney Foundation, Top Sector Life Sciences & Health of the Dutch Ministry of Economic Affairs.

Abnormal Iron Status and Adverse Outcome After Elective Cardiac Surgery: A Prospective, Observational Multicenter Study.

OBJECTIVES: To investigate the incidence of preoperative abnormal iron status and its association with packed red blood cell (PRBC) transfusion, postoperative major complications, and new onset of clinically significant disability in patients undergoing elective cardiac surgery. DESIGN: A prospective, observational multicenter cohort study. SETTING: Three cardiac surgical centers in the Netherlands between 2019 and 2021. Recruitment was on hold between March and May 2020 due to COVID-19. PATIENTS: A total of 427 patients aged 60 years and older who underwent elective on-pump cardiac surgery. MEASUREMENTS AND MAIN RESULTS: The primary endpoint was a 30-day PRBC transfusion. Secondary endpoints were postoperative major complications within 30 days (eg, acute kidney injury, sepsis), and new onset of clinically significant disability within 120 days of surgery. Iron status was evaluated before surgery. Abnormal iron status was present in 45.2% of patients (n = 193), and most frequently the result of iron deficiency (27.4%, n = 117). An abnormal iron status was not associated with PRBC transfusion (adjusted relative risk [ARR] 1.2; 95% CI 0.9-1.8: p = 0.227) or new onset of clinically significant disability (ARR 2.0; 95% CI 0.9-4.6: p = 0.098). However, the risk of postoperative major complications was increased in patients with an abnormal iron status (ARR 1.7; 95% CI 1.1-2.5: p = 0.012). CONCLUSIONS: An abnormal iron status before elective cardiac surgery was associated with an increased risk of postoperative major complications but not with PRBC transfusion or a new onset of clinically significant disability.

Participation rate and yield of two home-based screening methods to detect increased albuminuria in the general population in the Netherlands (THOMAS): a prospective, randomised, open-label implementation study.

BACKGROUND: Chronic kidney disease (CKD) has a rising global prevalence and is expected to become the fifth leading cause of death by 2030. Increased albuminuria defines the early stages of CKD and is among the strongest risk factors for progressive CKD and cardiovascular disease. The value of population screening for albuminuria to detect CKD in an early phase has yet to be studied. We aimed to evaluate the effectiveness of two home-based albuminuria population screening methods. METHODS: Towards Home-based Albuminuria Screening (THOMAS) was a prospective, randomised, open-label implementation study that invited Dutch adults aged 45-80 years for albuminuria screening. Individuals were randomly assigned (1:1) to screening by applying either a urine collection device (UCD) that was sent by post to a central laboratory for measurement of the albumin-to-creatinine ratio (ACR) by immunoturbidimetry or to screening via a smartphone application that measures the ACR with a dipstick method at home. Randomisation was done with a four-block method via a web-based system and was stratified by age, sex, and socioeconomic status. If two or more individuals per household were invited to participate, these individuals were randomly assigned to the same group. In case of confirmed increased albuminuria at home, participants were invited for an elaborate screening in a regional hospital (Amphia Hospital, Breda, Netherlands) for CKD and cardiovascular risk factors. When abnormalities were found, participants were referred to their general practitioner for treatment. The primary outcomes were the participation rate and yield of the home-based screening and elaborate screening. Participation rate was assessed in the intention-to-screen population (ie, all participants who were invited for the home-based screening or elaborate screening). Yield was assessed in the per-protocol population (ie, all individuals who participated in the home-based screening or elaborate screening). An exploratory analysis assessed the sensitivity and specificity of both home-based screening methods. To this end, an additional quantitative ACR test was performed among people participating in the elaborate screening, and a substudy was performed among participants with a first negative home-based screening test, who were invited for an additional test. The study is registered with ClinicalTrials.gov, NCT04295889. FINDINGS: 15 074 participants were enrolled between Nov 14, 2019, and March 19, 2021. 7552 (50·1%) were randomly assigned to home-based albuminuria screening by the UCD method and 7522 (49·9%) were assigned to albuminuria screening by the smartphone application method. The participation rate of the home-based screening was 4484 (59·4% [95% CI 58·3-60·5]) of the 7552 invited individuals for the UCD method and 3336 (44·3% [43·2-45·5]) of 7522 invited individuals for the smartphone application method (p<0·0001). Increased ACR was confirmed by home-based testing in 150 (3·3% [95% CI 2·9-3·9]) of 4484 individuals for the UCD method and 171 (5·1% [4·4-5·9]) of 3336 indivduals for the smartphone application method. 124 (82·7% [95% CI 75·8-87·9]) of 150 individuals assigned to the UCD method and 142 (83·0% [76·7-87·9]) of 171 participants assigned to the smartphone application method attended the elaborate screening. Sensitivity to detect increased ACR was 96·6% (95% CI 91·5-99·1) for the UCD method and 98·1% (89·9-99·9) for the smartphone application method, and specificity was 97·3% (94·7-98·8) for the UCD method and 67·9% (62·0-73·3) for the smartphone application method, indicating that the test characteristics of only the UCD method were sufficient for screening. Albuminuria, hypertension, hypercholesterolaemia, and decreased kidney function were newly diagnosed in 77 (62·1%), 44 (35·5%), 30 (24·2%), and 27 (21·8%) of 124 participants for the UCD method, respectively. Of the 124 participants assigned to the UCD method who completed elaborate screening, 111 (89·5%) were referred to their general practitioner for treatment because of newly diagnosed CKD or cardiovascular disease risk factors or known risk factors outside the target range. INTERPRETATION: Home-based screening of the general population for increased ACR using a UCD had a high participation rate and correctly identified individuals with increased albuminuria and yet unknown or known but outside target range CKD and cardiovascular risk factors. By contrast, the smartphone application method had a lower at-home participation rate than the UCD method and the test specificity was too low to accurately assess individuals for risk factors during the elaborate screening. The UCD screening strategy could allow for an early start of treatment to prevent progressive kidney function loss and cardiovascular disease in patients with CKD. FUNDING: Dutch Kidney Foundation, Top Sector Life Sciences & Health of the Dutch Ministry of Economic Affairs.

