Context.-Accurate measurement of serum creatinine is essential for estimating glomerular filtration rate (GFR), which is central to chronic kidney disease (CKD) detection, staging, and management. However, historical biases in creatinine measurement among several instrument/assay manufacturers compromised estimated GFR (eGFR) reliability, hindering CKD detection and management. Initiatives led by the National Kidney Disease Education Program (NKDEP) Laboratory Working Group and other stakeholders have standardized calibration of creatinine measurement and reporting of eGFR. Objective.-To summarize the journey toward global standardization of serum creatinine measurement procedures, detailing the coordinated efforts that led to calibration traceability to the NIST (National Institute of Standards and Technology) SRM (standard reference material) 967 reference material, examining the role of proficiency testing in monitoring progress, and contrasting the current state of creatinine assay standardization with that of other key kidney function biomarkers, such as cystatin C and urine albumin. Data sources.-A comprehensive review of peer-reviewed literature from 2005-2025 was conducted, including landmark studies from the Archives of Pathology & Laboratory Medicine, Clinical Chemistry, and clinical practice guidelines from several professional organizations. College of American Pathologists proficiency testing and LN24 survey data were included. Conclusions.-The global creatinine standardization initiative, driven by the NKDEP Laboratory Working Group, has dramatically reduced intermethod bias in serum creatinine measurement procedures, vastly improving the reliability of eGFR. While the metrologic traceability of calibration for creatinine measurement is now harmonized, challenges with analytical specificity, particularly for Jaffe-based methods, remain. The success of creatinine standardization serves as an important model for ongoing efforts to harmonize cystatin C and urine albumin measurements, ultimately leading to more accurate, equitable, and standardized care for patients with or at risk for CKD.

BACKGROUND:Multiplexed assays, which simultaneously measure many analytes from a single sample, have become increasingly significant for laboratory diagnosis. Many multiplexed assays with application to clinical diagnosis consist of analyte measurements that can fail individually. In these cases, traditional statistical quantitative quality control (QC) measures cannot be used without creating an unacceptably high false rejection rate. METHODS:We developed stochastic simulation software (qcsim) to calculate and visualize the detection power of complex QC rule combinations, including traditional Westgard rules as well as statistical tests of multiple QC repeats, with arbitrary degrees of multiplexing and levels of control. We used this approach to evaluate novel QC models that maintain stringent control of the bias and imprecision of each analyte in a highly multiplexed assay ("panel"). For classifier-based assays that use an algorithm to generate a small number of diagnostic outcomes from a large number of analytes (a "pattern"), we use perturbation analysis to assess the effect of different classifiers with a single analytical platform. RESULTS:Multiple QC approaches are able to overcome the challenge of highly multiplexed assays, and we demonstrate successful strategies that control the false rejection rate using either high analytical performance (low imprecision) or multiple QC replicates. We also demonstrate that, for pattern-based assays, algorithmic details of the specific classifier determine both critical analytes and the required stringency of the QC design. CONCLUSIONS:These results demonstrate multiple QC strategies that control highly multiplexed assays (1000-plex) at a level that is comparable to traditional QC schemes for single-analyte assays.

