Rationale & Objective: Arginine vasopressin (AVP) is an established driver of cyst growth in autosomal dominant polycystic kidney disease (ADPKD). Urine osmolality (osm) measures are surrogate markers of AVP activity. Both 24-hour and spot urine samples are used as indicators of AVP suppression. The agreement between these 2 measurements remains unclear. Study Design: A retrospective cohort study. Setting & Study Population: Three hundred and forty-nine patients with ADPKD with 839 urine samples from a tertiary care center. Selection Criteria for Study: Patients with ADPKD with records of spot and 24-hour urine measurements. Data Extraction: Consecutive patients' data from January 2018 to March 2023 were extracted from the quality assurance database of The Ottawa Hospital Cystic Kidney Disease Clinic. Analytical Approach: Discordance assessed at target urine osmolality of 250 and 270mmol/kg. Agreement assessed by Bland-Altman plots. The percentage of patients with difference in osmolality between the 2 measures for cutoff points of>50,>100,>150, and>200mmol/kg was calculated. Results: The mean 24-hour urine osm was 364mmol/kg, and the mean spot urine osm was 424 mosm/kg. Mean age of 46 years, 52% females, and 47 (13.5%) were on tolvaptan. Overall, in comparing spot urine osm to 24-hour urine osm, the discordance at 250 and 270mmol/kg was 24% with poor agreement on Bland-Altman plots. The differences between the 2 measures at varying cutoff points were 53.9% at 50mmol/kg, 35.8% at 100mmol/kg, 24.1% at 150mmol/kg, and 16.1% at 200mmol/kg. Results were similar when only a single measurement from each patient was used for analysis. Limitations: Total of 29% of patients did not have concurrent spot urine osmolality and 24-hour urine osmolality. The study was conducted at a single center. Limited number of patients were on tolvaptan. Conclusions: In adults with ADPKD, important differences exist between the 24-hour urine osmolality and spot urine osmolality that preclude interchangeable use. The method employed may impact clinical decision-making. More research is needed to determine, which urine osm should be used when assessing AVP suppression.

BACKGROUND:Clinicians caring for kidney transplant recipients (KTRs) most commonly use estimated glomerular filtration rate (eGFR) to guide medication dosing as it is the most readily available measure of kidney function. Which eGFR equations provide the most accurate medication dosing guidance for KTRs remains uncertain. METHODS:We studied 415 stable KTRs in Canada and New Zealand. Participants completed same-day measurements of creatinine and cystatin C and measured GFR (diethylenetriaminepentaacetic acid). Chronic Kidney Disease Epidemiology Collaboration, European Kidney Function Consortium, and transplant-specific eGFR equations were compared with both Cockcroft-Gault creatinine clearance (CrCl) and measured GFR. eGFR equations were assessed both indexed to a standardized body surface area (BSA) of 1.73 m 2 (milliliter per minute per 1.73 m 2 , as is conventional reporting from most clinical laboratories) and nonindexed (milliliter per minute) accounting for actual BSA. The primary outcome was the proportion of medication dosing discordance relative to Cockcroft-Gault CrCl or measured GFR for 8 commonly prescribed medications. Stratified analyses were performed on the basis of obesity status. RESULTS:Nonindexed eGFR equations (milliliter per minute) resulted in substantially lower medication dosing discordance compared with indexed eGFR equations (milliliter per minute per 1.73 m 2 ). These findings were most pronounced among KTRs with obesity, in whom underdosing was frequent. When compared with Cockcroft-Gault CrCl, the lowest proportion of discordance was found with the nonindexed 2023 transplant-specific equation. When compared with measured GFR, the lowest proportion of discordance was found with the nonindexed 2021 Chronic Kidney Disease Epidemiology Collaboration Cr/CysC equation. CONCLUSIONS:Nonindexed eGFR values accounting for actual BSA should be used by clinicians for medication dosing in KTRs. These findings may inform KT providers about which eGFR equations provide the safest, most accurate medication dosing guidance for KTRs.