Protocol for a randomized study assessing the feasibility of home-based albuminuria screening among the general population: The THOMAS study.

BACKGROUND: Chronic kidney disease (CKD) is a rising public health problem that may progress to kidney failure, requiring kidney replacement therapy. It is also associated with an increased incidence of cardiovascular disease (CVD). Because of its asymptomatic nature, CKD is often detected in a late stage. Population screening for albuminuria could allow early detection of people with CKD who may benefit from preventive treatment. In case such screening is performed in a general practitioner (GP) setting, this will result in relatively high costs. Home-based screening might be an effective and cost-effective alternative. AIM: The THOMAS study (Towards HOMe-based Albuminuria Screening) is designed to prospectively investigate two methods for home-based population screening for increased albuminuria to detect yet undiagnosed CKD and risk factors for progression and CVD. METHODS: This investigator initiated, randomized population-based study will include 15.000 individuals aged 45-80 years, who will be randomly assigned to be invited for a home-based screening test for albuminuria with a more conventional urine collection device or an innovative smartphone application. If the test result is positive upon confirmation (i.e., elevated albuminuria), participants are invited to a central screening facility for an elaborate screening for CKD and CVD risk factors. Participants are referred to their GP for appropriate treatment, if abnormalities are found. Primary endpoints are the participation rate, yield, and cost-effectiveness of the home-based screening and elaborate screening. CONCLUSIONS: The THOMAS study will evaluate the effectiveness and cost-effectiveness of home-based albuminuria screening in the general population for the early detection of CKD and CVD risk factors. It will provide insight into the willingness to participate in population screening for CKD and into the compliance of the general population to a corresponding screening protocol and compliance to participate. Thus, it may help to develop an attractive novel screening strategy for the early detection of CKD. TRIAL REGISTRATION: ClinicalTrials.gov, registration number NCT04295889, registered 05 March 2020. https://www.google.com/search?client=firefox-b-d&q=NCT04295889.

Rapid identification of SARS-CoV-2-infected patients at the emergency department using routine testing.

