Real-world big-data studies in laboratory medicine: Current status, application, and future considerations

Multidisciplinary approach to redefining thyroid hormone reference intervals with big data analysis

OBJECTIVES

This study aimed to employ big data analysis to harmonize reference intervals (RI) for thyroid function tests, with refinement to the TSH upper reference limit, and to optimize the TSH reflex algorithm to improve clinical management and test utilization.

DESIGN & METHODS

TSH, free T4, and free T3 results tested in Alberta, Canada, on Roche Cobas and Siemens Atellica were extracted from the laboratory information system (N = 1,144,155 for TSH, N = 183,354 for free T4 and N = 92,632 for free T3). Results from specialists, inpatients, or repeat testing, as well as from positive thyroid disease, autoimmune disease, and pregnancy biomarkers were excluded. RIs were derived using statistical models (Bhattacharya, refineR, and simple non-parametric) followed by endocrinology and laboratory review.

RESULTS

The TSH RIs for 0 to 7 days, 8 days to 1 year, and ≥1 year were 1.23 to 25.0 mIU/L, 1.00 to 6.80 mIU/L and 0.20 to 6.50 mIU/L, respectively. The free T4 RIs for 0 to 14 days, 15 to 29 days, and ≥30 days were 13.5 to 50.0 pmol/L, 8.7 to 32.5 pmol/L, and 10.0 to 25.0 pmol/L, respectively. An updated TSH reflex algorithm was developed based on the optimized TSH and free T4 RIs, with free T4 reflexed only at a TSH of <0.1 mIU/L.

CONCLUSIONS

The collaboration of a multidisciplinary team and the utilization of big data analysis led to the enhancement of thyroid function RIs, specifically resulting in the widening of the upper TSH reference limit to 6.50. Application of these optimized RIs with the TSH reflex algorithm will serve as a guide for improvement in interpretation of thyroid function tests.

Laboratory Preparation for Digital Medicine in Healthcare 4.0: An Investigation Into the Awareness and Applications of Big Data and Artificial Intelligence

Background

Healthcare 4.0. refers to the integration of advanced technologies, such as artificial intelligence (AI) and big data analysis, into the healthcare sector. Recognizing the impact of Healthcare 4.0 technologies in laboratory medicine (LM), we seek to assess the overall awareness and implementation of Healthcare 4.0 among members of the Korean Society for Laboratory Medicine (KSLM).

Methods

A web-based survey was conducted using an anonymous questionnaire. The survey comprised 36 questions covering demographic information (seven questions), big data (10 questions), and AI (19 questions).

Results

In total, 182 (17.9%) of 1,017 KSLM members participated in the survey. Thirty-two percent of respondents considered AI to be the most important technology in LM in the era of Healthcare 4.0, closely followed by 31% who favored big data. Approximately 80% of respondents were familiar with big data but had not conducted research using it, and 71% were willing to participate in future big data research conducted by the KSLM. Respondents viewed AI as the most valuable tool in molecular genetics within various divisions. More than half of the respondents were open to the notion of using AI as assistance rather than a complete replacement for their roles.

Conclusions

This survey highlighted KSLM members' awareness of the potential applications and implications of big data and AI. We emphasize the complexity of AI integration in healthcare, citing technical and ethical challenges leading to diverse opinions on its impact on employment and training. This highlights the need for a holistic approach to adopting new technologies.

Automatic kidney stone identification: an adaptive feature-weighted LSTM model based on urine and blood routine analysis

Kidney stones are the most common urinary system diseases, and early identification is of great significance. The purpose of this study was to use routine urine and blood detection indices to build a deep learning (DL) model to identify the presence of kidney stones in the early stage. A retrospective analysis was conducted on patients with kidney stones who were treated at West China Hospital of Sichuan University from January 2020 to June 2023. A total of 1130 individuals presenting with kidney stones and 1230 healthy subjects were enrolled. The first blood and urine laboratory data of participants at our hospital were collected, and the data were divided into a training dataset (80%) and a verification dataset (20%). Additionally, a long short-term memory (LSTM)-based adaptive feature weighting model was trained for the early identification of kidney stones, and the results were compared with those of other models. The performance of the model was evaluated by the area under the subject working characteristic curve (AUC). The important predictive factors are determined by ranking the characteristic importance of the predictive factors. A total of 17 variables were screened; among the top 4 characteristics according to the weight coefficient in this model, urine WBC, urine occult blood, qualitative urinary protein, and microcyte percentage had high predictive value for kidney stones in patients. The accuracy of the kidney stone (KS-LSTM) learning model was 89.5%, and the AUC was 0.95. Compared with other models, it has better performance. The results show that the KS-LSTM model based on routine urine and blood tests can accurately identify the presence of kidney stones. And provide valuable assistance for clinicians to identify kidney stones in the early stage.

