A Novel Enzymatic Hydrolysis Method for Urine Aldosterone Quantification: A Case for Reassessing Clinical Cut-Offs of Primary Aldosteronism.

BACKGROUND: Primary aldosteronism (PA) is a common endocrine cause of secondary hypertension. The aldosterone/renin ratio is an important tool for PA screening, and dynamic testing in serum or urine is used to confirm the diagnosis. While LC-MS/MS is considered the gold standard for testing, there is significant interlaboratory variability between the extraction procedures, which can impact diagnostic interpretation. To help overcome this, we present a simple and accurate LC-MS/MS method for the quantification of both serum and urine aldosterone using a novel enzymatic hydrolysis procedure. METHODS: Serum and urine aldosterone was extracted and measured by LC-MS/MS. Urine-conjugated aldosterone glucuronide was hydrolyzed using a genetically modified glucuronidase enzyme. The assay precision, accuracy, limit of quantification, recovery, and carryover were evaluated and the new assay cut-offs were proposed. RESULTS: The liquid chromatography method allowed for adequate separation of the aldosterone peak from closely eluting peaks. Significant in vitro aldosterone loss was observed during acid-catalyzed hydrolysis of urine, which was corrected with the addition of the internal standard to the urine before the hydrolysis step. Glucuronidase catalyzed hydrolysis of urine aldosterone glucuronide displays good correlation with the corrected acid-catalyzed hydrolysis. Serum aldosterone showed good agreement with reference values and the consensus range reported for external quality assessment specimens. CONCLUSION: A simple, fast, and highly accurate method for the detection of serum and urine aldosterone has been developed. The proposed novel enzymatic procedure allows for short hydrolysis time and compensates for urine aldosterone loss during the hydrolysis step.

The Present and Future of Lipid Testing in Cardiovascular Risk Assessment.

BACKGROUND: Lipids play a central role in the pathogenesis of cardiovascular disease (CVD), a leading cause of morbidity and mortality worldwide. Plasma lipids and lipoproteins are routinely measured to help identify individuals at high risk of developing CVD and to monitor patients' response to therapy. The landscape of lipid testing is rapidly changing, including new ways to estimate traditional lipid parameters (e.g., low-density lipoprotein-cholesterol [LDL-C] calculations) and new lipid parameters that show superiority for risk prediction (e.g., non-high-density lipoprotein-cholesterol [non-HDL-C], apolipoprotein B [apoB], and lipoprotein a [Lp(a)]). CONTENT: Various national guidelines for managing dyslipidemia to prevent CVD are available, which primarily focus on LDL-C for identifying those at high risk and setting thresholds for optimal response to therapy. However, LDL-C can be calculated and measured in various ways, each with advantages and disadvantages. Importantly, the recently established Sampson-NIH LDL-C equation appears to be superior to preceding calculations, as is clear from the literature and in guidelines. There is now a shift towards using lipid parameters other than LDL-C, such as non-HDL-C, apoB, and Lp(a), to identify high-risk patients and/or establish treatment targets. SUMMARY: The goal of this review is to discuss the present and future of lipid testing for CVD risk assessment through describing various national clinical guidelines, critically reviewing methods to calculate and measure LDL-C and discussing the clinical utility of additional lipid parameters.

Diverse psychotropic substances detected in drug and drug administration equipment samples submitted to drug checking services in Toronto, Ontario, Canada, October 2019-April 2020.

BACKGROUND: The overdose crisis has generated innovative harm reduction and drug market monitoring strategies. In Toronto, Ontario, Canada, a multi-site drug checking service (DCS) pilot project was launched in October 2019. The project provides people who use drugs with information on the chemical composition of their substances, thereby increasing their capacity to make more informed decisions about their drug use and avoid overdose. DCS also provides real-time market monitoring to identify trends in the unregulated drug supply. METHODS: Sample data were obtained through analyses of drug and used drug administration equipment samples submitted anonymously and free of charge to DCS in downtown Toronto from October 10, 2019, to April 9, 2020, representing the first six months of DCS implementation. Analyses were conducted in clinical laboratories using liquid chromatography- and/or gas chromatography-mass spectrometry (LC-MS, GC-MS) techniques. RESULTS: Overall, 555 samples were submitted, with 49% (271) of samples that were found to contain high-potency opioids, of which 87% (235) also contained stimulants. Benzodiazepine-type drugs were found in 21% (116) of all samples, and synthetic cannabinoids in 1% (7) of all samples. Negative effects (including overdose, adverse health events, and extreme sedation) were reported for 11% (59) of samples submitted for analysis. CONCLUSIONS: Toronto's DCS identified a range of high-potency opioids with stimulants, benzodiazepine-type drugs, and a synthetic cannabinoid, AMB-FUBINACA. This information can inform a range of evidence-informed overdose prevention efforts.