Introduction: Emerging data shows that early therapeutic intervention in patients with high risk smoldering multiple myeloma (SMM) may improve outcomes. While treatment of high-risk SMM is not standard of care in Canada, understanding the epidemiology of SMM in our region is important to contextualize the future healthcare resource implications of SMM treatment. Therefore, this study's primary aim was to assess the incidence and prevalence of clinically- detected SMM among the general population over time, diagnosed as part of routine standard of care. Though a recent population-level screening study has shown that among screened patients, the prevalence of SMM in the general population over 40 years old is 0.53% (Thorsteindottir et al. BCJ 2023), we hypothesize that the prevalence of clinically-detected SMM population will be much lower. Methods: This retrospective observational study included all patients in the Champlain Local Health Integration Network (LHIN) found to have an abnormal monoclonal protein (MCP) (an abnormal serum or urine protein electrophoresis or immunofixation, or free light chain [FLC] ratio) between January 1, 2010, and December 31, 2022. Charts were reviewed to identify patients initially diagnosed with or progressing to SMM. Healthcare services in Ontario, Canada are provided through regional LHINs. The Ottawa Hospital is the only tertiary care and hematology referral center in the Champlain LHIN, therefore, we assumed that all incident cases of SMM were evaluated at our center and included in this study. To provide population-based incidence and prevalence rates of SMM, the adult population of individuals living within the Champlain LHIN was taken from the published 2016 and 2021 census data. The incidence of SMM was defined as the date a patient met clinical criteria for SMM (bone marrow [BM] plasma cells 10-59% in the absence of myeloma defining events [MDE], or MCP ≥30 g/L and no MDE if no bone marrow performed). The prevalence of SMM was defined as the number of patients actively followed for SMM at our institution during the relevant time-period. The date of last follow up was July 1, 2024. Results: A total of 5948 people had an abnormal MCP between 2010-2022, of whom 229 (3.9%) were diagnosed with SMM. Of the 229 SMM patients, the initial diagnosis was MGUS in 74 (32%) and SMM in 155 (68%) patients. Among the 229 SMM patients, 9 (4%) met criteria for high risk SMM based on the Mayo 20/20/20 score (Lakshman et al. BCJ 2018). However, given the initial lack of availability of the FLC test during the entire study period, FLC data was missing in 82 (36%) of patients. Among SMM patients, the median MCP at the time of first evaluation was 17.7 (IQR 10.3-29.5) g/L if diagnosed between 2010-2014, 13.5 (IQR 6.7-21.9) if diagnosed between 2015-2019, and 9.3 (IQR 3.5-14.4) g/L if evaluated between 2020-2022. A baseline BM evaluation was performed among 79% (n=46) and 94% (n=157) of patients diagnosed with SMM between 2010-2014 (n=62) and 2015-2022 (n=167), respectively. Overall, 108 (47%) of SMM patients were alive and had been clinically evaluated since July 1, 2023. The incidence of SMM among per 100,000 people in the general population (n=1,230,655 in 2011; n=1,292,639 in 2016; n=1,394,070 in 2021) increased from 0.5 cases in 2011, to 0.9 cases in 2016, to 1.4 cases in 2021. If restricting to only people aged ≥40, the incidence of SMM per 100,000 people in the general population (n=623,370 in 2011, n=670,195 in 2016, and n=719,920 in 2021) was 1 case in 2011, 1.8 cases in 2016 and 2.8 cases in 2021. The prevalence of SMM per 100,000 people in the general population was 6.9 cases in 2011 (n=21), 7.7 (n=58) in 2016, and 11.3 (n=114) in 2021. The incidence of SMM among patients with an abnormal MCP increased from 1.6% (n=6 SMM; n=378 abnormal MCP) in 2011, 2.8% (n=12 SMM; n=12 abnormal MCP) in 2016, to 4.1% (n=20 SMM; n=486 abnormal MCP) in 2021. Conclusion: This large study shows that the prevalence of clinically-detected SMM among adults above age ≥40 is lower than reported in a universally screened patient population. While the incidence of SMM among patients with an abnormal MCP and among the general population has increased over time, the abnormal MCP size at diagnosis has decreased over time. This suggests that clinicians are more thoroughly evaluating patients for SMM even with smaller MCPs. Larger cohort studies are needed to evaluate if the incidence of clinically detected high-risk SMM has also increased over time.

When we hear the phrase ‘data-driven laboratory stewardship’, we often think of this as referring to using data on test use (e.g., test volumes or costs) to highlight opportunities for improvement. While these data are undoubtedly essential, there are many other potential data sources that can inform laboratory stewardship initiatives. Such data sources can be identified by drawing on key lessons from the rapidly developing field of implementation science (the scientific study of methods to facilitate the uptake of best practices into routine, everyday healthcare). Here, we introduce this field and outline some of its key lessons about relevant data to support (I) developing an understanding of the factors influencing over- or under-use of tests and enablers of change/barriers impeding change; and (II) selection of improvement strategies that are most suited to disrupting current patterns of test use and capitalizing on enablers of change/breaking down barriers impeding change. We also provide suggestions for how laboratory stewardship teams can put these lessons into practice as part of stewardship initiative development, and tools that can support these activities. The key lessons are couched within an over-arching framework that can be used to guide the development, implementation, and evaluation of laboratory stewardship initiatives as part of continuous improvement activities embedded within a learning health system.

BACKGROUND:The conventional single-analyte delta check, utilized for identifying intravenous fluid contamination and other preanalytical errors, is known to flag many specimens reflecting true patient status changes. This study aimed to derive delta check rules that more accurately identify contamination. METHODS:Results for calcium, creatinine, glucose, sodium, and potassium were retrieved from 326 103 basic or comprehensive metabolic panels tested between February 2021 and January 2022. In total, 7934 specimens showed substantial result changes, of which 1489 were labeled as either contaminated or non-contaminated based on chart review. These labeled specimens were used to derive logistic regression models and to select the most predictive single-analyte delta checks for 4 common contaminants. Their collective performance was evaluated using a test data set from October 2023 comprising 14 717 specimens. RESULTS:The most predictive single-analyte delta checks included a calcium change by ≤-24% for both saline and Plasma-Lyte A contamination, a potassium increase by ≥3.0 mmol/L for potassium contamination, and a glucose increase by ≥400 mg/dL (22.2 mmol/L) for dextrose contamination. In the training data sets, multi-analyte logistic regression models performed better than single-analyte delta checks. In the test data set, logistic regression models and single-analyte delta checks demonstrated collective alert rates of 0.58% (95% CI, 0.46%-0.71%) and 0.60% (95% CI, 0.49%-0.74%), respectively, along with collective positive predictive values of 79% (95% CI, 70%-89%) and 77% (95% CI, 68%-87%). CONCLUSIONS:Single-analyte delta checks selected by logistic regression demonstrated a low false alert rate.

BACKGROUND:Machine learning solutions offer tremendous promise for improving clinical and laboratory operations in pathology. Proof-of-concept descriptions of these approaches have become commonplace in laboratory medicine literature, but only a scant few of these have been implemented within clinical laboratories, owing to the often substantial barriers in validating, implementing, and monitoring these applications in practice. This mini-review aims to highlight the key considerations in each of these steps. CONTENT:Effective and responsible applications of machine learning in clinical laboratories require robust validation prior to implementation. A comprehensive validation study involves a critical evaluation of study design, data engineering and interoperability, target label definition, metric selection, generalizability and applicability assessment, algorithmic fairness, and explainability. While the main text highlights these concepts in broad strokes, a supplementary code walk-through is also provided to facilitate a more practical understanding of these topics using a real-world classification task example, the detection of saline-contaminated chemistry panels.Following validation, the laboratorian's role is far from over. Implementing machine learning solutions requires an interdisciplinary effort across several roles in an organization. We highlight the key roles, responsibilities, and terminologies for successfully deploying a validated solution into a live production environment. Finally, the implemented solution must be routinely monitored for signs of performance degradation and updated if necessary. SUMMARY:This mini-review aims to bridge the gap between theory and practice by highlighting key concepts in validation, implementation, and monitoring machine learning solutions effectively and responsibly in the clinical laboratory.