Background: Leptin, resistin, and adiponectin are critical adipokines involved in the pathophysiology of obesity and its related disorders, including type 2 diabetes. Although these biomarkers have historically been quantified using immunoassays, the specificity of antibody-based methods has frequently been questioned. As a result, there is an increasing interest in developing reliable, multiplexed clinical assays that utilize mass spectrometry for improved accuracy. In this study, we present a multiplexed immunoaffinity liquid chromatography-tandem mass spectrometry (multi-IA-LC-MS/MS) assay designed for the sensitive and selective measurement of leptin, resistin, and adiponectin in human plasma. Methods: Leptin, resistin, and adiponectin were selectively enriched from plasma samples using an antibody cocktail composed of monoclonal antibodies targeting each respective adipokine. The enriched adipokines underwent enzymatic digestion, and the resulting tryptic peptides were quantified using LC-MS/MS. The validated assay was subsequently applied to plasma samples collected from a cohort of subjects representing various weight categories, including normal weight, overweight, and obesity. Results: The lower limits of quantification for the assay were determined to be 0.5 ng/mL for both leptin and resistin, and 50 ng/mL for adiponectin. Intra- day, inter- day, and total imprecision measurements were all < 15 %, while spike recovery consistently exceeded 83 %. Comparative analysis with individual immunoassays demonstrated strong correlation, with all correlation coefficients (r) being equal to or greater than 0.869. Notably, when comparing subjects with obesity to those with normal weight, there was an approximately ninefold increase in circulating leptin levels and a similar to 1.6-fold decrease in circulating adiponectin levels. Conclusions: A multi-IA-LC-MS/MS assay was developed for the simultaneous and sensitive measurement of leptin, resistin, and adiponectin in clinical samples. This quantitative method shows significant potential for applications related to obesity and could facilitate improved clinical management and understanding of obesity- related conditions.

BACKGROUND:Multianalyte machine learning (ML) models can potentially identify previously undetectable wrong blood in tube (WBIT) errors, improving upon current single-analyte delta check methodology. However, WBIT detection model performance has not been assessed in a real-world, low-prevalence context. To estimate real-world positive predictive values, we propose a methodology to assess WBIT detection models by evaluating the impact of missing data and by using a "low prevalence" validation data set. METHODS:We trained a range of model specifications using various predictors in a pediatric setting. We assessed the top-performing model on a modified, "low prevalence" validation data set across a range of probability thresholds. Model performance was also compared to a pre-positive patient identification (pre-PPID) dataset. RESULTS:An Extreme Gradient Boosting (XGBoost) model with minimal preprocessing performed the best for both complete blood count with differential white cell count (CBC with Diff) tests (accuracy 0.9715) and complete blood count without differential white cell count (CBC without Diff) tests (accuracy 0.9647). Assessment on a downsampled, "low prevalence" validation data set resulted in estimated positive predictive values ranging from 0.01 to 0.67 (CBC with Diff) and 0.01 to 0.75 (CBC without Diff), depending on the probability threshold chosen. A comparison of prospective performance to PPID data demonstrated a large decrease in estimated WBIT errors. CONCLUSIONS:We find that ML models can accurately predict WBITs in a primarily pediatric setting. Evaluating model performance across a range of probability thresholds minimizes the number of false positives while still providing added safety benefits. The decrease in estimated WBITS post-PPID implementation shows the potential safety benefits of a WBIT model for hospitals not using PPID when collecting laboratory specimens.

[This corrects the article DOI: 10.1016/j.jmsacl.2024.01.004.].

Objectives: Ketone bodies (KBs) serve as important energy sources that spare glucose, providing the primary energy for cardiac muscle, skeletal muscle during aerobic exercise, and the brain during periods of catabolism. The levels and relationships between the KBs are critical indicators of metabolic health and disease. However, challenges in separating isomeric KBs and concerns about sample stability have previously limited their clinical measurement. Methods: A novel 6.5-minute liquid chromatography-mass spectrometry-based assay was developed, enabling the precise measurement of alpha-, betaand gamma-hydroxybutyrate, beta-hydroxyisobutyrate, and acetoacetate. This method was fully validated for human serum and plasma samples by investigating extraction efficiency, matrix effects, accuracy, recovery, intra- and inter-precision, linearity, lower limit of quantitation (LLOQ), carryover, specificity, stability, and more. From 107 normal samples, reference ranges were established for all analytes and the beta-hydroxybutyrate/acetoacetate ratio. Results: All five analytes were adequately separated chromatographically. An extraction efficiency between 80 and 120 % was observed for all KBs. Accuracy was evaluated through spike and recovery using 10 random patient samples, with an average recovery of 85-115 % for all KBs and a coefficient of variation of <= 3 %. Coefficients of variation for intra- and inter-day imprecision were < 5 %, and the total imprecision was < 10 %. No significant interferences were observed. Specimens remained stable for up to 6 h on ice or 2 h at room temperature. Conclusions: The developed method is highly sensitive and robust. It has been validated for use with human serum and plasma, overcoming stability concerns and providing a reliable and efficient quantitative estimation of ketone bodies.