Overutilization in laboratory medicine: tackling the problem with quality improvement science.

Overutilization of tests and treatments is a widespread problem in contemporary heath care, and laboratory medicine is no exception. It is estimated that 10-70% of laboratory tests may be unnecessary, with estimates in the literature varying depending on the situation and the laboratory test. Inappropriate use of laboratory tests can lead to further unnecessary testing, adverse events, inaccurate diagnoses, and inappropriate treatments. Altogether, this increases the risk of harm to a patient, which can be physical, psychological, or financial in nature. Overutilization in healthcare is driven by complex factors including care delivery models, litigious practice environments, and medical and patient culture. Quality improvement (QI) methods can help to tackle overutilization. In this review, we outline the global healthcare problem of laboratory overutilization, particularly in the developed world, and describe how an understanding of and application of quality improvement principles can help to address this challenge.

Xylazine detected in unregulated opioids and drug administration equipment in Toronto, Canada: clinical and social implications.

BACKGROUND: The North American opioid overdose crisis is driven in large part by the presence of unknown psychoactive adulterants in the dynamic, unregulated drug supply. We herein report the first detection of the psychoactive veterinary compound xylazine in Toronto, the largest urban center in Canada, by the city's drug checking service. METHODS: Toronto's Drug Checking Service launched in October 2019. Between then and February 2021, 2263 samples were submitted for analysis. The service is offered voluntarily at harm reduction agencies that include supervised consumption services. Samples were analyzed using gas chromatography-mass spectrometry or liquid chromatography-high resolution mass spectrometry. Targeted and/or untargeted screens for psychoactive substances were undertaken. RESULTS: In September 2020, xylazine was first detected by Toronto's Drug Checking Service. Among samples analyzed from September 2020 to February 2021 expected to contain fentanyl in isolation (610) or in combination with methamphetamine (16), xylazine was detected in 46 samples (7.2% and 12.5% of samples, respectively). Samples were predominantly drawn from used drug equipment. Three of the samples containing xylazine (6.5%) were associated with an overdose. CONCLUSION: We present the first detection of xylazine in Toronto, North America's fourth-largest metropolitan area. The increased risk of overdose associated with use of xylazine and its detection within our setting highlights the importance of drug checking services in supporting rapid responses to the emergence of potentially harmful adulterants. These data also highlight the clinical challenges presented by the dynamic nature of unregulated drug markets and the concomitant need to establish regulatory structures to reduce their contribution to overdose morbidity and mortality.

Glucosamine induces ER stress by disrupting lipid-linked oligosaccharide biosynthesis and N-linked protein glycosylation.

Glucosamine is an essential substrate for N-linked protein glycosylation. However, elevated levels of glucosamine can induce endoplasmic reticulum (ER) stress. Glucosamine-induced ER stress has been implicated in the development of diabetic complications, including atherosclerosis and hepatic steatosis. In this study, we investigate the potential relationship between the effects of glucosamine on lipid-linked oligosaccharide (LLO) biosynthesis, N-linked glycosylation, and ER homeostasis. Mouse embryonic fibroblasts (MEFs) were cultured in the presence of 0-5 mM glucosamine for up to 18 h, and LLO biosynthesis was monitored by fluorescence-assisted carbohydrate electrophoresis. ER stress was determined by quantification of unfolded protein response (UPR) gene expression. We found that exposure of MEFs to ≥1 mM glucosamine significantly impaired the biosynthesis of mature (Glc3Man9GlcNAc2) LLOs before the activation of the UPR, which resulted in the accumulation of an LLO intermediate (Man3GlcNAc2). The addition of 4-phenylbutyric acid (4-PBA), a chemical chaperone, was able to alleviate ER stress but did not rescue LLO biosynthesis. Other ER stress-inducing agents, including dithiothreitol and thapsigargin, had no effect on LLO levels. Together, these data suggest that elevated concentrations of glucosamine induce ER stress by interfering with lipid-linked oligosaccharide biosynthesis and N-linked glycosylation. We hypothesize that this pathway represents a causative link between hyperglycemia and the development of diabetic complications.

