Data flow in clinical laboratories: could metadata and peridata bridge the gap to new AI-based applications?

In the last decades, clinical laboratories have significantly advanced their technological capabilities, through the use of interconnected systems and advanced software. Laboratory Information Systems (LIS), introduced in the 1970s, have transformed into sophisticated information technology (IT) components that integrate with various digital tools, enhancing data retrieval and exchange. However, the current capabilities of LIS are not sufficient to rapidly save the extensive data, generated during the total testing process (TTP), beyond just test results. This opinion paper discusses qualitative types of TTP data, proposing how to divide laboratory-generated information into two categories, namely metadata and peridata. Being both metadata and peridata information derived from the testing process, it is proposed that the first is useful to describe the characteristics of data, while the second is for interpretation of test results. Together with standardizing preanalytical coding, the subdivision of laboratory-generated information into metadata or peridata might enhance ML studies, also by facilitating the adherence of laboratory-derived data to the Findability, Accessibility, Interoperability, and Reusability (FAIR) principles. Finally, integrating metadata and peridata into LIS can improve data usability, support clinical utility, and advance AI model development in healthcare, emphasizing the need for standardized data management practices.

Critical Results in Laboratory Medicine.

BACKGROUND: Timely and accurate notification of critical results is crucial in laboratory medicine and mandated by accreditation standards like ISO15189. Alert lists do, however, vary widely and clinical laboratories typically rely on a combination of in-house agreed and/or literature-based critical values. Communication by phone is still the preferred method of notification, but digital communication could help improve communication of critical results. CONTENT: We review the available evidence concerning critical result thresholds and critical result notification practices. The evidence is ranked using an adaptation of the Stockholm Hierarchy. In addition, we propose an evidence-based list of critical result thresholds for hospitalized patients that laboratories can use as a starter list and further customize based on the clinical needs of their patient population. SUMMARY: A clear distinction between critical results and significantly abnormal results is essential for effective and timely healthcare interventions. Implementation of a policy using differentiated thresholds taking into account individual patient characteristics and how fast medical attention is needed, and the use alternative communication methods could enhance communication efficiency and reduce notification fatigue.

IFN-α2 Autoantibody Screening and Functional Evaluation in Viral and Bacterial Infections.

BACKGROUND: The presence of anti-interferon (IFN)-α2 autoantibodies is a strong indicator of severe disease course during viral infections and is observed in autoimmune diseases (e.g., myasthenia gravis). Detection of these autoantibodies during severe bacterial infections is understudied. Multiple anti-IFN-α2 autoantibody screening assays are available. However, the results do not always correlate with the neutralizing capacity of the autoantibodies. METHODS: Anti-IFN-α2 antibodies were measured by a Luminex-based assay in serum samples from individuals admitted to the intensive care unit infected with influenza (n = 38), invasive bacteria (n = 152), and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (n = 52). Anti-IFN-α2 antibodies were also studied in individuals with myasthenia gravis (n = 22) and in healthy individuals (n = 37). Individuals testing positive by Luminex were subsequently tested by enzyme-linked immunosorbent assay (ELISA) and tested for nonspecific reactivity and neutralization. RESULTS: Three of 16 Luminex-positive samples had nonspecific reactivity, 11/16 were positive by ELISA, and 10/16 had neutralizing activity. Anti-IFN-α2 antibodies were found in individuals infected with SARS-CoV-2 (7/52), influenza (3/38), invasive bacteria [2/152, of which 1 was Legionella pneumophilia and was 1 Escherichia coli (E. coli) (out of 39 E. coli infections)], and in individuals with myasthenia gravis (2/22). CONCLUSIONS: Anti-IFN-α2 autoantibodies were detected in viral infections, myasthenia gravis, and rarely in bacterial infections. ELISA and Luminex screening assays do not give similar results. Nonspecific reactivity and functional assays are necessary to validate the screening test result.

Laboratory workflow for optimizing leukocyte count and differential in synovial fluids on Sysmex XN-1000.