Objectives Lab testing is a high-volume activity that is often overused, leading to wasted resources and inappropriate care. Improving test ordering practices in tertiary care involves deciding where to focus scarce intervention resources, but clear guidance on how to optimize these resources is lacking. We aimed to explore context-sensitive factors and processes that inform individual decisions about laboratory stewardship interventions by speaking to key interest holders in this area.Methods We conducted semi-structured interviews with test-ordering intervention development experts and authors of test-ordering guidance documents to explore five broad topics: 1) processes used to prioritize tests for intervention; 2) factors considered when deciding which tests to target; 3) measurement of these factors; 4) interventions selected; 5) suggestions for a framework to support these decisions. Transcripts were double coded using directed-content and thematic analysis.Results We interviewed 14 intervention development experts. Experts noted they frequently consider test volume, test value, and patient care when deciding on a test to target. Experts indicated that quantifying many relevant factors was challenging. Processes to support these decisions often involved examining local data, obtaining buy-in, and relying on an existing guideline. Suggestions for building a framework emphasized the importance of collaboration, consideration of context and resources, and starting with "easy wins" to gain support and experience.Conclusions Our study provides insight into the factors and processes experts consider when deciding which tests to target for intervention and can inform the development of a framework to guide the selection of tests for intervention and guideline development.

Key Points Nearly half of all patients with CKD who progress to kidney failure initiate dialysis in an unplanned fashion, which is associated with poor outcomes. Machine learning models using routinely collected data can accurately predict 6- to 12-month kidney failure risk among the population with advanced CKD. These machine learning models retrospectively deliver advanced warning on a substantial proportion of unplanned dialysis events. Background Approximately half of all patients with advanced CKD who progress to kidney failure initiate dialysis in an unplanned fashion, which is associated with high morbidity, mortality, and health care costs. A novel prediction model designed to identify patients with advanced CKD who are at high risk for developing kidney failure over short time frames (6–12 months) may help reduce the rates of unplanned dialysis and improve the quality of transitions from CKD to kidney failure. Methods We performed a retrospective study using machine learning random forest algorithms incorporating routinely collected age and sex data along with time-varying trends in laboratory measurements to derive and validate 6- and 12-month kidney failure risk prediction models in the population with advanced CKD. The models were comprehensively characterized in three independent cohorts in Ontario, Canada—derived in a cohort of 1849 consecutive patients with advanced CKD (mean [SD] age 66 [15] years, eGFR 19 [7] ml/min per 1.73 m 2 ) and validated in two external advanced CKD cohorts ( n =1356; age 69 [14] years, eGFR 22 [7] ml/min per 1.73 m 2 ). Results Across all cohorts, 55% of patients experienced kidney failure, of whom 35% involved unplanned dialysis. The 6- and 12-month models demonstrated excellent discrimination with area under the receiver operating characteristic curve of 0.88 (95% confidence interval [CI], 0.87 to 0.89) and 0.87 (95% CI, 0.86 to 0.87) along with high probabilistic accuracy with the Brier scores of 0.10 (95% CI, 0.09 to 0.10) and 0.14 (95% CI, 0.13 to 0.14), respectively. The models were also well calibrated and delivered timely alerts on a significant number of patients who ultimately initiated dialysis in an unplanned fashion. Similar results were found upon external validation testing. Conclusions These machine learning models using routinely collected patient data accurately predict near-future kidney failure risk among the population with advanced CKD and retrospectively deliver advanced warning on a substantial proportion of unplanned dialysis events. Optimal implementation strategies still need to be elucidated.