Establishment and evaluation of Voting algorithm-based internal quality control (ViQC), a patient-based real-time quality control

BACKGROUND

Patient-Based Real-Time Quality Control (PBRTQC) has emerged as a supplementary programme to traditional internal quality control (iQC) mechanisms. Despite its growing popularity, practical applications in clinical settings reveal several challenges. The primary objective of this research is to introduce and develop an Artificial Intelligence (AI)-based method, named Voting algorithm based iQC (ViQC), designed to enhance the precision and reliability of existing PBRTQC systems.

METHODS

In this study, we conducted a retrospective analysis of 111,925 inpatient serum glucose test results from Nanjing Drum Tower Hospital, Nanjing, China, to provide an unbiased data set. The Voting iQC (ViQC) algorithm, established by the principles of the Voting algorithm, was then developed. Its analytical performance was evaluated through the calculation of random errors (RE). Subsequently, its clinical efficacy was assessed by comparison with five statistical algorithms: Moving Average (MA), Exponentially Weighted Moving Average (EWMA), Moving Median (movMed, MM), Moving Quartile (MQ), and Moving Standard Deviation (MovSD).

RESULTS

The ViQC model incorporates a variety of machine learning models, including logistic regression, Bayesian methods, K-Nearest Neighbor, decision trees, random forests, and gradient boosting decision trees, to establish a robust predictive framework. This model consistently maintains a false positive rate below 0.002 across all six evaluated error factors, showcasing exceptional precision. Notably, its performance further excels with an error factor of 3.0, where the false positive rate drops below 0.001, and achieves an accuracy rate as high as 0.965 at an error factor of 2.0. The classification effectiveness of ViQC model is evaluated by an area under the curve (AUC) exceeding 0.97 for all error factors. In comparison to five conventional PBRTQC statistical methods, ViQC significantly enhances error detection efficiency, maximum reducing the trimmed average number of patient samples required for detecting errors from 724 to 168, thereby affirming its superior error detection capability.

CONCLUSION

The new established PBRTQC using artificial intelligence yielded satisfactory performance compared to the traditional PBBTQC in real world setting.

Development of next-generation reference interval models to establish reference intervals based on medical data: current status, algorithms and future consideration

Evidence derived from laboratory medicine plays a pivotal role in the diagnosis, treatment monitoring, and prognosis of various diseases. Reference intervals (RIs) are indispensable tools for assessing test results. The accuracy of clinical decision-making relies directly on the appropriateness of RIs. With the increase in real-world studies and advances in computational power, there has been increased interest in establishing RIs using big data. This approach has demonstrated cost-effectiveness and applicability across diverse scenarios, thereby enhancing the overall suitability of the RI to a certain extent. However, challenges persist when tests results are influenced by age and sex. Reliance on a single RI or a grouping of RIs based on age and sex can lead to erroneous interpretation of results with significant implications for clinical decision-making. To address this issue, the development of next generation of reference interval models has arisen at an historic moment. Such models establish a curve relationship to derive continuously changing reference intervals for test results across different age and sex categories. By automatically selecting appropriate RIs based on the age and sex of patients during result interpretation, this approach facilitates clinical decision-making and enhances disease diagnosis/treatment as well as health management practices. Development of next-generation reference interval models use direct or indirect sampling techniques to select reference individuals and then employed curve fitting methods such as splines, polynomial regression and others to establish continuous models. In light of these studies, several observations can be made: Firstly, to date, limited interest has been shown in developing next-generation reference interval models, with only a few models currently available. Secondly, there are a wide range of methods and algorithms for constructing such models, and their diversity may lead to confusion. Thirdly, the process of constructing next-generation reference interval models can be complex, particularly when employing indirect sampling techniques. At present, normative documents pertaining to the development of next-generation reference interval models are lacking. In summary, this review aims to provide an overview of the current state of development of next-generation reference interval models by defining them, highlighting inherent advantages, and addressing existing challenges. It also describes the process, advanced algorithms for model building, the tools required and the diagnosis and validation of models. Additionally, a discussion on the prospects of utilizing big data for developing next-generation reference interval models is presented. The ultimate objective is to equip clinical laboratories with the theoretical framework and practical tools necessary for developing and optimizing next-generation reference interval models to establish next-generation reference intervals while enhancing the use of medical data resources to facilitate precision medicine.