INTRODUCTION: Falsely elevated synovial white blood cell (WBC) counts using automated hematology analyzers have been reported particularly in the setting of joint arthroplasty. We evaluated the implementation of a laboratory workflow based on Sysmex XN-1000-automated cell counting and scattergram interpretation. METHODS: WBC and differential were measured for 76 synovial fluid samples (29 native joints and 47 with joint arthroplasties) with Sysmex XN-1000 and manual methods. All scattergrams were evaluated for possible incorrect WBC and/or differential according to our implemented workflow. A specific finding was the "banana-shape" scattergram, which indicates possible interferences. The European Bone & Joint Infection Society (EBJIS) criteria were applied to identify possible prosthetic joint infections (PJIs) in patients with joint arthroplasties. RESULTS: Correlation between automated and manual WBC counts, calculated for samples with WBC count <50 000/µL, was higher for native joints (r = 0.938) compared with patients known with arthroplasty (r = 0.906). Scattergrams classified as OK showed overall a higher correlation compared with scattergrams, which were interpreted as NOT OK. "Banana-shape" scattergrams (n = 19) showed falsely elevated automated WBC count, and the patterns were mainly seen in prosthesis patients (17/19 [89%]). Six of 47 (13%) patients with joint arthroplasties were reclassified from "confirmed" to "unlikely" PJI according to the EBJIS criteria. CONCLUSION: Our workflow based on scattergram interpretation resulted in accurate WBC counts in synovial fluid using automated/and or manual methods. It is important to identify the presence of "banana-shape" scattergrams to avoid overestimated automated WBC counts. Overall, automated synovial WBC count can be used, even for patients with arthroplasty, but after visual inspection of the scattergram to exclude possible interferences.

Endotyping of IgE-Mediated Polyethylene Glycol and/or Polysorbate 80 Allergy.

BACKGROUND: Polyethylene glycol (PEG) and polysorbate 80 (PS80) allergy preclude from SARS-CoV-2 vaccination. The mechanism(s) governing cross-reactivity and PEG molecular weight dependence remain unclear. OBJECTIVES: To evaluate PEGylated lipid nanoparticle (LNP) vaccine (BNT162b2) tolerance and explore the mechanism of reactivity in PEG and/or PS80 allergic patients. METHODS: PEG/PS80 dual- (n = 3), PEG mono- (n = 7), and PS80 mono-allergic patients (n = 2) were included. Tolerability of graded vaccine challenges was assessed. Basophil activation testing on whole blood (wb-BAT) or passively sensitized donor basophils (allo-BAT) was performed using PEG, PS80, BNT162b2, and PEGylated lipids (ALC-0159). Serum PEG-specific IgE was measured in patients (n = 10) and controls (n = 15). RESULTS: Graded BNT162b2 challenge in dual- and PEG mono-allergic patients (n = 3/group) was well tolerated and induced anti-spike IgG seroconversion. PS80 mono-allergic patients (n = 2/2) tolerated single-dose BNT162b2 vaccination. Wb-BAT reactivity to PEG-containing antigens was observed in dual- (n = 3/3) and PEG mono- (n = 2/3), but absent in PS80 mono-allergic patients (n = 0/2). BNT162b2 elicited the highest in vitro reactivity. BNT162b2 reactivity was IgE mediated, complement independent, and inhibited in allo-BAT by preincubation with short PEG motifs, or detergent-induced LNP degradation. PEG-specific IgE was only detectable in dual-allergic (n = 3/3) and PEG mono-allergic (n = 1/6) serum. CONCLUSION: PEG and PS80 cross-reactivity is determined by IgE recognizing short PEG motifs, whereas PS80 mono-allergy is PEG-independent. PS80 skin test positivity in PEG allergics was associated with a severe and persistent phenotype, higher serum PEG-specific IgE levels, and enhanced BAT reactivity. Spherical PEG exposure via LNP enhances BAT sensitivity through increased avidity. All PEG and/or PS80 excipient allergic patients can safely receive SARS-CoV-2 vaccines.

Potentials and pitfalls of ChatGPT and natural-language artificial intelligence models for the understanding of laboratory medicine test results. An assessment by the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) Working Group on Artificial Intelligence (WG-AI).