Multiple myeloma frequently impacts the kidneys, with up to half of patients presenting with reduced kidney function at diagnosis.1Kyle R.A. Gertz M.A. Witzig T.E. et al.Review of 1027 patients with newly diagnosed multiple myeloma.Mayo Clin Proc. 2003; 78: 21-33https://doi.org/10.4065/78.1.21Abstract Full Text Full Text PDF PubMed Scopus (1759) Google Scholar, 2Finkel K.W. Cohen E.P. Shirali A. Abudayyeh A. American Society of Nephrology Onco-Nephrology ForumParaprotein-related kidney disease: evaluation and treatment of myeloma cast nephropathy.Clin J Am Soc Nephrol. 2016; 11: 2273-2279https://doi.org/10.2215/CJN.01640216Crossref PubMed Scopus (37) Google Scholar, 3Knudsen L.M. Hippe E. Hjorth M. Holmberg E. Westin J. Renal function in newly diagnosed multiple myeloma--a demographic study of 1353 patients. The Nordic Myeloma Study Group.Eur J Haematol. 1994; 53: 207-212https://doi.org/10.1111/j.1600-0609.1994.tb00190.xCrossref PubMed Scopus (223) Google Scholar, 4Dimopoulos M.A. Sonneveld P. Leung N. et al.International Myeloma Working Group recommendations for the diagnosis and management of myeloma-related renal impairment.J Clin Oncol. 2016; 34: 1544-1557https://doi.org/10.1200/JCO.2015.65.0044Crossref PubMed Scopus (269) Google Scholar The most common cause of kidney damage is light chain cast nephropathy,5Ecotiere L. Thierry A. Debiais-Delpech C. et al.Prognostic value of kidney biopsy in myeloma cast nephropathy: a retrospective study of 70 patients.Nephrol Dial Transplant. 2016; 31: 64-72https://doi.org/10.1093/ndt/gfv283Crossref PubMed Scopus (35) Google Scholar,6Bridoux F. Leung N. Belmouaz M. et al.Management of acute kidney injury in symptomatic multiple myeloma.Kidney Int. 2021; 99: 570-580https://doi.org/10.1016/j.kint.2020.11.010Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar and Bence Jones proteinuria is present in 64%-96% of cases at diagnosis.7Pascali E. Pezzoli A. The incidence of Bence-Jones proteinuria in multiple myeloma.Br J Haematol. 1986; 64: 847-850https://doi.org/10.1111/j.1365-2141.1986.tb02248.xCrossref PubMed Scopus (4) Google Scholar, 8Sinclair D. Dagg J.H. Smith J.G. Stott D.I. The incidence and possible relevance of Bence-Jones protein in the sera of patients with multiple myeloma.Br J Haematol. 1986; 62: 689-694https://doi.org/10.1111/j.1365-2141.1986.tb04092.xCrossref PubMed Scopus (18) Google Scholar, 9Smith D. Yong K. Multiple myeloma.BMJ. 2013; 346: f3863https://doi.org/10.1136/bmj.f3863Crossref PubMed Scopus (69) Google Scholar Nephrologists must be keenly aware of multiple myeloma as a potential diagnosis in cases of undifferentiated kidney dysfunction. Because of this, some employ a commonly taught “clinical pearl” assessing for discrepancy between the urinary protein-creatinine ratio (UPCR) and urinary albumin-creatinine ratio (UACR) as a surrogate screen for multiple myeloma. The theory behind this is that light chains in the urine will lead to discordance between the UPCR and the UACR, since overall protein levels will increase but albumin levels will not; however, other causes of non-albumin proteinuria are also known to occur (eg, tubular proteinuria). It remains unknown whether the absolute difference between UPCR and UACR, which we will refer to as the urine protein-to-albumin gap, is useful as a screen for multiple myeloma. We conducted a population-level retrospective cohort study of residents in Ontario, Canada, in 2009-2021 to examine the association between urine protein-to-albumin gap and subsequent diagnosis of multiple myeloma using linked databases held at ICES. ICES is an independent, not-for-profit research institute whose legal status under Ontario’s health information privacy law allows it to collect and analyze health care and demographic data, without consent, for health system evaluation and improvement. ICES captures data on all Ontario residents who undergo a health care encounter, including health care visits, laboratory tests, hospitalizations, and vital statistics. These datasets were linked using unique encoded identifiers and analyzed at ICES (Table S1). The use of data in this project is authorized under section 45 of Ontario’s Personal Health Information Protection Act and does not require review by a research ethics board. All Ontario residents aged 18-105 years with same-day quantifiable measurements of UPCR and UACR and without a prior history of multiple myeloma or monoclonal gammopathy from April 1, 2009, through March 31, 2021 were included (Fig S1). Individuals were categorized by quartile of urine protein-to-albumin gap and stratified at a UPCR threshold of 50 mg/mmol, the KDIGO definition of severely increased proteinuria.10Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work GroupKDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease.Kidney Int Suppl. 2013; 3: 1-150Abstract Full Text Full Text PDF Scopus (1846) Google Scholar Multivariable time-to-event models measured the association between urine protein-to-albumin gap and subsequent diagnosis of multiple myeloma. Full methods are in Item S1. Baseline characteristics are given in Table S2. Mean age was 63 ± 18 (SD) years and 49% of individuals were female. Mean serum creatinine and estimated glomerular filtration rate (eGFR) were 1.28 ± 0.96 mg/dL and 71 ± 32 mL/min/1.73 m2, respectively. Median UPCR and UACR were 27 (IQR, 12-58) mg/mmol and 4 (IQR, 1-19) mg/mmol, respectively. Among 28,231 eligible individuals, 116 were diagnosed with multiple myeloma (0.4%) a median of 31 days from UACR and UPCR measurement. Among the overall population, with each successive quartile of urine protein-to-albumin gap, myeloma incidence rate (Table 1) and risk (Fig 1A) were progressively greater. We found evidence of effect modification of the 50 mg/mmol UPCR threshold on the association between urine protein-to-albumin gap quartiles and multiple myeloma (P = 0.04). Among individuals with UPCR >50 mg/mmol, the association between successive quartile of urine protein-to-albumin gap and subsequent myeloma diagnosis was strengthened (Fig 1B). In contrast, among those with UPCR ≤50 mg/mmol, there was no significant association (Fig 1C), though potentially this relates to the lower number of cases poststratification. As this population-based study was based on administrative health care data, limitations included an inability to determine the reason a provider ordered same-day UACR and UPCR measurements (which may impact the pretest probability for multiple myeloma), and a lack of standardization of albuminuria and proteinuria measurements across the many laboratories involved (which may have impacted performance metrics).Table 1Incidence Rates of Multiple Myeloma by Urine Protein-to-Albumin Gap Quartile in the Overall Population and Upon Stratification at a UPCR Threshold of 50 mg/mmolUrine Protein-to-Albumin GapNo. of PatientsIncidence Rate of Multiple Myeloma (95% CI)P for TrendOverallMyeloma CasesOverall population<0.001 Quartile 1: ≤9.3 mg/mmol7,174145.63 (3.34-9.51) Quartile 2: 9.4-18.0 mg/mmol6,955167.06 (4.32-11.52) Quartile 3: 18.1-43.4 mg/mmol7,0502712.39 (8.50-18.07) Quartile 4: >43.4 mg/mmol7,0525928.54 (22.11-36.84)UPCR >50 mg/mmol<0.001 Quartile 1: ≤34.2 mg/mmol1,908≤5aIn accordance with ICES privacy policies, cell sizes ≤5 cannot be reported.6.14 (2.31-16.37) Quartile 2: 34.3-57.5 mg/mmol1,8926-1015.30 (7.65-30.59) Quartile 3: 57.6-111.8 mg/mmol1,8991534.66 (20.90-57.50) Quartile 4: >111.8 mg/mmol1,9003479.91 (57.10-111.84)UPCR ≤50 mg/mmol0.2 Quartile 1: ≤7.9 mg/mmol5,252137.03 (4.08-12.11) Quartile 2: 8.0-12.6 mg/mmol5,121116.53 (3.62-11.79) Quartile 3: 12.7-23.7 mg/mmol5,117106.23 (3.35-11.58) Quartile 4: >23.7 mg/mmol5,1422111.50 (7.50-17.63)Incidence rate is given per 10,000 person-years; P for trend is across quartiles.a In accordance with ICES privacy policies, cell sizes ≤5 cannot be reported. Open table in a new tab Incidence rate is given per 10,000 person-years; P for trend is across quartiles. In summary, these results show that among individuals with UPCR >50 mg/mmol, the risk for myeloma was significantly higher for urine protein-to-albumin gap greater than ˜50 mg/mmol (˜4-fold higher risk), particularly when this gap exceeded ˜100 mg/mmol (˜11-fold higher risk). However, when these thresholds are used as screening cutoffs for multiple myeloma, their performance is modest. For UPCR >50 mg/mmol with urine protein-to-albumin gap >50 mg/mmol, sensitivity was 85%; specificity, 43%; positive likelihood ratio, 1.49; and negative likelihood ratio, 0.35. For UPCR >50 mg/mmol and urine protein-to-albumin gap >100 mg/mmol, sensitivity was 61%; specificity, 72%; positive likelihood ratio, 2.16; and negative likelihood ratio, 0.55. These results may help guide clinicians who use UPCR-UACR discrepancy as a screen for potential multiple myeloma in cases of undifferentiated kidney dysfunction. In turn, this may provide information to guide when further testing to evaluate for multiple myeloma is imperative. Study concept and design: GLH, HI, MMS, AA; data acquisition/analysis: GLH, HI, MMS, AA; data interpretation: GLH, HI, AV, AM, GK, MB, MC, DM, PT, CM, MMS, AA. MMS and AA contributed equally to this work. Each author contributed important intellectual content during manuscript drafting or revision and agrees to be personally accountable for the individual’s own contributions and to ensure that questions pertaining to the accuracy or integrity of any portion of the work, even one in which the author was not directly involved, are appropriately investigated and resolved, including with documentation in the literature if appropriate. Dr Hundemer is supported by the Lorna Jocelyn Wood Chair for Kidney Research. Dr Sood is supported by the Jindal Research Chair for the Prevention of Kidney Disease. This study was supported by ICES. The funders had no role in study design; collection, analysis, and interpretation of the data; writing the report; or the decision to submit the report for publication. Dr Sood has received speaker fees from AstraZeneca. Dr Akbari has received speaker fees from AstraZeneca and holds research grants from Otsuka. The other authors declare that they have no relevant financial interests. ICES is funded by an annual grant from the Ontario Ministry of Health (MOH) and the Ministry of Long-Term Care (MLTC). This document used data adapted from the Statistics Canada Postal CodeOM Conversion File, which is based on data licensed from Canada Post Corporation, and/or data adapted from the Ontario Ministry of Health Postal Code Conversion File, whic contains data copied under license from Canada Post Corporation and Statistics Canada. We would like to thank IQVIA Solutions Canada Inc for use of their Drug Information File. Parts of this material are based on data and/or information compiled and provided by the Canadian Institute for Health Information, the Ontario MOH, and Ontario Health. The research was conducted by members of the ICES Kidney, Dialysis and Transplantation team, at the ICES Ottawa facility. The analyses, conclusions, opinions, and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended or should be inferred. Received July 28, 2022. Evaluated by 2 external peer reviewers, with direct editorial input from a Statistics/Methods Editor, an Associate Editor, and the Editor-in-Chief. Accepted in revised form November 6, 2022. Download .pdf (.32 MB) Help with pdf files Supplementary File (PDF) Figure S1; Item S1; Tables S1-S2.