Integrating Patient-Based Real-Time Quality Control (PBRTQC) in a New Field: Inter-Comparison between Biochemical Instrumentations with LDL-C

BACKGROUND

Patient-based real-time quality control (PBRTQC) can be a valuable tool in clinical laboratories due to its cost-effectiveness and constant monitoring. More focus is placed on discovering and improving algorithms that compliment conventional internal control techniques. The practical implementation of PBRTQC with a biochemical instrument comparison is lacking. We aim to evaluate PBRTQC's efficacy and practicality by comparing low-density lipoprotein cholesterol (LDL-C) test results to ensure consistent real-time monitoring across biochemical instrumentations in clinical laboratories.

METHOD

From 1 September 2021 to 30 August 2022, the First Affiliated Hospital of Xi'an Jiaotong University collected data from 158,259 both healthy and diseased patients, including 84,187 male and 74,072 female patients, and examined their LDL-C results. This dataset encompassed a group comprising 50,556 individuals undergoing health examinations, a group comprising 42,472 inpatients (IP), and a group comprising 75,490 outpatients (OP) for the PBRTQC intelligent monitoring platform to conduct daily tests, parameter configuration, program development, real-time execution, and performance validation of the patients' data. Moreover 40 patients' LDL-C levels were assessed using two biochemical analyzers, designated as the reference and comparator instruments. A total of 160 LDL-C results were obtained from 40 both healthy and diseased patients, including 14 OP, 16 IP, and 10 health examination attendees, who were selected to represent LDL-C levels broadly. Two biochemical instruments measured LDL-C measurements from the same individuals to investigate consistency and reproducibility across patient statuses and settings. We employed exponentially weighted moving average (EWMA) and moving median (MM) methods to calculate inter-instrument bias and ensure analytical accuracy. Inter-instrument bias for LDL-C measurements was determined by analyzing fresh serum samples, different concentrations of quality control (QC), and commercialized calibrators, employing both EWMA and MM within two assay systems. The assessment of inter-instrumental bias with five different methods adhered to the external quality assessment standards of the Clinical Laboratory Center of the Health Planning Commission, which mandates a bias within ±15.0%.

RESULT

We calculated inter-instrument comparison bias with each of the five methods based on patient big data. The comparison of fresh serum samples, different concentrations of QC, commercialized calibrators, and EWMA were all in the permissive range, except for MM. MM showed that the bias between two biochemical instruments in the concentration ranges of 1.5 mmoL/L-6.2 mmoL/L exceeded the permissible range. This was mainly due to the small number of specimens, affected by variations among individual patients, leading to increased false alarms and reduced effectiveness in monitoring the consistency of the inter-instrumental results. Moreover, the inter-comparison bias derived from EWMA was less than 3.01%, meeting the 15% range assessment criteria. The bias result for MM was lower than 24.66%, which was much higher than EWMA. Thus, EWMA is better than MM for monitoring inter-instrument comparability. PBRTQC can complement the use of inter-comparison bias between biochemical analyzers. EWMA has comparable inter-instrument comparability monitoring efficacy.

CONCLUSIONS

The utilization of AI-based PBRTQC enables the automated real-time comparison of test results across different biochemical instruments, leading to a reduction in laboratory operating costs, enhanced work efficiency, and improved QC. This advanced technology facilitates seamless data integration and analysis, ultimately contributing to a more streamlined and efficient laboratory workflow in the biomedical field.