OBJECTIVES: ChatGPT, a tool based on natural language processing (NLP), is on everyone's mind, and several potential applications in healthcare have been already proposed. However, since the ability of this tool to interpret laboratory test results has not yet been tested, the EFLM Working group on Artificial Intelligence (WG-AI) has set itself the task of closing this gap with a systematic approach. METHODS: WG-AI members generated 10 simulated laboratory reports of common parameters, which were then passed to ChatGPT for interpretation, according to reference intervals (RI) and units, using an optimized prompt. The results were subsequently evaluated independently by all WG-AI members with respect to relevance, correctness, helpfulness and safety. RESULTS: ChatGPT recognized all laboratory tests, it could detect if they deviated from the RI and gave a test-by-test as well as an overall interpretation. The interpretations were rather superficial, not always correct, and, only in some cases, judged coherently. The magnitude of the deviation from the RI seldom plays a role in the interpretation of laboratory tests, and artificial intelligence (AI) did not make any meaningful suggestion regarding follow-up diagnostics or further procedures in general. CONCLUSIONS: ChatGPT in its current form, being not specifically trained on medical data or laboratory data in particular, may only be considered a tool capable of interpreting a laboratory report on a test-by-test basis at best, but not on the interpretation of an overall diagnostic picture. Future generations of similar AIs with medical ground truth training data might surely revolutionize current processes in healthcare, despite this implementation is not ready yet.

Case report: Myocarditis in congenital STAT1 gain-of function.

Autosomal dominant Signal transducer and activator of transcription 1 (STAT1) gain-of-function (GOF) mutations result in an inborn error of immunity characterized by chronic mucocutaneous candidiasis, recurrent viral and bacterial infections, and diverse autoimmune manifestations. Current treatment consists of chronic antifungal therapy, antibiotics for concomitant infections, and immunosuppressive therapy in case of autoimmune diseases. More recently, treatment with Janus kinases 1 and 2 (JAK1/2) inhibitors have shown promising yet variable results. We describe a STAT1 GOF patient with an incidental finding of elevated cardiac troponins, leading to a diagnosis of a longstanding, slowly progressive idiopathic myocarditis, attributed to STAT1 GOF. Treatment with a JAK-inhibitor (baricitinib) mitigated cardiac inflammation on MRI but was unable to alter fibrosis, possibly due to the diagnostic and therapeutic delay, which finally led to fatal arrhythmia. Our case illustrates that myocarditis could be part of the heterogeneous disease spectrum of STAT1 GOF. Given the insidious presentation in our case, a low threshold for cardiac evaluation in STAT1 GOF patients seems warranted.

Complementarity determining regions in SARS-CoV-2 hybrid immunity.

Pre-vaccination SARS-CoV-2 infection can boost protection elicited by COVID-19 vaccination and post-vaccination breakthrough SARS-CoV-2 infection can boost existing immunity conferred by COVID-19 vaccination. Such 'hybrid immunity' is effective against SARS-CoV-2 variants. In order to understand 'hybrid immunity' at the molecular level we studied the complementarity determining regions (CDR) of anti-RBD (receptor binding domain) antibodies isolated from individuals with 'hybrid immunity' as well as from 'naive' (not SARS-CoV-2 infected) vaccinated individuals. CDR analysis was done by liquid chromatography/mass spectrometry-mass spectrometry. Principal component analysis and partial least square differential analysis showed that COVID-19 vaccinated people share CDR profiles and that pre-vaccination SARS-CoV-2 infection or breakthrough infection further shape the CDR profile, with a CDR profile in hybrid immunity that clustered away from the CDR profile in vaccinated people without infection. Thus, our results show a CDR profile in hybrid immunity that is distinct from the vaccination-induced CDR profile.

Antinuclear antibodies in individuals with COVID-19 reflect underlying disease: Identification of new autoantibodies in systemic sclerosis (CDK9) and malignancy (RNF20, RCC1, TRIP13).