Objectives: The novel coronavirus disease 19 (COVID-19), caused by SARS-CoV-2, spreads rapidly across the world. The exponential increase in the number of cases has resulted in overcrowding of emergency departments (ED). Detection of SARS-CoV-2 is based on an RT-PCR of nasopharyngeal swab material. However, RT-PCR testing is time-consuming and many hospitals deal with a shortage of testing materials. Therefore, we aimed to develop an algorithm to rapidly evaluate an individual's risk of SARS-CoV-2 infection at the ED. Methods: In this multicenter retrospective study, routine laboratory parameters (C-reactive protein, lactate dehydrogenase, ferritin, absolute neutrophil and lymphocyte counts), demographic data and the chest X-ray/CT result from 967 patients entering the ED with respiratory symptoms were collected. Using these parameters, an easy-to-use point-based algorithm, called the corona-score, was developed to discriminate between patients that tested positive for SARS-CoV-2 by RT-PCR and those testing negative. Computational sampling was used to optimize the corona-score. Validation of the model was performed using data from 592 patients. Results: The corona-score model yielded an area under the receiver operating characteristic curve of 0.91 in the validation population. Patients testing negative for SARS-CoV-2 showed a median corona-score of 3 vs. 11 (scale 0-14) in patients testing positive for SARS-CoV-2 (p<0.001). Using cut-off values of 4 and 11 the model has a sensitivity and specificity of 96 and 95%, respectively. Conclusions: The corona-score effectively predicts SARS-CoV-2 RT-PCR outcome based on routine parameters. This algorithm provides the means for medical professionals to rapidly evaluate SARS-CoV-2 infection status of patients presenting at the ED with respiratory symptoms.

Fast 0/1-h algorithm for detection of NSTEMI: are current high-sensitivity cardiac troponin assays fit for purpose? An EQA-based evaluation.

Background High-sensitivity cardiac troponin T/I (hs-cTnT/I) assays have improved analytical sensitivity for the detection of myocardial infarction (MI). To gain clinical specificity and sensitivity, interpretation of changes in cTn concentrations over time is crucial. The 2015 ESC NSTEMI guideline defines absolute delta values as additional rule-in and rule-out criteria for MI. A critical assumption for application of this rule is that total analytical imprecision within the delta period, including inter-instrument bias, is comparable to analytical imprecision in the validation studies. Methods Data from the Dutch External Quality Assessment Scheme (EQAS) were used to calculate inter-instrument bias and estimate imprecision for the measuring range where the proposed delta values are relevant: for Roche Elecsys hs-cTnT, 5-52 and 5-12 ng/L; for Abbott Architect hs-cTnI, 2-52 and 2-5 ng/L for rule-in and rule-out, respectively. Results For Elecsys, the median inter-instrument bias is 0.3 ng/L (n = 33 laboratories), resulting in reference change values (RCVs) of 3.0 and 1.7 ng/L, respectively, for rule-in and rule-out with imprecision as claimed by the manufacturer. With RCVs smaller than the guideline's delta thresholds, 100% of the laboratories have adequate specifications. RCVs for rule-in/rule-out increased to 4.6 ng/L/2.5 ng/L, respectively, with individual imprecisions as estimated from EQA data, resulting in 64% and 82% of laboratories with adequate specifications. For Architect, 40% of instruments (n = 10) might falsely qualify the result as clinically relevant; hence, inter-instrument bias could not be determined. Conclusions We advise laboratories that use the fast 0/1-h algorithm to introduce stringent internal quality procedures at the relevant/low concentration level, especially when multiple analyzers are randomly used.

Cumulative dose of bevacizumab associates with albuminuria rather than podocyturia in cancer patients.