Journal Article The Case for Including Data and Code with ML Publications in Laboratory Medicine Get access Stephen R Master Stephen R Master Department of Pathology and Laboratory Medicine, Children’s Hospital of Philadelphia, Philadelphia, PA, USAPerelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA Address correspondence to this author at: 34th and Civic Center Blvd., Room 5144, Philadelphia, PA 19104-4399, USA; E-mail masters@chop.edu. https://orcid.org/0000-0002-6808-631X Search for other works by this author on: Oxford Academic Google Scholar The Journal of Applied Laboratory Medicine, Volume 8, Issue 1, January 2023, Pages 213–216, https://doi.org/10.1093/jalm/jfac088 Published: 04 January 2023 Article history Received: 11 July 2022 Accepted: 21 September 2022 Published: 04 January 2023

BACKGROUND & AIMS: Altered plasma acylcarnitine levels are well-known biomarkers for a variety of mitochondrial fatty acid oxidation disorders and can be used as an alternative energy source for the intestinal epithelium when short-chain fatty acids are low. These membrane-permeable fatty acid in-termediates are excreted into the gut lumen via bile and are increased in the feces of patients with inflammatory bowel disease (IBD). METHODS: Herein, based on studies in human subjects, ani-mal models, and bacterial cultures, we show a strong positive correlation between fecal carnitine and acylcarnitines and the abundance of Enterobacteriaceae in IBD where they can be consumed by bacteria both in vitro and in vivo. RESULTS: Carnitine metabolism promotes the growth of Escherichia coli via anaerobic respiration dependent on the cai operon, and acetylcarnitine dietary supplementation increases fecal carnitine levels with enhanced intestinal colonization of the enteric pathogen Citrobacter rodentium. CONCLUSIONS: In total, these results indicate that the increased luminal concentrations of carnitine and acylcarni-tines in patients with IBD may promote the expansion of pathobionts belonging to the Enterobacteriaceae family, thereby contributing to disease pathogenesis.

BACKGROUNDMitochondria are cytosolic organelles within most eukaryotic cells. Mitochondria generate the majority of cellular energy in the form of adenosine triphosphate (ATP) through oxidative phosphorylation (OxPhos). Pathogenic variants in mitochondrial DNA (mtDNA) and nuclear DNA (nDNA) lead to defects in OxPhos and physiological malfunctions (Nat Rev Dis Primer 2016;2:16080.). Patients with primary mitochondrial disorders (PMD) experience heterogeneous symptoms, typically in multiple organ systems, depending on the tissues affected by mitochondrial dysfunction. Because of this heterogeneity, clinical diagnosis is challenging (Annu Rev Genomics Hum Genet 2017;18:257-75.). Laboratory diagnosis of mitochondrial disease depends on a multipronged analysis that can include biochemical, histopathologic, and genetic testing. Each of these modalities has complementary strengths and limitations in diagnostic utility.CONTENTThe primary focus of this review is on diagnosis and testing strategies for primary mitochondrial diseases. We review tissue samples utilized for testing, metabolic signatures, histologic findings, and molecular testing approaches. We conclude with future perspectives on mitochondrial testing.SUMMARYThis review offers an overview of the current biochemical, histologic, and genetic approaches available for mitochondrial testing. For each we review their diagnostic utility including complementary strengths and weaknesses. We identify gaps in current testing and possible future avenues for test development.