Deletion of Myeloid GSK3α Attenuates Atherosclerosis and Promotes an M2 Macrophage Phenotype.

OBJECTIVE: Glycogen synthase kinase (GSK)-3α/ß has been implicated in the pathogenesis of diabetes mellitus, cancer, Alzheimer, and atherosclerosis. The tissue- and homolog-specific functions of GSK3α and ß in atherosclerosis are unknown. This study examines the effect of hepatocyte or myeloid cell deletion of GSK3α or GSK3ß on atherosclerosis in low-density lipoprotein receptor (LDLR)(-/-) mice. APPROACH AND RESULTS: We ablated GSK3α or GSK3ß expression in hepatic or myeloid cells of LDLR(-/-) mice, and mice were fed a high-fat diet for 10 weeks. GSK3α or GSK3ß deficiency in hepatic or myeloid cells did not affect metabolic parameters, including plasma lipid levels. Hepatic deletion of GSK3α or GSK3ß did not affect the development of atherosclerosis or hepatic lipid content. Myeloid deletion of GSK3α, but not of GSK3ß, reduced atherosclerotic lesion volume and lesion complexity. Mice lacking GSK3α in myeloid cells had a less inflammatory and more anti-inflammatory plasma cytokine profile. Macrophages within atherosclerotic lesions of myeloid GSK3α-deficient mice, but not of GSK3ß-deficient mice, displayed reduced expression of markers associated with M1 macrophage polarization and enhanced expression of the M2 markers. Finally, bone marrow-derived macrophages were isolated and differentiated into classical M1 macrophages or alternative M2 macrophages in vitro. GSK3α deletion, but not GSK3ß deletion, attenuated the expression of genes associated with M1 polarization while promoting the expression of genes associated with M2 polarization by modulating STAT3 and STAT6 activation. CONCLUSIONS: Our findings suggest that deletion of myeloid GSK3α attenuates the progression of atherosclerosis by promoting an M2 macrophage phenotype.

Detection and quantification of endoplasmic reticulum stress in living cells using the fluorescent compound, Thioflavin T.

The endoplasmic reticulum (ER) plays a central role in the co- and post-translational modification of many proteins. Disruption of these processes can lead to the accumulation of misfolded proteins in the endoplasmic reticulum - a condition known as endoplasmic reticulum stress. In recent years, the association of endoplasmic reticulum stress with a number of disease pathologies has increased interest in the study of this condition. Current methods to detect endoplasmic reticulum stress are indirect and retrospective. Here we describe a new method to detect and quantify endoplasmic reticulum stress in live cells using Thioflavin T (ThT), a small molecule that exhibits enhanced fluorescence when it binds to protein aggregates. We show that enhanced ThT-fluorescence correlates directly with established indicators of unfolded protein response activation. Furthermore, enhanced ThT-fluorescence can be detected in living cells within 20 min of application of an endoplasmic reticulum stress-inducing agent. ThT is capable of detecting endoplasmic reticulum stress induced by distinctly different conditions and compounds, in different cultured cell types as well as in mouse tissue samples. Pre-treatment with a potent endoplasmic reticulum stress-reducing agent, 4-phenylbutyric acid, mitigates the enhanced ThT signal. This new tool will be useful in future research investigating the role of protein misfolding in the development and/or progression of human diseases.

Oral glucosamine sulfate supplementation does not induce endoplasmic reticulum stress or activate the unfolded protein response in circulating leukocytes of human subjects.

Glucosamine sulfate is a dietary supplement that is marketed as a treatment for osteoarthritis. Recent evidence from animal and cell culture models have suggested that glucosamine treatment can promote the misfolding of proteins and the activation of the unfolded protein response (UPR). We investigated whether glucosamine sulfate supplementation activates the UPR in circulating leukocytes of human subjects. Cultured Thp1 human monocytes were exposed to increasing concentrations of glucosamine (0, 0.25, 1.0, 4.0 mmol · L(-1)) for 18 h. We observed a dose-dependent increase in intracellular glucosamine levels as well as the activation of UPR. To test the effect of glucosamine sulfate supplementation in humans, 14 healthy human subjects took 1500 mg · day(-1) glucosamine sulfate for 14 days. Metabolic parameters and blood samples were collected before and after supplementation. In humans, glucosamine sulfate supplementation did not alter metabolic parameters including lipid levels and glucose tolerance. Further, glucosamine sulfate supplementation did not affect intracellular glucosamine levels or activate the UPR in the leukocytes of human subjects. Our results indicate that in healthy human subjects, the recommended dose of glucosamine sulfate (1500 mg · day(-1)) for 14 days does not significantly alter intracellular glucosamine levels and does not activate the UPR in circulating leukocytes.