Big data-based parathyroid hormone (PTH) values emphasize need for age correction

PURPOSE

We aimed to study the relationship between aging and increased parathyroid hormone (PTH) values.

METHODS

We performed a retrospective cross-sectional study with data from patients who underwent outpatient PTH measurements performed by a second-generation electrochemiluminescence immunoassay. We included patients over 18 years of age with simultaneous PTH, calcium, and creatinine measurements and 25-OHD measured within 30 days. Patients with glomerular filtration rate < 60 mL/min/1.73 m2, altered calcemia, 25-OHD level < 20 ng/mL, PTH values > 100 pg/mL or using lithium, furosemide or antiresorptive therapy were excluded. Statistical analyses were performed using the RefineR method.

RESULTS

Our sample comprised 263,242 patients for the group with 25-OHD ≥ 20 ng/mL, that included 160,660 with 25-OHD ≥ 30 ng/mL. The difference in PTH values among age groups divided by decades was statistically significant (p < 0.0001), regardless of 25-OHD values, ≥ 20 or ≥ 30 ng/mL. In the group with 25-OHD ≥ 20 ng/mL and more than 60 years, the PTH values were 22.1-84.0 pg/mL, a different upper reference limit from the reference value recommended by the kit manufacturer.

CONCLUSION

We observed a correlation between aging and PTH increase, when measured by a second-generation immunoassay, regardless of vitamin D levels, if greater than 20 ng/mL, in normocalcemic individuals without renal dysfunction.

Establishment of a reference interval for total carbon dioxide using indirect methods in Chinese populations living in high-altitude areas: A retrospective real-world analysis

BACKGROUND

Hypoxia leads to different concentrations of the bicarbonate buffer system in Tibetan people. Indirect methods were used to establish the reference interval (RI) for total carbon dioxide (tCO2) based on big data from the adult population of Tibet, a high-altitude area in Western China.

METHODS

Anonymous tCO2 test data (n = 442,714) were collected from the People's Hospital of the Tibet Autonomous Region from January 2018, to December 2021. Multiple linear regression and variance component analyses were performed to assess the effects of sex, age, and race on tCO2 levels. Indirect methods, including Hoffmann, Bhattacharya, expectation maximization (EM), kosmic and refineR, were used to calculate the total RI and ethnicity-partitioned RI.

RESULTS

A total of 230,821 real-world tCO2 test results were eligible. Sex, age, and race were significantly associated with the tCO2 levels. The total and ethnically-partitioned RIs estimated using the five indirect methods were comparable. The total RI of tCO2 was 14-24 mmol/L (calculated using Hoffmann and refineR) and 15-24 mmol/L (Bhattacharya, EM and kosmic). For Han nationality, the RIs were 14-25 mmol/L (calculated using Hoffmann and Bhattacharya), 16-23 mmol/L (EM), 15-24 mmol/L (kosmic), and 14.2-24.5 mmol/L (refineR). For the Tibetan population, the RIs were 14-24 mmol/L (calculated using Hoffmann and refineR), 15-24 mmol/L (Bhattacharya and kosmic), and 15-23 mmol/L (EM). The established RIs were significantly lower than those living at lower altitudes area (22-29 mmol/L) that was provided by the manufacturer.

CONCLUSION

The tCO2 RI of the populations living on the Tibetan Plateau was significantly lower than those at the lower altitudes. The RIs established using indirect methods are suitable for clinical applications in Tibet.

Current status and challenges in establishing reference intervals based on real-world data

Reference intervals (RIs) are the cornerstone for evaluation of test results in clinical practice and are invaluable in judging patient health and making clinical decisions. Establishing RIs based on clinical laboratory data is a branch of real-world data mining research. Compared to the traditional direct method, this indirect approach is highly practical, widely applicable, and low-cost. Improving the accuracy of RIs requires not only the collection of sufficient data and the use of correct statistical methods, but also proper stratification of heterogeneous subpopulations. This includes the establishment of age-specific RIs and taking into account other characteristics of reference individuals. Although there are many studies on establishing RIs by indirect methods, it is still very difficult for laboratories to select appropriate statistical methods due to the lack of formal guidelines. This review describes the application of real-world data and an approach for establishing indirect reference intervals (iRIs). We summarize the processes for establishing iRIs using real-world data and analyze the principle and applicable scope of the indirect method model in detail. Moreover, we compare different methods for constructing growth curves to establish age-specific RIs, in hopes of providing laboratories with a reference for establishing specific iRIs and giving new insight into clinical laboratory RI research. (201 words).