Background: Plasma and RBC zinc values are unrelated in hospitalized patients. The independent association of these values with important patient outcomes is unknown. Objectives: Measure the independent association of plasma and RBC zinc with outcomes in hospitalized patients.Methods: Plasma and RBC zinc concentrations were prospectively measured within 48 h of hospitalization in consenting patients. Data were linked deterministically with population-based health administrative data to measure each association of zinc measures with 2 outcomes (time to death from any cause and likelihood of death or urgent readmission to hospital within 30-d of discharge) after adjusting for validated outcome risk scores.Results: In total, 250 people admitted to medical services were studied. Patients were ill with a 1-y baseline expected death risk (IQR) of 19.9% (6.3%-37.2%). The observed 1-y and 2-y all-cause death risks were 24.5% (95% CI: 19.6%, 30.3%) and 33.2% (95% CI: 27.3%, 39.9%), respectively. Death risk increased significantly as plasma zinc concentrations decreased (P = 0.0001). This association persisted even after adjusting for the baseline expected death risk (P = 0.02) with every 2-& mu;mol/L decrease in plasma zinc concentrations being independently associated with, on average, a 35% increase in the death risk. RBC zinc concentrations were not associated with the death risk. Neither plasma nor RBC zinc concentrations were significantly associated with the 30-d death or urgent readmission rate.Conclusions: Plasma, but not RBC, zinc concentrations are independently associated with the all-cause death risk in hospitalized medical patients. Further study is required to determine whether this association is causal and identify its potential causal pathways. Curr Dev Nutr 2023;x:xx.