Canadian Society of Clinical Chemists Harmonized Pediatric Lipid Reporting Recommendations for Clinical Laboratories.

Detecting dyslipidemia early is important because atherosclerosis originates in childhood and early treatment can improve outcomes. In 2022, the Canadian Cardiovascular Society (CCS) and Canadian Pediatric Cardiology Association (CPCA) published a clinical practice update to detect, evaluate, and manage pediatric dyslipidemia. However, guidance on its translation into clinical laboratories is lacking. The Canadian Society of Clinical Chemists Working Group on Reference Interval Harmonization Lipid Team aims to assist guideline implementation and promote harmonized pediatric lipid reporting across Canada. The 2022 CCS/CPCA clinical practice update, 2011 National Heart, Lung, and Blood Institute integrated guidelines, and new data analysis (Canadian pediatric reference values from the Canadian Laboratory Initiative on Pediatric Reference Intervals [CALIPER] and retrospective patient data from large community laboratories) were incorporated to develop 5 key recommendations. These include recommendations to: (1) offer nonfasting and fasting lipid testing; (2) offer a lipid panel including total cholesterol, low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), non-HDL-C, and triglycerides, with apolipoprotein B and lipoprotein(a) available as individually orderable tests; (3) flag total cholesterol, LDL-C, and non-HDL-C results ≥ 95th percentile, and HDL-C results < 10th percentile, as recommended by CCS/CPCA/National Heart, Lung, and Blood Institute and validated by CALIPER, and flag apolipoprotein B and nonfasting triglyceride results ≥ 95th percentile on the basis of CALIPER, and do not flag Lp(a) results but mention the adult cutoff in the interpretive comments; (4) implement interpretive comments listed in the current report; and (5) implement the National Institutes of Health LDL-C equation. The Canadian Society of Clinical Chemists Working Group on Reference Interval Harmonization Lipid Team will support clinical laboratories to implement these recommendations using knowledge translation strategies. Harmonizing pediatric lipid reporting across Canadian clinical laboratories will optimize clinical decision-making and improve cardiovascular risk management in youth.

Drug checking services as a surveillance tool for clinical laboratories: Examining trends in the unregulated fentanyl supply.

OBJECTIVES: Timely assessment and understanding of drug trends is essential for clinical laboratories to effectively respond to the overdose epidemic. In this proof-of-concept study, we sought to determine whether information obtained through Toronto's Drug Checking Services (DCS) and cross-provincial urine drug testing (UDT) data can be used as a surveillance tool for clinical laboratories and discuss the value of collaboration between the clinical laboratory, clinicians, and community partners to optimize patient care. DESIGN & METHODS: Mass spectrometry-based UDT data from LifeLabs Ontario (n = 127,529) and British Columbia (n = 14,848), and drug checking data from Toronto DCS (n = 3,308 drugs or used paraphernalia) was collected between August 2020 and October 2021. Fentanyl co-positivity with toxic adulterants such as benzodiazepine-related drugs and fentanyl analogues were examined. RESULTS: The percent co-positivity of fentanyl with etizolam, flualprazolam, flubromazolam, carfentanil, and acetylfentanyl in both Ontario UDT and DCS drugs/used paraphernalia showed similar trends. Regional differences in co-positivity with etizolam and fentanyl analogues were noted between Ontario and British Columbia UDT with patterns consistent over the entire 15-month collection period. CONCLUSIONS: Clinical laboratories should connect with their local DCS, if available, to understand and monitor unregulated drug trends. These data can be used as an important tool to help clinical laboratories tailor their UDT menus and thereby provide a community-focused service to improve patient care.

Analytical Unreliability of 25 Hydroxy Vitamin D Measurements in Pre-Term Neonates.