Utilization of five data mining algorithms combined with simplified preprocessing to establish reference intervals of thyroid-related hormones for non-elderly adults

BACKGROUND

Despite the extensive research on data mining algorithms, there is still a lack of a standard protocol to evaluate the performance of the existing algorithms. Therefore, the study aims to provide a novel procedure that combines data mining algorithms and simplified preprocessing to establish reference intervals (RIs), with the performance of five algorithms assessed objectively as well.

METHODS

Two data sets were derived from the population undergoing a physical examination. Hoffmann, Bhattacharya, Expectation Maximum (EM), kosmic, and refineR algorithms combined with two-step data preprocessing respectively were implemented in the Test data set to establish RIs for thyroid-related hormones. Algorithm-calculated RIs were compared with the standard RIs calculated from the Reference data set in which reference individuals were selected following strict inclusion and exclusion criteria. Objective assessment of the methods is implemented by the bias ratio (BR) matrix.

RESULTS

RIs of thyroid-related hormones are established. There is a high consistency between TSH RIs established by the EM algorithm and the standard TSH RIs (BR = 0.063), although EM algorithms seems to perform poor on other hormones. RIs calculated by Hoffmann, Bhattacharya, and refineR methods for free and total triiodo-thyronine, free and total thyroxine respectively are close and match the standard RIs.

CONCLUSION

An effective approach for objectively evaluating the performance of the algorithm based on the BR matrix is established. EM algorithm combined with simplified preprocessing can handle data with significant skewness, but its performance is limited in other scenarios. The other four algorithms perform well for data with Gaussian or near-Gaussian distribution. Using the appropriate algorithm based on the data distribution characteristics is recommended.

Disruption vs. evolution in laboratory medicine. Current challenges and possible strategies, making laboratories and the laboratory specialist profession fit for the future

Since beginning of medical diagnostics, laboratory specialists have done an amazing job, continuously improving quality, spectrum and speed of laboratory tests, currently contributing to the majority of medical decision making. These improvements are mostly of an incremental evolutionary fashion, meaning improvements of current processes. Sometimes these evolutionary innovations are of a radical fashion, such as the invention of automated analyzers replacing manual testing or the implementation of mass spectrometry, leading to one big performance leap instead of several small ones. In few cases innovations may be of disruptive nature. In laboratory medicine this would be applicable to digitalization of medicine or the decoding of the human genetic material. Currently, laboratory medicine is again facing disruptive innovations or technologies, which need to be adapted to as soon as possible. One of the major disruptive technologies is the increasing availability and medical use of artificial intelligence. It is necessary to rethink the position of the laboratory specialist within healthcare settings and the added value he or she can provide to patient care. The future of the laboratory specialist profession is bright, as it the only medical profession comprising such vast experience in patient diagnostics. However, laboratory specialists need to develop strategies to provide this expertise, by adopting to the quickly evolving technologies and demands. This opinion paper summarizes some of the disruptive technologies as well as strategies to secure and/or improve the quality of diagnostic patient care and the laboratory specialist profession.

Improved implementation of medical tests - barriers and opportunities

Applying the concept of a value proposition to medical testing is just one of the many ways to identify and monitor the value of tests. A key part of this concept focusses on processes that should take place after a test is introduced into routine local practice, namely test implementation. This process requires identification of the clinical pathway, the stakeholders and the benefits or disbenefits that accrue to those stakeholders. There are various barriers and challenges to test implementation. Implementation requires the process of clinical audit which involves measurement of outcomes external to the laboratory but this is not widely performed in laboratory medicine. A second key challenge is that implementation requires liaison with stakeholders outside of the laboratory including clinicians and other healthcare professional such as finance managers. Many laboratories are remote from clinical care and other stakeholders making such liaison difficult. The implementation process is based on data which again will be primarily on processes outside of the laboratory. However the recent enthusiasm for so-called real world data and new data mining techniques may represent opportunities that will facilitate better test implementation. A final barrier is that a range of new tools not currently in the education curriculum of the laboratory professional is required for implementation such as those of preparing a business case to support the introduction of a test and health economic analysis. The professional bodies in laboratory medicine could assist with education in these areas.