Abstract Background Evidence indicates substantial inappropriate and inefficient use of resources in laboratory medicine. Commonly-used approaches that can be effective in improving practice include education, changes to processes or systems, and audit & feedback. The overall aims of this project were to survey the experiences of individuals implementing laboratory stewardship activities across a regional lab system, focusing on identifying key enablers and barrier to success. Methods We used a mixed methods approach beginning with a questionnaire to gather brief details about implementation efforts (what was done, how interventions were implemented, when, and by whom). These preliminary findings served as a precursor to conducting focus groups with implementation teams at each of the sites. We probed the implementation process, drivers of success, and barriers faced. Focus group transcripts were analyzed using thematic analysis. Results Forty individuals took part in the surveys and focus groups. Results revealed a range of progress with lab stewardship initiatives across sites, with a variety of initiatives having been implemented. A wide range of factors impacting progress with laboratory stewardship, both negatively and positively, were discussed. These were grouped into four broad topics: 1) Roles, responsibilities, and relationships—the “Who”; 2) Generating interest in stewardship—the “Why”; 3) Implementation strategies and processes—the “What”; 4) Project management and resources—the “How”. Key barriers identified included: lack of a clear vision which reduces relative priority, lack of resources, emphasis on financial outcomes, and that over-use problems have complex causes. Key enablers identified included active participation of hospital clinical and administrative champions, ability to implement system level changes (e.g., changing electronic orders and processes), focusing on patient safety outcomes, using established change methodologies which support systematic identification of root causes to help select appropriate change strategies as well as formal evaluation. Conclusion While some sites progressed with stewardship initiatives with relative ease, most faced challenges along the way. Surfacing these challenges supports the development of strategies to help address them such that lab stewardship activities can have maximal impact in improving quality of care. Structured and systematic approaches to laboratory stewardship are more likely to be successful than those without designed root cause analysis with targeted interventions and clear outcome metrics.

Rationale:The differential diagnosis for a patient with high-anion-gap metabolic acidosis (HAGMA) is broad; lactic acidosis is an important entity to screen for and treat. An elevated serum lactate is often used as a marker of inadequate tissue perfusion in critically ill patients but can also be indicative of decreased lactate utilization or poor hepatic clearance. Investigating for the underlying cause such as diabetic ketoacidosis, malignancy, or culprit medications is essential to establish the diagnosis and treatment plan.Presenting concerns of the patient:A 60-year-old man with a history of substance use and end-stage kidney disease treated with hemodialysis presented to hospital with confusion, altered level of consciousness, and hypothermia. Initial laboratory investigations were significant for a severe HAGMA with elevated serum lactate and β-hydroxybutyrate levels, but toxicology screen was negative, and there was no clear underlying precipitant. Urgent hemodialysis was arranged to mitigate his severe acidosis.Diagnoses:He had an initial single dialysis treatment for 4 hours, with posthemodialysis labs showing significant improvement in his acidosis, serum lactate level, and clinical status (cognition, hypothermia). Given this rapid resolution, a sample from his predialysis blood work was sent for analysis of plasma metformin and returned significantly elevated at 60 mcg/mL (therapeutic range 1-2 mcg/mL).Interventions and outcomes:On careful medication reconciliation in the dialysis unit, the patient stated he had never heard of the medication metformin, and there was no record of a filled prescription at his pharmacy. Given his living situation with shared accommodations, it was presumed that he had taken medications that were prescribed to a roommate. Several of his other medications including his antihypertensives were subsequently given after dialysis on dialysis days to improve adherence.Teaching points:Maintain a broad differential diagnosis for patients presenting with a clinical syndrome consistent with an acute toxicity even if no culprit medications are identifiable on history, especially in patients with a suggestive social history.Anion-gap metabolic acidosis (AGMA) is common in hospitalized patients but sometimes requires further history and/or confirmatory testing to elucidate the root cause underlying typical causes of AGMA such as lactic acidosis or ketoacidosis.The main treatment of metformin toxicity is resuscitation and supportive care; however, metformin's biochemical properties make it readily dialyzable via either diffusion or convection.The Extracorporeal Treatments In Poisoning group recommends hemodialysis for metformin toxicity when there is a serum lactate >20 mmol/L, a blood pH <7.0, a failure of standard therapy, end-organ damage (hepatic or renal insufficiency), or a decreased level of consciousness.

Abstract Background Plasma has long been the recommended specimen type for K measurement. However, serum usage persists, including use due to recent plasma tube shortages. In comparison to plasma K, serum K is generally right-shifted (higher). This shift varies across individual samples according to a normal distribution. Consequentially, K results outside of the population reference interval (RI) for plasma (hypokalemia or hyperkalemia) are unlikely to be strictly concordant with classification in serum according to the serum RI. Serum may thus have a lower sensitivity to detect either hypokalemia or hyperkalemia relative to plasma. We examined this premise from a theoretical standpoint by simulation. Methods We used longstanding and widely implemented textbook K reference intervals (Tietz, 4th Ed.) for plasma (PRI = 3.4–4.5 mmol/L) and serum (SRI = 3.5–5.1 mmol/L). The difference between plasma K and serum K is characterized by a normally distributed function where serum = plasma + 0.35 ± 0.3 mmol/L. This transformation was applied to an at-large patient data distribution from a large academic medical center to generate a theoretical serum K distribution from real world plasma patient K results. The simulated serum K data were characterized as to whether hypokalemic and hyperkalemic plasma K specimens as defined by PRI were also classified in serum as being either below or above lower and upper limits of SRI, respectively. Results Primary data were a plasma K patient distribution for a three month interval (n = 59 570; median = 4.1 mmol/L; hypokalemia = 6.5%; hyperkalemia = 16.7%). Simulated serum K data from the plasma transformation (n = 100 000) yielded a right-shifted distribution (median = 4.4 mmol/L) with 4.1% of results below the lower limit of SRI, and 11.4% of results above the upper limit of SRI. For samples originating as hypokalemic according to the PRI, sensitivity for detection by being flagged as below the lower limit of the SRI was 44.1% (specificity = 98.7%). For samples originating as hyperkalemic according to the PRI, sensitivity for detection by being flagged as above the upper limit of the SRI was 58.7% (specificity = 97.6%). Conclusions Simulation results indicate that serum K should best be thought of as an inferior substitute marker for plasma K. Depending on reference intervals employed, sensitivity of serum RI for detection either of plasma hypokalemia or hyperkalemia may be significantly less than 100%. This can result in a significant degree of misidentification of hyperkalemia or hyperkalemia as being within SRI. In our simulation, 55.9% of plasma hypokalemia results were in serum classified as being within the SRI; 41.3% of plasma hyperkalemia results were in serum classified as being within the SRI. Additionally, note that a switch from serum to plasma will likely precipitate a significant increase in flag rates for plasma K outside of the PRI compared to flag rates observed for serum K outside of the SRI. These results follow simply from the fact that serum K includes a random component not present in plasma.