Angiogenesis inhibition with bevacizumab, a monoclonal antibody against vascular endothelial growth factor A (VEGF-A), is an anticancer treatment associated with hypertension and renal glomerular toxicity referred to as a preeclampsia-like syndrome. In preeclampsia, podocyturia predates proteinuria and clinical features of preeclampsia, and is regarded as a biomarker of ongoing glomerular injury. Using a quantitative polymerase chain reaction of the podocyte-specific molecules nephrin, podocin, and VEGF-A in the urine, we examined whether podocyturia is present in bevacizumab-treated cancer patients, and whether it relates to proteinuria and the cumulative dose of bevacizumab. Urine samples were cross-sectionally collected from 43 bevacizumab-treated patients, 21 chemotherapy-treated patients, and 7 healthy controls. Urinary protein-to-creatinine ratio (mean and range) was 32.0 mg/mmol (5.2-284.4) in the bevacizumab group, compared with 11.4 mg/mmol (1.1-21.0) in the chemotherapy group and 7.4 mg/mmol (3.9-16.5) (P < .05) in healthy controls, whereas urinary albumin-to-creatinine ratio values in the three groups were, respectively, 18.9 mg/mmol (0.1-227.7), 1.5 mg/mmol (0.2-3.5), and 0.2 mg/mmol (0.1-0.4) (P < .05). The cumulative dose of bevacizumab ranged from 550 to 93,628 mg. Urinary podocin mRNA expression was undetectable in 59% of participants, urinary nephrin mRNA expression per mmol creatinine ranged from 0.0 to 5.3 and urinary VEGF-A mRNA expression from 0.0 to 2.7. Urinary nephrin mRNA expression did not correlate to the albumin-to-creatinine ratio or the cumulative dose of bevacizumab, whereas the latter correlated with the albumin-to-creatinine ratio (r = 0.77; P < .001). Our results demonstrate that the cumulative dose of bevacizumab is closely correlated with albuminuria but not with podocyturia as measured with the quantitative polymerase chain reaction technique, challenging the feasibility of this measurement to monitor ongoing glomerular injury in patients chronically treated with bevacizumab.

Technical and clinical validation of the Greiner FC-Mix glycaemia tube.

BACKGROUND: Measurement of adequate glucose concentrations is complicated by in vitro breakdown of glucose due to glycolysis. Unlike the commonly used NaF-EDTA and NaF-oxalate phlebotomy tubes, citrated NaF-EDTA tubes are reported to directly and thereby completely inhibit glycolysis. Recently, Greiner introduced the Vacuette® FC-Mix NaF-EDTA-citrate tube, currently the only NaF-citrate tube without volume-disturbing liquid additions available on the European market. Here we present its potential as alternative for the laborious and therefore unfeasible conditions for glucose sampling as recommended by the World Health Organization (WHO). METHODS: The FC-Mix tube was tested against the WHO recommended method of optimal laboratory conditions, both in healthy volunteers and pregnant woman undergoing oral glucose tolerance test (oGTT) for screening of gestational diabetes mellitus (GDM). Glucose concentrations were measured after different incubation times (0-48 h) and temperatures (room temperature, 37 °C), both in uncentrifuged whole blood and centrifuged material. RESULTS: Deming regression analysis shows that glucose concentrations measured in the FC-Mix tube correlate to the WHO recommended method. Stability is maintained at room temperature for 48 h and at least 24 h at 37 °C. The use of the FC-Mix tube was also validated in screening for GDM and proved comparable to the WHO recommended method in diagnostic outcome. CONCLUSIONS: The new Greiner FC-Mix tube combines the easy handling of a routine tube with dry additive with the ability to immediately inhibit glycolysis as in the WHO method for optimal pre-analytical and analytical conditions and performs equally to those conditions when screening for GDM.

Expressing analytical performance from multi-sample evaluation in laboratory EQA.

BACKGROUND: To provide its participants with an external quality assessment system (EQAS) that can be used to check trueness, the Dutch EQAS organizer, Organization for Quality Assessment of Laboratory Diagnostics (SKML), has innovated its general chemistry scheme over the last decade by introducing fresh frozen commutable samples whose values were assigned by Joint Committee for Traceability in Laboratory Medicine (JCTLM)-listed reference laboratories using reference methods where possible. Here we present some important innovations in our feedback reports that allow participants to judge whether their trueness and imprecision meet predefined analytical performance specifications. METHODS: Sigma metrics are used to calculate performance indicators named 'sigma values'. Tolerance intervals are based on both Total Error allowable (TEa) according to biological variation data and state of the art (SA) in line with the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) Milan consensus. RESULTS: The existing SKML feedback reports that express trueness as the agreement between the regression line through the results of the last 12 months and the values obtained from reference laboratories and calculate imprecision from the residuals of the regression line are now enriched with sigma values calculated from the degree to which the combination of trueness and imprecision are within tolerance limits. The information and its conclusion to a simple two-point scoring system are also graphically represented in addition to the existing difference plot. CONCLUSIONS: By adding sigma metrics-based performance evaluation in relation to both TEa and SA tolerance intervals to its EQAS schemes, SKML provides its participants with a powerful and actionable check on accuracy.