A high prevalence of antinuclear antibodies (ANA) in COVID-19 has been insinuated, but the nature of the target antigens is poorly understood. We studied ANA by indirect immunofluorescence in 229 individuals with COVID-19. The target antigens of high titer ANA (≥1:320) were determined by immunoprecipitation (IP) combined with liquid-chromatography-mass spectrometry (MS). High titer ANA (≥1:320) were found in 14 (6%) of the individuals with COVID-19. Of the 14 COVID-19 cases with high titer ANA, 6 had an underlying autoimmune disease and 5 a malignancy. IP-MS revealed known target antigens associated with autoimmune disease as well as novel autoantigens, including CDK9 (in systemic sclerosis) and RNF20, RCC1 and TRIP13 (in malignancy). The novel autoantigens were confirmed by IP-Western blotting. In conclusion, in depth analysis of the targets of high titer ANA revealed novel autoantigens in systemic sclerosis and in malignant disease.

Detection of macroenzymes: establishing upper reference limits for eight enzymes after polyethylene glycol precipitation.

Introduction: The presence of macroenzymes in blood can cause diagnostic confusion. Therefore, confirming the presence of macroenzymes is important to reduce unnecessary (non-)invasive investigations. Polyethylene glycol (PEG) precipitation is a simple and fast first-line method for the detection of macroenzymes. However, there is no consensus on the upper reference limit for the PEG-precipitable activity (%PPA) of monomeric enzymes. The aim of this study was to verify a PEG precipitation protocol for the detection of macroenzymes in our laboratory by establishing upper reference limits (URLs) and determining imprecision for eight enzymes after PEG precipitation. In addition, we aimed to clinically verify the URLs using samples containing macroenzymes as identified by electrophoresis. Materials and methods: Per enzyme, at least 40 leftover blood samples from adult patients with either normal or increased enzyme activities were diluted 1:1 with 25% PEG 6000 and 1:1 with 0.9% NaCl. Mixtures were incubated for 10 min at 37°C and centrifuged. Supernatant enzyme activity was measured on Cobas c702 and the %PPA was calculated. Results: The following URLs were obtained: 26% PPA for amylase, 29% PPA for alkaline phosphatase (ALP), 61% PPA for alanine aminotransferase, 48% PPA for aspartate aminotransferase, 24% PPA for creatine kinase (CK), 55% PPA for gamma-glutamyltransferase, 65% PPA for lactate dehydrogenase, and 56% PPA for lipase. The within-lab imprecision was < 15%. Regarding the clinical verification, the two historical samples with proven macroCK showed a %PPA of 69% and 43%, respectively, and a sample with proven macroALP had a %PPA of 52%. Conclusion: In this study, URLs for monomeric enzyme activities after PEG precipitation for eight different enzymes were established. The URLs are suitable for clinical use, but are only partially in line with other studies. Therefore, our data highlight the importance of establishing laboratory-specific upper reference limits for %PPA to allow a correct interpretation.

Laboratory biomarkers in the diagnosis and follow-up of treatment of allergic bronchopulmonary aspergillosis in cystic fibrosis.

Allergic bronchopulmonary aspergillosis (ABPA), a severe inflammatory respiratory disease, is caused by a hypersensitivity reaction to the colonization of the airways with Aspergillus fumigatus. It is most often described in patients with asthma or cystic fibrosis. The diagnosis of ABPA is based on a combination of clinical, radiological, and immunological findings that have been included in different diagnostic criteria over the years. In this paper, we review the biomarkers included in these diagnostic criteria and novel research biomarkers that may be used in the diagnosis and treatment follow-up of ABPA in cystic fibrosis.

Alpha-amylase as the culprit in an occupational mealworm allergy case.

Background: Occupational allergy has been described in employees working in contact with mealworms in pet stores, live fish bait or infested stored grains and recently, in mealworm farming for animal feed and human consumption. Mealworm allergens linked to occupational allergy are troponin C, cockroach-like allergen, tropomyosin, arginine kinase, early-staged encapsulation inducing- and larval cuticle proteins. Objective: We report a case of occupational mealworm allergy and studied the culprit component. Methods: Diagnosis was done by skin prick, specific IgE, basophil activation and lung function testing. Allergen purification was performed by anion-exchange chromatography and immunoblotting with patient IgE. Allergens were identified by in-gel trypsin digest and tandem mass spectrometry. Allergenicity and specificity further confirmed by IgE inhibition and passive basophil activation experiments. Results: We describe a new case of occupational mealworm allergy in a laboratory worker, with sensitization to different developmental stages and derivates of the mealworm. In basophil activation tests, the majority of patient's basophils (69%-91%) degranulated upon stimulation with the lowest concentration of mealworm extracts (0.16 µg/ml). Despite strong sensitization to mites, the patient did not show cross-reactivity to other insects. We were able to identify alpha-amylase as the main allergen and through inhibition experiments, we demonstrated that low amounts (0.1 µg/ml) of this allergen could strongly inhibit mealworm specific IgE by 79.1%. Moreover, passive BAT experiments demonstrated the IgE-alpha-amylase interaction to be functional, inducing up to 25.5% degranulation in healthy donor basophils. Conclusion: Alpha-amylase can be identified as the responsible allergen in this specific case of occupational mealworm allergy.