BACKGROUND: Vitamin D supplementation is common practice for neonates and infants due to limited stores of vitamin D at birth. Although not commonly encountered, vitamin D toxicity can occur due to over-supplementation. However, toxic concentrations are often not included in method validation experiments, and assays often are not validated in the neonatal population. METHODS: We compared serial 25 hydroxy vitamin D [25(OH)D] measurements in pre-term neonates receiving 25(OH)D supplementation and identified 12 patients wherein concentrations of 25(OH)D were above 50 ng/mL (125 nM) that required additional investigations as the 25(OH)D results did not match the clinical picture. Available samples were compared across 4 immunoassay platforms (LIAISON XL, Roche Cobas e602, Abbott Alinity i, and Siemens Centaur XP) and LC-MS/MS. RESULTS: Concentrations of 25(OH)D observed on one individual immunoassay platform (LIAISON XL) fluctuated substantially between subsequent blood draws in select neonates with elevated concentrations. Serum samples from these patients showed variable agreement between LC-MS/MS and other immunoassay platforms. These fluctuations were not explained by the presence of 3-epimer-25(OH)D or 24,25(OH)2D. CONCLUSIONS: Although we were unable to identify a cause for the variable elevated results, our findings suggest that neonatal 25(OH)D measurements alone should not be used for assessment of nutritional monitoring, and that clinical correlation and other laboratory parameters including ionized calcium should be considered.

Validating the NIH LDL-C equation in a specialized lipid cohort: Does it add up?

BACKGROUND: Guideline recommendations for the management of lipids in patients at risk for cardiovascular disease is largely based on low-density lipoprotein cholesterol (LDL-C) concentration. LDL-C is commonly calculated by the Friedewald equation, which has many limitations. The National Institutes of Health (NIH) equation better estimates LDL-C, particularly in patients with hypertriglyceridemia and/or low LDL-C. We validated the NIH LDL-C equation at the first Canadian clinical laboratory to implement this equation. METHODS: A total of 3161 lipid ultracentrifugation results from a specialized lipid cohort of 2836 patients were included. LDL-C was calculated using the NIH and Friedewald equations and compared to LDL-C measured by ultracentrifugation. We determined the accuracy of these equations at treatment thresholds and developed NIH equation restriction criteria to ensure only accurate results are reported. RESULTS: Ultracentrifugation LDL-C more strongly correlated with NIH-calculated LDL-C (r2 = 0.889) than Friedewald-calculated LDL-C (r2 = 0.807) and resulted in fewer non-sensical negative LDL-C values. The correlation for NIH-calculated LDL-C improved to r2 = 0.975 after applying our restriction criteria. The NIH equation showed equivalent or superior concordance with ultracentrifugation at treatment thresholds. The LDL-C mean absolute difference increased with increasing TG and decreasing LDL-C concentrations, although the NIH equation was more robust under both conditions. CONCLUSIONS: We validated the NIH equation against ultracentrifugation in a cohort with a wide lipid concentration range, which supported its superiority over the Friedewald equation. We recommend clinical implementing the NIH equation for all patients except those with type III hyperlipoproteinemia or TG > 9.04 mmol/L, with an LDL-C lower reporting limit of <0.50 mmol/L.

Bias in IGF-1 concentrations and interpretation across three different clinical laboratory assays.

In this study, we compared the DiaSorin LiaisonXL IGF-1 immunoassay to both the Roche Elecsys IGF-1 immunoassay and to the liquid chromatography-high resolution mass spectrometry (LC-MS) IGF-1 assay. Our study shows a constant positive bias in DiaSorin compared to the Roche immunoassay (mean 42 µg/L, 24%), and a proportional positive bias in DiaSorin compared to the LC-MS method (mean 49 µg/L, 29%). Further, we demonstrate the potential clinical impact of this bias by evaluating 43 adult samples, collected over a 2-month period, which were shown to be discrepant based on a chart review. Despite the positive analytical bias in the Diasorin assay compared to the LC-MS assay, the Diasorin assay upper reference limits were lower than those of the LC-MS assay. This effect caused nine out of forty-three samples to show falsely elevated results when they were clinically diagnosed as negative for acromegaly. Discussed in the context of previous literature, our findings emphasize the importance of adjusting reference intervals for IGF-1 assays based on the clinical needs of a patient population.

Canadian Society of Clinical Chemists Harmonized Clinical Laboratory Lipid Reporting Recommendations on the Basis of the 2021 Canadian Cardiovascular Society Lipid Guidelines.