Background Machine learning (ML) has been applied to an increasing number of predictive problems in laboratory medicine, and published work to date suggests that it has tremendous potential for clinical applications. However, a number of groups have noted the potential pitfalls associated with this work, particularly if certain details of the development and validation pipelines are not carefully controlled. Methods To address these pitfalls and other specific challenges when applying machine learning in a laboratory medicine setting, a working group of the International Federation for Clinical Chemistry and Laboratory Medicine was convened to provide a guidance document for this domain. Results This manuscript represents consensus recommendations for best practices from that committee, with the goal of improving the quality of developed and published ML models designed for use in clinical laboratories. Conclusions The committee believes that implementation of these best practices will improve the quality and reproducibility of machine learning utilized in laboratory medicine. Summary We have provided our consensus assessment of a number of important practices that are required to ensure that valid, reproducible machine learning (ML) models can be applied to address operational and diagnostic questions in the clinical laboratory. These practices span all phases of model development, from problem formulation through predictive implementation. Although it is not possible to exhaustively discuss every potential pitfall in ML workflows, we believe that our current guidelines capture best practices for avoiding the most common and potentially dangerous errors in this important emerging field.

Journal Article Low Plasma Citrulline Guiding the Diagnosis of a Mitochondrial Disorder Get access Parith Wongkittichote, Parith Wongkittichote Department of Pathology and Laboratory Medicine, Children’s Hospital of Philadelphia, Philadelphia, PA, United StatesDivision of Human Genetics, Children’s Hospital of Philadelphia, Philadelphia, PA, United States https://orcid.org/0000-0002-7321-307X Search for other works by this author on: Oxford Academic Google Scholar Rebecca D Ganetzky, Rebecca D Ganetzky Department of Pathology and Laboratory Medicine, Children’s Hospital of Philadelphia, Philadelphia, PA, United StatesDivision of Human Genetics, Children’s Hospital of Philadelphia, Philadelphia, PA, United States https://orcid.org/0000-0001-6238-8109 Search for other works by this author on: Oxford Academic Google Scholar Matthew M Demczko, Matthew M Demczko Division of Human Genetics, Children’s Hospital of Philadelphia, Philadelphia, PA, United States Search for other works by this author on: Oxford Academic Google Scholar Xinying Hong, Xinying Hong Department of Pathology and Laboratory Medicine, Children’s Hospital of Philadelphia, Philadelphia, PA, United States https://orcid.org/0000-0002-7517-5193 Search for other works by this author on: Oxford Academic Google Scholar Miao He, Miao He Department of Pathology and Laboratory Medicine, Children’s Hospital of Philadelphia, Philadelphia, PA, United States Search for other works by this author on: Oxford Academic Google Scholar Stephen R Master Stephen R Master Department of Pathology and Laboratory Medicine, Children’s Hospital of Philadelphia, Philadelphia, PA, United States Address correspondence to this author at: Department of Pathology and Laboratory Medicine, Children’s Hospital of Philadelphia, 34th and Civic Center Blvd., Room 5144, Philadelphia, PA 19104, United States. Tel 215-426-7280; e-mail masters@chop.edu. https://orcid.org/0000-0002-6808-631X Search for other works by this author on: Oxford Academic Google Scholar Clinical Chemistry, Volume 69, Issue 6, June 2023, Pages 661–664, https://doi.org/10.1093/clinchem/hvad039 Published: 01 June 2023 Article history Received: 18 January 2023 Accepted: 09 March 2023 Published: 01 June 2023

Journal Article Do You See What I See? Automated IFE Interpretation Using Machine Learning Get access Stephen R Master, Stephen R Master Department of Pathology and Laboratory Medicine, Childrens Hospital of Philadelphia, Philadelphia, PA, United StatesPerelman School of Medicine, University of Pennsylvania, Philadelphia, PA, United States Address correspondence to this author at: Department of Pathology and Laboratory Medicine, Childrens Hospital of Philadelphia, 34th and Civic Center Blvd., Room 5144, Philadelphia, PA 19104-4399. E-mail: masters@chop.edu. https://orcid.org/0000-0002-6808-631X Search for other works by this author on: Oxford Academic Google Scholar Shannon Haymond Shannon Haymond Ann Robert H. Lurie Childrens Hospital of Chicago, Chicago, IL, United StatesDepartment of Pathology, Northwestern University Feinberg School of Medicine, Chicago, IL, United States Search for other works by this author on: Oxford Academic Google Scholar Clinical Chemistry, Volume 69, Issue 2, February 2023, Pages 113–115, https://doi.org/10.1093/clinchem/hvac202 Published: 28 December 2022 Article history Received: 18 November 2022 Accepted: 21 November 2022 Published: 28 December 2022