Background Development of a short timeframe (6-12 months) kidney failure risk prediction model may serve to improve transitions from advanced chronic kidney disease (CKD) to kidney failure and reduce rates of unplanned dialysis. The optimal model for short timeframe kidney failure risk prediction remains unknown. Methods This retrospective study included 1757 consecutive patients with advanced CKD (mean age 66 years, estimated glomerular filtration rate 18 mL/min/1.73 m(2)). We compared the performance of Cox regression models using (a) baseline variables alone, (b) time-varying variables and machine learning models, (c) random survival forest, (d) random forest classifier in the prediction of kidney failure over 6/12/24 months. Performance metrics included area under the receiver operating characteristic curve (AUC-ROC) and maximum precision at 70% recall (PrRe70). Top-performing models were applied to 2 independent external cohorts. Results Compared to the baseline Cox model, the machine learning and time-varying Cox models demonstrated higher 6-month performance [Cox baseline: AUC-ROC 0.85 (95% CI 0.84-0.86), PrRe70 0.53 (95% CI 0.51-0.55); Cox time-varying: AUC-ROC 0.88 (95% CI 0.87-0.89), PrRe70 0.62 (95% CI 0.60-0.64); random survival forest: AUC-ROC 0.87 (95% CI 0.86-0.88), PrRe70 0.61 (95% CI 0.57-0.64); random forest classifier AUC-ROC 0.88 (95% CI 0.87-0.89), PrRe70 0.62 (95% CI 0.59-0.65)]. These trends persisted, but were less pronounced, at 12 months. The random forest classifier was the highest performing model at 6 and 12 months. At 24 months, all models performed similarly. Model performance did not significantly degrade upon external validation. Conclusions When predicting kidney failure over short timeframes among patients with advanced CKD, machine learning incorporating time-updated data provides enhanced performance compared with traditional Cox models.

Objectives: Serum potassium (K) exhibits a positive shift relative to plasma K due to a variable amount of K release associated with clotting. Because of this variation, plasma K results outside of the reference interval (RI) for plasma (hypokalemia or hyperkalemia) in individual samples may not produce classification-concordant results in serum according to the serum RI. We examined this premise from a theoretical standpoint by simulation. Design & methods: We used textbook K reference intervals for plasma (PRI = 3.4-4.5 mmol/L) and serum (SRI = 3.5-5.1 mmol/L). The difference between PRI and SRI is characterized by a normal distribution: serum K = plasma K + 0.35 & PLUSMN; 0.308 mmol/L. This transformation was applied by simulation to an observed patient data distribution for plasma K to generate a corresponding theoretical serum K distribution. Individual samples were tracked for comparison with respect to classification (below, within, above RI) for plasma and serum. Results: Primary data were an all-comers plasma K patient distribution (n = 41,768; median = 4.1 mmol/L; 7.1% below PRI (hypokalemia); 15.5% above PRI (hyperkalemia)). Simulation to obtain the associated serum K yielded a right-shifted distribution (median = 4.4 mmol/L; 4.8% below SRI; 10.8% above SRI). Sensitivity for detection in serum (flagged below SRI) for samples originating as hypokalemic in plasma was 45.7% (specificity = 98.3%). Sensitivity for detection in serum (flagged above SRI) for samples originating as hyperkalemic in plasma was 56.6% (specificity = 97.6%). Conclusions: Simulation results indicate that serum K should best be thought of as an inferior substitute marker for plasma K. These results follow simply from the variable component of serum K compared to plasma K. Plasma should be the preferred specimen type for K assessment.

Objectives There is bountiful evidence of ambulatory patient over-testing, both in terms of the tests ordered and their frequency. To be thorough and efficient, many physicians pay short shrift to sustainable test ordering. We have developed a calculus that transforms sequential intra-patient test results into total variation comprising its preanalytical, analytical and biological components. We have used this approach to compare the analytical and clinical performance of multiple analytical systems. Here, we determine the total test variation in patients with test separation intervals of 90, 180, 270 and 365 days. We demonstrate that annual testing with phlebotomies scheduled at the same time of day minimizes the variation of the sequential test results and permits optimal discrimination of a statistically significant change in the test results. Method We studied 5 years (2014-2019) of Ottawa Hospital adult outpatient general chemistry test pairs of albumin, alanine aminotransferase, aspartate aminotransferase, calcium, chloride, bicarbonate, creatinine, potassium, magnesium, sodium, phosphate, total bilirubin and urea. Intrapatient data pairs (separated by 88-92, 178-182, 268-272 and 363-367 days), were divided into 2 groups: those sampled on weekdays at the same time of day (+ 2 hr) and those with 3 to 8 hr sampling differences . The between pair variations were derived calculated with Dahlberg's formula. Results The median numbers of patient pairs/analyte for the matched sampling times were 1034 (90 days), 696 (180 days), 84 (270 days) and 364 (365 days). For the unmatched sampling times, the median numbers of patient pairs were much smaller, ranging between 38 to 290 and their average variations exceeded the matched pair variations by as much as 30% (phosphate and creatinine). The matched pair 90, 180, 270 and 365 day variations were highly correlated (r ranged from 0.965 to 0.977). Compared to the other periods, the 180 day variations were smaller (-7.5%) and the 270 day variations were larger (+11.9%). Overall, the 365 day magnitudes of variations were closest to the magnitudes of variation of the other periods. Conclusions At least £2.2 billion are spent annually on UK pathology services, a significant portion wasted on suboptimal testing including: 1) tests are ordered too closely, providing little new information 2) tests that are ordered 9 months later and have been influenced by seasonal factors or 3) sequential tests that are ordered at different times of the day and are influenced by diurnal variation. Biological variation can be a guide to reducing test wastage and setting a standard for testing intervals that are minimally one year with the patient being sampled at the same time of day. Simple algorithms incorporating minimal sequential data would be able to discern real laboratory trends far more readily than 'thorough and efficient' but over-testing clinicians.