Automated Interlaboratory Comparison of Therapeutic Drug Monitoring Data and Its Use for Evaluation of Published Therapeutic Reference Ranges

Therapeutic drug monitoring is a tool for optimising the pharmacological treatment of diseases where the therapeutic effect is difficult to measure or monitor. Therapeutic reference ranges and dose-effect relation are the main requirements for this drug titration tool. Defining and updating therapeutic reference ranges are difficult, and there is no standardised method for the calculation and clinical qualification of these. The study presents a basic model for validating and selecting routine laboratory data. The programmed algorithm was applied on data sets of antidepressants and antipsychotics from three public hospitals in Denmark. Therapeutic analytical ranges were compared with the published therapeutic reference ranges by the Arbeitsgemeinschaft für Neuropsychopharmakologie und Pharmakopsychiatrie (AGNP) and in additional literature. For most of the drugs, the calculated therapeutic analytical ranges showed good concordance between the laboratories and to published therapeutic reference ranges. The exceptions were flupentixol, haloperidol, paroxetine, perphenazine, and venlafaxine + o-desmethyl-venlafaxine (total plasma concentration), where the range was considerably higher for the laboratory data, while the calculated range of desipramine, sertraline, ziprasidone, and zuclopenthixol was considerably lower. In most cases, we identified additional literature supporting our data, highlighting the need of a critical re-examination of current therapeutic reference ranges in Denmark. An automated approach can aid in the evaluation of current and future therapeutic reference ranges by providing additional information based on big data from multiple laboratories.

Comparison of age- and sex-dependent reference limits derived from distinct sources for metabolic measurands in basic liver diagnostics

BACKGROUND

Reference intervals for basic liver laboratory diagnostic rely on manufacturers' information, remaining unchanged for more than 20 years. This ignores known age and sex dependencies.

METHODS

We performed a retrospective cross-sectional study to compare the age-dependent distribution of flagged and non-flagged laboratory findings between reference limits from 3 distinct sources: manufacturer, published reference study, and the truncated maximum likelihood method applied on a cohort of inpatients aged 18-100 years. Discordance rates adjusted for the permissible analytical uncertainty are reported for serum levels of albumin (n= 150,550), alkaline phosphatase (n= 433,721), gamma-GT (n=580,012), AST (n= 510,620), and ALT (n= 704,546).

RESULTS

The number of flagged findings differed notably between reference intervals compared, except for alkaline phosphatase. AST and alkaline phosphatase increased with age in women. Overall discordance for AP, AST, and ALT remained below 10%, respectively, in both sexes. Albumin decreased with age which led to discordant flags in up to 22% in patients ≥70 years. GGT and ALT peaked in 50-59-year-old men with up to 23.5% and 22.8% discordant flags, respectively.

CONCLUSION

We assessed the impact of different reference limits on liver related laboratory results and found up to 25 % discordant flags. We suggest to further analyse the diagnostic and economic effects of reference limits adapted to the population of interest even for well-established basic liver diagnostics.

RIbench: A Proposed Benchmark for the Standardized Evaluation of Indirect Methods for Reference Interval Estimation

BACKGROUND

Indirect methods leverage real-world data for the estimation of reference intervals. These constitute an active field of research, and several methods have been developed recently. So far, no standardized tool for evaluation and comparison of indirect methods exists.