GA1 (OMIM# 231670) is an organic aciduria caused by defective Glutaryl-CoA dehydrogenase (GCDH), encoded by GCDH. Early detection of GA1 is crucial to prevent patients from developing acute encephalopathic crisis and subsequent neurologic sequelae. Diagnosis of GA 1 relies on elevated glutarylcarnitine (C5DC) in plasma acylcarnitine analysis and hyperexcretion of glutaric acid (GA) and 3-hydroxyglutaric acid (3HG) in urine organic acid (UOA) analysis. Low excretors (LE), however, exhibit subtly elevated or even normal plasma C5DC and uri-nary GA levels, leading to screening and diagnostic challenges. The measurement of 3HG in UOA is thus often used as the 1st tier test for GA1. We described a case of LE detected via newborn screen with normal excretion of GA, absent of 3HG and increased 2-methylglutaconic acid (2MGA), which was detected at 3 mg/g creatinine (reference interval <1 mg/g creatinine) without appreciable ketones. We retrospectively examined UOA of 8 other GA1 patients and the 2MGA level ranged from 2.5 to 27.39 mg/g creatinine, which is significantly higher than normal controls (0.05-1.61 mg/g creatinine). Although the underlying mechanism of 2MGA formation in GA1 is unclear, our study suggests 2MGA is a biomarker for GA1 and should be monitored by routine UOA to evaluate its diagnostic and prognostic value.(c) 2023 Elsevier Inc. All rights reserved.