Acidification of 24-hour urine in urolithiasis risk testing: An obsolete relic?

BACKGROUND: Recommendations on the optimal preservation of 24 h urine for the metabolic work-up in urolithiasis patients are very heterogeneous. In case two such tests with different storage condition recommendations are being analysed, multiple collections would be needed, challenging especially elderly and very young patients. We therefore aimed to evaluate the stability of urine constituents under different storage conditions. MATERIAL AND METHODS: We collected urine samples from ten healthy volunteers and prepared aliquots to be stored either at room temperature or 4 °C. Some aliquots were preserved using hydrochloric acid prior to storage, some thereafter, some using the BD Urine preservation tube and some were not preserved at all. Storage duration was 0, 24, 48 or 72 h. In all samples calcium, magnesium, phosphorus, creatinine, oxalate, citrate and uric acid were measured and compared to the according reference sample. RESULTS: We could not find any significant deviation for any of the analytes and preanalytical treatment conditions compared to the associated reference sample. CONCLUSION: Preservation of 24 h urine for the metabolic evaluation in stone formers might not be necessary for sample storage up to 72 h.

Delayed diagnosis and treatment of extreme hypertriglyceridemia due to rejection of a lipemic sample.

INTRODUCTION: Most laboratories routinely determine haemolysis, icterus and lipemia indices to identify lipemic samples and reject potentially affected results. Hypertriglyceridemia is the most common cause of lipemia and severe hypertriglyceridemia (≥ 11.3 mmol/L) is a major risk factor of acute pancreatitis. LABORATORY ANALYSIS: A 56-year-old woman attended the outpatient clinic for a follow-up visit 1 month after a kidney transplantation. Her immunosuppressive therapy consisted of corticosteroids, cyclosporine, and mycophenolic acid. The routine clinical chemistry sample was rejected due to extreme lipemia. The comment "extreme lipemic sample" was added on the report, but the requesting physician could not be reached. The Cobas 8000 gave a technical error (absorption > 3.3) for the HIL-indices (L-index: 38.6 mmol/L) which persisted after high-speed centrifugation. The patient was given a new appointment 2 days later. The new sample was also grossly lipemic and gave the same technical error (L-index: 35.9 mmol/L). WHAT HAPPENED: The second sample was manually diluted 20-fold after centrifugation to obtain a result for triglycerides within the measuring range (0.10-50.0 mmol/L). Triglycerides were 169.1 mmol/L, corresponding to very severe hypertriglyceridemia. This result was communicated to the nephrologist and the patient immediately recalled to the hospital. She received therapeutic plasma exchange the next day and did not develop acute pancreatitis. MAIN LESSON: This case illustrates the delicate balance between avoiding the release of unreliable results due to lipemia and the risk of delayed diagnosis when results are rejected. Providing an estimate of the degree of hypertriglyceridemia might be preferable to rejecting the result.

How to meet ISO15189:2012 pre-analytical requirements in clinical laboratories? A consensus document by the EFLM WG-PRE.