METHODS

We provide RIbench, a benchmarking suite for quantitative evaluation of any existing or novel indirect method. The benchmark contains simulated test sets for 10 biomarkers mimicking routine measurements of a mixed distribution of non-pathological (reference) values and pathological values. The non-pathological distributions represent 4 common distribution types: normal, skewed, heavily skewed, and skewed-and-shifted. To identify strengths and weaknesses of indirect methods, test sets have varying sample sizes and pathological distributions differ in location, extent of overlap, and fraction. For performance evaluation, we use an overall benchmark score and sub-scores derived from absolute z-score deviations between estimated and true reference limits. We illustrate the application of RIbench by evaluating and comparing the Hoffmann method and 4 modern indirect methods -TML (Truncated-Maximum-Likelihood), kosmic, TMC (Truncated-Minimum-Chi-Square), and refineR- against one another and against a nonparametric direct method (n = 120).

RESULTS

For the modern indirect methods, pathological fraction and sample size had a strong influence on the results: With a pathological fraction up to 20% and a minimum sample size of 5000, most methods achieved results comparable or superior to the direct method.

CONCLUSIONS

We present RIbench, an open-source R-package, for the systematic evaluation of existing and novel indirect methods. RIbench can serve as a tool for enhancement of indirect methods, improving the estimation of reference intervals.

An innovative approach based on real-world big data mining for calculating the sample size of the reference interval established using transformed parametric and non-parametric methods

Background

Currently, the direct method is the main approach for establishment of reference interval (RI). However, only a handful of studies have described the effects of sample size on establishment of RI and estimation of sample size. We describe a novel approach for estimation of the sample size when establishing RIs using the transformed parametric and non-parametric methods.

Methods

A total of 3,697 healthy participants were enrolled in this study. We adopted a two-layer nested loop sample size estimation method to determine the effects of sample size on RI, using thyroid-related hormone as an example. The sample size was selected as the calculation result when the width of the confidence interval (CI) of the upper and lower limit of the RI were both stably < 0.2 times the width of RI. Then, we calculated the sample size for establishing RIs via transformed parametric and non-parametric methods for thyroid-related hormones.

Results

Sample sizes for thyroid stimulating hormone (TSH), as required by parametric and non-parametric methods to establish RIs were 239 and 850, respectively. Sample sizes required by the transformed parametric method for free triiodothyronine (FT3), free thyroxine (FT4), total triiodothyronine (TT3) and total thyroxine (TT4) were all less than 120, while those required by the non-parametric method were more than 120.

Conclusion

We describe a novel approach for estimating sample sizes for establishment of RI. A corresponding open-source code has been developed and is available for applications. The established method is suitable for most analytes, with evidence based on thyroid-related hormones indicating that different sample sizes are required to establish RIs using different methods for analytes with different variations.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12874-022-01751-1.

Clinlabomics: leveraging clinical laboratory data by data mining strategies

The recent global focus on big data in medicine has been associated with the rise of artificial intelligence (AI) in diagnosis and decision-making following recent advances in computer technology. Up to now, AI has been applied to various aspects of medicine, including disease diagnosis, surveillance, treatment, predicting future risk, targeted interventions and understanding of the disease. There have been plenty of successful examples in medicine of using big data, such as radiology and pathology, ophthalmology cardiology and surgery. Combining medicine and AI has become a powerful tool to change health care, and even to change the nature of disease screening in clinical diagnosis. As all we know, clinical laboratories produce large amounts of testing data every day and the clinical laboratory data combined with AI may establish a new diagnosis and treatment has attracted wide attention. At present, a new concept of radiomics has been created for imaging data combined with AI, but a new definition of clinical laboratory data combined with AI has lacked so that many studies in this field cannot be accurately classified. Therefore, we propose a new concept of clinical laboratory omics (Clinlabomics) by combining clinical laboratory medicine and AI. Clinlabomics can use high-throughput methods to extract large amounts of feature data from blood, body fluids, secretions, excreta, and cast clinical laboratory test data. Then using the data statistics, machine learning, and other methods to read more undiscovered information. In this review, we have summarized the application of clinical laboratory data combined with AI in medical fields. Undeniable, the application of Clinlabomics is a method that can assist many fields of medicine but still requires further validation in a multi-center environment and laboratory.

Within-subject biological variation estimates using an indirect data mining strategy. Spanish multicenter pilot study (BiVaBiDa)

OBJECTIVES

The estimates of biological variation (BV) have traditionally been determined using direct methods, which present limitations. In response to this issue, two papers have been published addressing these limitations by employing indirect methods. Here, we present a new procedure, based on indirect methods that analyses data collected within a multicenter pilot study. Using this method, we obtain CV_I estimates and calculate confidence intervals (CI), using the EFLM-BVD CV_I estimates as gold standard for comparison.