INTRODUCTION: The antiphospholipid syndrome (APS), initially described nearly 4 decades ago, remains enigmatic with unclear pathophysiology. In addition to the lack of a mechanistic diagnostic test, APS exhibits considerable phenotypic heterogeneity, including the occurrence of “non-consensus” clinical manifestations (i.e. other than vascular thrombosis and pregnancy complications). Despite the evidence that phospholipid-binding factors are highly likely to be central to APS pathogenesis, the full complement of phospholipid-binding proteins has yet to be systemically assayed. Proteomic profiling by label-free quantitation (LFQ) may identify patterns of protein binding to phospholipid that are central to APS pathogenesis and thereby elucidate a basis for the heterogeneous clinical presentations of APS. METHODS: C18 silica beads were coated with phospholipid by incubation with a suspension of 30% phosphatidylserine/70% phosphatidylcholine. Plasmas from 4 groups: 1) APS patients (n=11), 2) APL positive patients who lacked clinical manifestations (“APL”) (n=11), 3) non-APL patients with thrombosis (“nonAPL”)(n=11), and 4) normal healthy controls (“normal”) (n=6) were incubated with the phospholipid-coated beads, and eluates were processed by gel electrophoresis followed by tryptic digestion. The resulting peptides were analyzed by LC-MS/MS on a Thermo Orbitrap Fusion Lumos. Protein identification, false discovery rate (FDR) estimation, and LFQ were performed using Proteome Discoverer 2.2 (Thermo). LFQ values were normalized by overall peptide abundance, and relative quantitation was obtained. Groupwise comparisons were performed using PD 2.2 in order to identify differentially enriched proteins. RESULTS: A total of 1,210 proteins were identified, with 875 proteins identified at high confidence (q<0.01). Based on gel electrophoresis, all samples had comparable amounts of total protein, and normalization of LFQ data on the basis of total proteins was therefore performed. Differentially expressed proteins between groups were identified using both fold-change and -log p value cutoffs (Figure 1a). Primary analysis focused on identifying proteins that were consistently preferentially bound from the APS plasmas compared to the normal controls. Novel findings included the following proteins related to coagulation, immune response, and to lipid binding that bound preferentially from APS plasmas compared to normal plasmas (APS:control ratios shown in parentheses): coagulation factor VII (10.4), heparin cofactor 2 (9.6), CD14 (6.1), apolipoprotein L1 (6.0), and apolipoprotein L2 (5.5). Additional proteins that preferentially bound from APS plasmas included OAF (13.7), BRSK1 (13.2), MINPP1 (12.5), fetuin B (10.5), SHBG (8.9), ADAMTS-like protein 4 (8.5), SERPIN A4 (7.6), alpha-2-HS glycoprotein (6.7) and inter-alpha-trypsin inhibitor heavy chains (H3, H4 and H1; 8.5, 6.6, 6.5 respectively). Consistent with prior knowledge of APS, and validating this LFQ approach immunoglobulins and complement proteins were also among the preferentially bound proteins and included: Ig heavy chain 1-3 (15.6), complement C9 (13.3), complement C4-B, (7.1), complement component C8 beta chain (5.6), Ig kappa variable 1D-8 (5.6), and Ig lambda-like polypeptide 1 (3.9). Interestingly, a number of identified proteins showed intermediate abundance (between APS and control) in the APL positive patients without clinical manifestations (“APL”) and APL negative patients with thrombosis (“nonAPL”) (see figure 1B-D for examples), suggesting a potential dose response effect. CONCLUSIONS: 1) These results are a first demonstration of concept for the ability of LFQ proteomics to address the complexity of APS through analysis of phospholipid-bound proteins. 2) A number of phospholipid-binding proteins, many of which have not been previously implicated in APS pathogenesis, are preferentially present in APS patient plasma 3) Some plasmas from APL and non-APL patients also show increased binding of some of these proteins compared to the normal controls, suggesting that LFQ may identify markers for subgroups of patients within these 2 categories.

Abstract Background Inherited mitochondrial diseases, such as Leigh Syndrome cause defects in the energy metabolism, resulting in abnormalities in the mitochondrial redox state, defined as the ratio between nicotinamide adenine dinucleotide (NADH) and oxidized nicotinamide adenine dinucleotide (NAD+). Accurate and reproducible measurement of the NADH/NAD+ ratio directly is challenging due to its instability. The hepatic intramitochondrial NADH/NAD+ ratio is directly proportional to the ratio between the ketone bodies: beta-hydroxybutyrate (BHB) and acetoacetate (AcAc). Therefore, measuring the ketone body ratio is a more stable approach to estimate the mitochondrial redox state. Due to AcAc stability challenges and the presence of BHB structural isomers (i.e., alpha-hydroxybutyrate (AHB), gamma-hydroxybutyrate (GHB), and beta-hydroxyisobutyrate (BHIB)), no clinically validated assay that measures BHB, AcAc, and their ratio was available as a single test. Existing assays quantify AcAc and BHB separately using spectrophotometry, enzymology, or indirectly via gas chromatography, resulting in significantly reduced accuracy and large specimen quantity requirements. Lack of integration is particularly challenging for calculating a reproducible ratio. To improve clinical care, a novel liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS) based ketone body panel requiring only 10 µL of serum or plasma was developed and clinically validated. Methods 10 µL of serum or plasma was taken and kept on ice, followed by a protein precipitation, and drying down of the acquired supernatant. Thereafter, all five analytes were separated using a 6.5 min reversed phased LC column (Acquity, Waters) and detected by MS/MS analysis (Xevo TQ-S, Waters). Data were processed and curated with Ascent V4 (Indigo). Patient samples from the mitochondrial clinic (Children’s Hospital of Philadelphia) were analyzed and results were paired with known diagnoses. Results A new robust and sensitive ketone body panel was developed that not only accurately and reproducibly measures concentrations of BHB, AcAc, and their ratio, but also the BHB isomers AHB, GHB, and BHIB. To ensure quality, a thorough validation and stability study has been performed. All analytes reported a linear range over three orders of magnitude, i.e., 0.0025–1.5 mM (AcAc and AHB), 0.0050–1.5 mM (BHB and BHIB), and 0.0025–1.2 mM (GHB), and a dilution up to 50× is permitted. The precision was assayed over 20× days, 2× replicates, and 2× injections for each analyte at three levels determining the inter-assay (day), intra-assay (replicate), and intra-assay (injection) coefficient of variation (CV); all the %CVs were reported below 6.5%. Spike and recovery were performed to determine the accuracy, results were obtained from ten random patient samples; an average of 99.9% (AcAc), 102.7% (BHB), 95.3% (AHB), 85.7% (GHB), and 87.5% (BHIB) was measured. Additionally, at least 40 samples were cross-examined for BHB using two independent assays, a mean bias of 0.01 mM and a Pearson’s R = 0.996 were reported. Specificity was determined for all analytes against 16 common therapeutic drugs, bilirubin, hemoglobin, triglycerides, and high protein levels at <15% bias. Finally, new reference ranges have been established and were used to calculate accuracy rates. Conclusion Improved mitochondrial disease screening and diagnostics is established utilizing the LC-MS/MS based ketone body panel.