There is limited guidance on laboratory reporting and interpretation of lipids and lipoproteins used in cardiovascular risk stratification. This contributes to inconsistencies in lipid reporting across clinical laboratories. Recently, the Canadian Cardiovascular Society (CCS) published the 2021 CCS guidelines for the management of dyslipidemia for the prevention of cardiovascular disease in the adult. A subcommittee of the Working Group on Reference Interval Harmonization of the Canadian Society of Clinical Chemists has developed harmonized lipid reporting recommendations that are aligned with the 2021 CCS guidelines, to improve the standardization of lipid assessment and clinical decision-making. The proposed harmonized lipid reporting recommendations were critically reviewed by a broad range of laboratory and clinical experts across Canada. Feedback from approximately 30 expert reviewers was reviewed by the Working Group on Reference Interval Harmonization lipid subcommittee, and consensus decisions were incorporated into the 2021 harmonized lipid reporting recommendations. In this position statement, we provide 6 recommendations for laboratory reporting of lipid parameters. These recommendations include implementing the new National Institutes of Health equation to replace the Friedewald equation for calculating low-density lipoprotein cholesterol, offering lipoprotein (a), either as an in-house or send-out test, and using assays that report lipoprotein (a) in molar units (nmol/L). We also developed a harmonized lipid reporting format with interpretive comments that includes flagging results based on screening patients using treatment decision thresholds in a primary prevention setting. Overall, harmonized lipid reporting will help bridge the gap between clinical guideline recommendations and clinical laboratory reporting and interpretation, and will improve cardiovascular risk assessment across Canada.

Revising Ferritin Lower Limits: It's Time to Raise the Bar on Iron Deficiency.

Ferritin is a key diagnostic marker of iron deficiency (ID), but the interpretative guidance provided to physicians varies significantly. Clear discrepancies exist between clinical guidelines that recommend evidence-based ferritin cutoffs and clinical laboratories that report highly variable ferritin reference intervals (RIs) derived from apparently healthy populations. In this study, clinical laboratories across North America were surveyed to assess the RIs provided with ferritin results. Although clinical guidelines often recommend ferritin cutoffs of 15 or 30 µg/L to identify uncomplicated ID, the survey showed that 18 of 23 responding laboratories reported female RI lower limits well below 15 µg/L. To understand the clinical impact, we analyzed 52 027 unique patient ferritin values over a 5-year period (2013-2017) from a tertiary care hospital. In this population, the 90th percentile ferritin cutoff to identify ID anemia in adults was 24 µg/L in female patients and 25 µg/L in male patients. Distribution of ferritin results in female patients showed that menopausal status had a significant effect on median values, which increased 2- to 3-fold in the postmenopausal state. Furthermore, sorting the data for female patients by physician specialty showed the highest prevalence of low ferritin values in patients seen in obstetrics and gynecology. This study highlights the discrepancy between clinical guidelines and clinical laboratory practice for ferritin reporting and indicates that ferritin RIs, particularly for female patients, are set to an inappropriately low threshold in most clinical laboratories in North America; this level provides good specificity but poor sensitivity when screening for ID.

Reducing free thyroid hormone testing through multiple Plan-Do-Study-Act cycles.

OBJECTIVES: Free thyroid hormones (fT4 and fT3) are one of the most commonly ordered laboratory tests and often ordered when not clinically meaningful. Based on this, many studies have sought to identify strategies to reduce inappropriate fT4 and fT3 testing. The goal of the current study was to implement a quality improvement (QI) framework to identify an optimal approach to reducing inappropriate free thyroid hormone testing through multiple change ideas and Plan-Do-Study-Act (PDSA) cycles. The aim was to reduce fT4 and fT3 30% from baseline at a large tertiary hospital within 12 months. METHODS: The Model for Improvement Framework was used to implement a total of 3 change ideas in the first and second PDSA cycles. Change ideas included implementation and refinement of a free thyroid hormone forced function reflex system, modifications to test requisitions/order-entry interfaces, and a TSH-only option. Process and balancing measures were evaluated to fine-tune the change interventions. Data was continuously monitored pre and post interventions to assess progress, impact and potential errors. RESULTS: In the first PDSA cycle, laboratory testing of fT4 was decreased by 24% and fT3 by 18%. Soliciting physician feedback and assessing balancing measures was important in refining the approach. In the second PDSA cycle, fT4 was decreased by an additional 16% and fT3 by 29%. An audit of the process showed that phone calls to the laboratory to add-on free thyroid hormones did not increase after the second PDSA, averaging 2 calls per month. CONCLUSIONS: To achieve optimal reductions in free thyroid hormone testing, multiple PDSA cycles were required alongside assessing process and balancing measures. Overall, fT4 and fT3 testing was decreased by 39% and 47%, respectively.