Recommendation for the review of biological reference intervals in medical laboratories.

This document is based on the original recommendation of the Expert Panel on the Theory of Reference Values of the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC), updated guidelines were recently published under the auspices of the IFCC and the Clinical and Laboratory Standards Institute (CLSI). This document summarizes proposals for recommendations on: (i) The terminology, which is often confusing, noticeably concerning the terms of reference limits and decision limits. (ii) The method for the determination of reference limits according to the original procedure and the conditions, which should be used. (iii) A simple procedure allowing the medical laboratories to fulfill the requirements of the regulation and standards. The updated document proposes to verify that published reference limits are applicable to the laboratory involved. Finally, the strengths and limits of the revised recommendations (especially the selection of the reference population, the maintenance of the analytical quality, the choice of the statistical method used…) will be briefly discussed.

Accreditation process in European countries - an EFLM survey.

BACKGROUND: Accreditation is a valuable resource for medical laboratories. The development of quality systems based on ISO 15189 has taken place in many laboratories in the European countries but data about accreditation remain scarce. The EFLM Working Group "Accreditation and ISO/CEN standards" conducted a survey that reviews the current state of the accreditation process in European countries. METHODS: An on-line questionnaire was addressed to delegates of 39 EFLM scientific societies in March 2014. One answer by country was taken into account. The survey was dealing with mandatory status, number of accredited medical laboratories in each country, possibility of flexible scope and concerned medical fields. The status of point-of-care testing (POCT) in each country was also studied. RESULTS: Twenty-nine responses (74%) were registered. All the assessed countries (100%) have begun an accreditation process in various ways. All the national accreditation bodies (NAB) offer or are working to offer an ISO 15189 accreditation. The accreditation process most often concerns all phases of the examination and various medical fields. Medical laboratories are responsible for POCT in 20 (69%) countries. The accreditation process for POCT, according to ISO 15189 and ISO 22870, is also developing. CONCLUSIONS: While there are several variations in the approaches to accreditation of medical laboratories in the European countries, the ISO 15189 accreditation project has been widely accepted. The use of a unique standard and the cooperation among countries due to scientific societies, EFLM, accreditation bodies and EA enable laboratory professionals to move toward uniform implementation of the accreditation concept.

Pregnancy diabetes: A comparison of diagnostic protocols based on point-of-care, routine and optimized laboratory conditions.

In vitro glycolysis poses a problem during diabetes screening, especially in remote laboratories. Point-of-care analysis (POC) may provide an alternative. We compared POC, routine and STAT analysis and a feasible protocol during glucose tolerance test (GTT) for pregnancy diabetes (GDM) screening. In the routine protocol, heparin tubes were used and turn-around-time (TAT) was unsupervised. In the STAT protocol, tubes were processed immediately. The feasible protocol comprised of citrated tubes with a TAT of 1 hour. Outcome was defined as glucose concentration and clinical diagnosis. Glucose measured by POC was higher compared to routine analysis at t = 0 (0.25 mM) and t = 120 (1.17 mM) resulting in 17% more GDM diagnoses. Compared to STAT analysis, POC glucose was also higher, although less pronounced (0.06 and 0.9 mM at t = 0 and t = 120 minutes, respectively) and misclassification was only 2%. Glucose levels and clinical diagnosis were similar using the feasible protocol and STAT analysis (0.03 mM and -0.07 mM at t = 0 and t = 120, 100% identical diagnoses). POC is an viable alternative for STAT glucose analysis in GDM screening (sensitivity: 100%, specificity: 98%). A feasible protocol (citrated phlebotomy tubes with a TAT of 60 minutes) resulted in 100% identical outcome and provides the best alternative.