Abstract Background Sequential patient data can be transformed into the variable component of the reference change value (RCV) calculation: (CVA2 + CVI2)1/2 (PMID 35137000). We have analyzed sequential healthy and abnormal outpatient ALT to derive RCVs of Ottawa Hospital outpatients (OHOP) who had sequential ALT measured by Siemens or Roche assays. Methods We used the American College of Gastroenterology definition of normal ALT with supplementation of P5P as <25 U/L and <33 U/L in females and males, respectively. We defined abnormal as between the upper limit of normal (ULN) and 3xULN. 5 years of Siemens ALT measurements and subsequently 3 years of Roche ALT were abstracted from the Ottawa Hospital EMR. For the 8 patient cohorts, we tabulated consecutive pairs of intrapatient ALT by time intervals of separation: 0–1 weeks, 1–2 weeks, 2–3 weeks . . ., up to 51–52 weeks. For each interval, we determined the standard deviation of duplicates (SDD) between the paired intrapatient results. SDD was graphed against the midpoints of the weekly interval. Linear regression was used to determine the y-intercept (yo). The 95% RCVs of the male and female, normal and abnormal populations were calculated for the Siemens (with P5P) and Roche (non P5P) assays. Results The Table describes each cohort, yo and RCV. According to the Roche/Siemens cross-over study, Roche ALT values were 10% less than Siemens. This Roche bias is reflected in the normal outpatient ALT means and less so in those with elevated ALT. Also presented are the RCV of low ALT (normal) patients as documented by Ricos and Carobene. Conclusion Our RCVs are realistic: normal ALT patients demonstrate lower RCVs, close to Ricos’ but much larger than Carobene’s which are unachievable due to diurnal variation (over 40% of OHOP ALT are drawn at different times of the day). Furthermore, Carobene’s CVA is tiny.

Objectives To examine local patient safety events related to the administration of anti-Rh(D) immune globin (RhIG) during pregnancy, and to follow-up with targeted educational intervention to improve knowledge of this process.Background Administration RhIG is established treatment for the prevention of haemolytic disease of the foetus and newborn (HDFN). However, patient safety events in relation to its correct use continue to occur.Methods A retrospective audit of patient safety events related to RhIG administration during pregnancy was performed. Targeted educational intervention in the form of PowerPoint (R) presentation were given to nursing staff, laboratory staff and physicians and evaluated with pre- and post-tests using multiple-choice questions given immediately before and after the presentation.Results An annual incidence of 0.24% of patient safety events related to the administration of RhIG during pregnancy was found. These events were mostly in the preanalytical phase, for example mislabelled samples or samples for D-rosette/Kleihauer-Betke testing drawn from the baby, not the mother. Using Bayesian analysis, the probability of positive effect for the targeted educational intervention was 100% with a median improved score of 29%. This was compared with a control group using standard curriculum education intervention based on the current curriculum for nursing, laboratory and medical students which showed a median improved score of only 4.4%.Conclusions Administration of RhIG during pregnancy is a multistep process involving health care professionals of several disciplines providing opportunities to enhance the curriculum for nursing, laboratory and medical students and to ensure on-going education.

BACKGROUND AND OBJECTIVES:Regression models incorporating laboratory tests treat unordered tests as missing and are often imputed. Imputation typically assumes that data are "missing at random" (MAR, test's order status is unrelated to its result after accounting for other variables). This study examined the validity of this assumption. METHODS:We included 14 biochemistry tests. All tests were measured regardless of test order status. Test-stratified multiple linear regression determined the independent association between test result and order status after adjusting for patient age, sex, comorbidities, and patient location. Testing likelihood models were created for all tests using hospital-wide data. RESULTS:Four hundred thirty-four patients were included (mean age [standard deviation] 60.7 [19.1], 50.5% female). In 9 of 14 tests (64.2%), test results were significantly associated with order status after adjustment. Results were significantly more abnormal when tests were ordered for 6 tests and significantly more normal for 3 tests. Test abnormality increased as testing likelihood decreased. CONCLUSIONS:These data suggest that laboratory data are often not MAR. The direction and extent of differences in missing laboratory test values varies between tests. Overall the abnormality of ordered tests increased as testing likelihood decreased. These results suggest that imputating missing laboratory data may return biased values.

Abstract Introduction Some laboratory testing practices may be of low value, leading to wasted resources and potential patient harm. Our scoping review investigated factors and processes that developers report using to inform decisions about what tests to target for practice improvement. Methods We searched Medline on May 30th, 2019 and June 28th, 2021 and included guidelines, recommendation statements, or empirical studies related to test ordering practices. Studies were included if they were conducted in a tertiary care setting, reported making a choice about a specific test requiring intervention, and reported at least one factor informing that choice. We extracted descriptive details, tests chosen, processes used to make the choice, and factors guiding test choice. Results From 114 eligible studies, we identified 30 factors related to test choice including clinical value, cost, prevalence of test, quality of test, and actionability of test results. We identified nine different processes used to inform decisions regarding where to spend intervention resources. Conclusions Intervention developers face difficult choices when deciding where to put scarce resources intended to improve test utilization. Factors and processes identified here can be used to inform a framework to help intervention developers make choices relevant to improving testing practices.