The International Organization for Standardization (ISO) 15189:2012 standard aims to improve quality in medical laboratories through standardization of all key elements in the total testing process, including the pre-analytical phase. It is hence essential that accreditation bodies, assessing laboratories against ISO15189:2012, pay sufficient attention to auditing pre-analytical activities. However, there are significant differences in how technical auditors interpret the pre-analytical requirements described in ISO15189:2012. In this consensus document, the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) Working Group for Pre-analytical Phase (WG-PRE) sets out to review pre-analytical requirements contained in ISO15189:2012 and provide guidance for laboratories on how to meet these requirements. The target audience for this consensus document is laboratory professionals who wish to improve the quality of the pre-analytical phase in their laboratory. For each of the ISO requirements described in ISO15189:2012, members of EFLM WG-PRE agreed by consensus on minimal recommendations and best-in-class solutions. The minimal consensus recommendation was defined as the minimal specification which laboratories should implement in their quality management system to adequately address the pre-analytical requirement described in ISO15189:2012. The best-in-class solution describes the current state-of-the-art in fulfilling a particular pre-analytical requirement in ISO15189:2012. We fully acknowledge that not every laboratory has the means to implement these best-in-class solutions, but we hope to challenge laboratories in critically evaluating and improving their current procedures by providing this expanded guidance.

A case of severe pseudohyperkalaemia due to muscle contraction.

INTRODUCTION: Severe hyperkalaemia is a serious medical condition requiring immediate medical attention. Before medical treatment is started, pseudohyperkalaemia has to be ruled out. CASE DESCRIPTION: A 10-month old infant presented to the emergency department with fever and coughing since 1 week. Routine venous blood testing revealed a severe hyperkalaemia of 6.9 mmol/L without any indication of haemolysis. Reanalysis of the plasma sample confirmed the hyperkalaemia (7.1 mmol/L). Based on these results, the clinical pathologist suggested to perform a venous blood gas analysis and electrocardiogram (ECG) which revealed a normal potassium of 3.7 mmol/L and normal ECG, ruling out a potentially life-treating hyperkalaemia. The child was diagnosed with pneumonia. The paediatrician had difficulty to perform the first venous blood collection due to excessive movement of the infant during venipuncture. The muscle contractions of the child in combination with venous stasis most probably led to a local increase of potassium in the sampled limbs. The second sample collected under optimal preanalytical circumstances had a normal potassium. Since muscle contraction typically does not cause severe hyperkalaemia, other causes of pseudohyperkalaemia were excluded. K3-EDTA contamination and familial hyperkalaemia were ruled out and the patient did not have extreme leucocytosis or thrombocytosis. By exclusion a diagnosis of pseudohyperkalaemia due to intense muscle movement and venous stasis was made. CONCLUSION: This case suggests that intense muscle contraction and venous stasis can cause severe pseudohyperkalemia without hemolysis. Once true hyperkalemia has been ruled out, a laboratory work-up can help identify the cause of pseudohyperkalaemia.

Resolving DOAC interference on aPTT, PT, and lupus anticoagulant testing by the use of activated carbon.

BACKGROUND: Direct oral anticoagulants (DOACs) affect laboratory coagulations tests. Activated carbon (AC) can be used for adsorption of DOACs during acute human intoxications. OBJECTIVES: This study evaluates whether AC can also be used to resolve DOAC interference on in vitro clotting tests (prothrombin time [PT], activated partial thromboplastin time [aPTT], and lupus anticoagulant [LA] assays). PATIENTS/METHODS: Interference on PT, aPTT, Liquid anti-FXa, DTI, and LA screening/confirmation (SCT and dRVVT) was determined by spiking citrated plasma from 5 adult controls with 0, 20, 40, 80, 120, or 160 mg/mL AC. DOAC concentrations, PT, and aPTT were compared before and after AC addition to citrated plasma from patients receiving DOACs (n = 29), low molecular weight heparin (n = 10), and coumarin (n = 10) therapy. Samples from 69 LA screened patients were compared before and after AC addition. RESULTS: A concentration of 20 mg/mL AC had the lowest interference and was selected for further experiments. After AC addition, all DOAC concentrations were below the limit of quantification in the 29 treated patients, except for 2 apixaban samples. AC removed DOAC interference on PT and aPTT but had no impact on results obtained during coumarin or low molecular weight heparin therapy. Of 15 LA samples with interference resulting from DOAC therapy, 14 samples became negative and 1 positive after AC addition. Interference from coumarin therapy was not resolved. All 19 LA negative samples remained negative. AC treatment of the negative pooled plasma was required to avoid false-negative LA results in 21 known LA-positive samples. CONCLUSIONS: AC selectively removes DOAC interference on PT, aPTT, and LA assays.