Recent studies demonstrating problems with COVID (1) and sepsis (2) prediction models have helped raise awareness (3) about the need for better practices in developing and reporting machine learning (ML; or artificial intelligence) methods in healthcare. This so-called reproducibility (or replication) crisis has been recognized and extends beyond clinical applications (4). In fact, this issue is not specific to ML. Many scientific journals, including Clinical Chemistry, have adopted principles (specified in the article submission guidelines) to facilitate reproducibility, rigor, and transparency in published findings. Although there are common elements that support transparency and rigor in science, each technology has its own set of pitfalls that must be addressed. This is true for the rapidly developing field of ML. The growth and development of open-source software and digitized and publicly available data sources have made the application of ML methods highly accessible. While this has facilitated a surge in interest and publications, it has reduced the requirement for developers to have the necessary foundational or subject matter knowledge needed for quality publications and innovations. When ML methods are published in clinical journals, peer reviewers and editors may lack the expertise to appropriately evaluate technical aspects of submissions. Thus, we need best practices that help educate clinical experts and govern how ML for laboratory medicine should be developed and communicated. This is critical for ensuring practicality and reproducibility of ML applications and, ultimately, their successful translation to clinical practice.

BackgroundA 4-week-old Caucasian male presented with failure to thrive, grunting, irritability, and feeding difficulties.The pregnancy was complicated by intrauterine growth restriction and oligohydramnios; fetal echocardiogram was normal.He was born at term, passed his hearing test, and his newborn screening for 37 disorders was unremarkable.The family history was significant for an unexplained neonatal death of a male maternal first cousin once removed.A chest radiograph on presentation showed cardiomegaly.An echocardiogram demonstrated moderate dilation and hypertrophy of the left ventricle with a severely reduced ejection fraction (13%; normal 55%).Laboratory testing showed increased B-type natriuretic peptide (BNP) [1868.2pg/ml; reference interval (RI): 0.0-100.0pg/ml] and blood lactic acid (3.8 mmol/L; RI: 0.5-2.0mmol/L).Complete blood cell counts revealed borderline low absolute neutrophil count at 1160/mL (RI: 1108-5450/mL).Results of liver and renal function tests, plasma amino acids analysis (PAA), plasma free and total carnitine, and plasma acylcarnitine profile were normal.Urine organic acids analysis (UOA) showed borderline increases in 3methylglutaconic acid (3MGA) (25 mmol/mol creatinine; normal <22 mmol/mol creatinine) and 3-methylglutaric acid (3MG) (3.3 mmol/mol creatinine; normal