METHODS

Data were collected over a 18-month period for 7 measurands, from 3 Spanish hospitals; inclusion criteria: patients 18-75 years with more than two determinations. For each measurand, four different strategies were carried out based on the coefficient of variation ratio (rCoeV) and based on the use of the bootstrap method (OS1, RS2 and RS3). RS2 and RS3 use symmetry reference change value (RCV) to clean database.

RESULTS

RS2 and RS3 had the best correlation for the CV_I estimates with respect to EFLM-BVD. RS2 used the symmetric RCV value without eliminating outliers, while RS3 combined RCV and outliers. When using the rCoeV and OS1 strategies, an overestimation of the CV_I value was obtained.

CONCLUSIONS

Our study presents a new strategy for obtaining robust CV_I estimates using an indirect method together with the value of symmetric RCV to select the target population. The CV_I estimates obtained show a good correlation with those published in the EFLM-BVD database. Furthermore, our strategy can resolve some of the limitations encountered when using direct methods such as calculating confidence intervals.

Traceable machine learning real-time quality control based on patient data

Objectives

Patient-based real-time quality control (PBRTQC) has gained attention as an alternative/integrative tool for internal quality control (iQC). However, it is still doubted for its performance and its application in real clinical settings. We aim to generate a newly and easy-to-access patient-based real-time QC by machine learning (ML) traceable to standard reference data with assigned values by National Institute of Metrology of China (NIM), and to compare it with PBRTQC for clinical validity evaluation.

Methods

For five representative biochemistry analytes, 1,195 000 patient testing results each were collected. After data processing, independent training and test sets were divided. Machine learning internal quality control (MLiQC) was set up by Random Forest in ML and was validated by way of both metrology algorithm traceability and 4 PBRTQC methods recommended by IFCC analytical working group.

Results

MLiQC were established. As an example of albumin (ALB) at the critical bias, the uncertainty of MLiQC was 0.14%, which was evaluated by standard reference data produced by NIM. Compared with four optimal PBRTQC methods at critical bias, the average of the number of patient samples from a bias introduced until detected (ANPed) of MLiQC averagely decreased from 600 to 20. The median and 95 quantiles of NPeds (MNPed and 95NPed) of MLiQC were superior to all optimal PBRTQCs above 90% for all test items.

Conclusions

MLiQC is highly superior to PBRTQC and well-suited in real settings. The validation of the model from two aspects of algorithm traceability and clinical effectiveness confirms its satisfactory performance.

Establishment of Reference Interval and Aging Model of Homocysteine Using Real-World Data

Objective

The level of Homocysteine (Hcy) in males is generally higher than that of females, but the same reference interval (RI) is often used in clinical practice. This study aims to establish a sex-specific RI of Hcy using five data mining algorithms and compare these results. Furthermore, age-related continuous RI was established in order to show the relationship between Hcy concentration distribution and age.

Methods

A total of 20,801 individuals were included in the study and Tukey method was used to identify outliers in subgroups by sex and age. Multiple linear regression and standard deviation ratio (SDR) was used to determine whether the RI for Hcy needs to be divided by sex and age. Five algorithms including Hoffmann, Bhattacharya, expectation maximization (EM), kosmic and refineR were utilized to establish the RI of Hcy. Generalized Additive Models for Location Scale and Shape (GAMLSS) algorithm was used to determine the aging model of Hcy and calculate the age-related continuous RI.

Results

RI of Hcy needed to be partitioned by sex (SDR = 0.735 > 0.375). RIs established by Hoffmann, Bhattacharya, EM (for females) and kosmic are all within the 95% CI of reference limits established by refine R. The Sex-specific aging model of Hcy showed that the upper limits of the RI of Hcy declined with age beginning at age of 18 and began to rise approximately after age of 40 for females and increased with age for males.

Conclusion

The RI of Hcy needs to be partitioned by sex. The RIs established by the five data mining algorithms showed good consistency. The dynamic sex and age-specific model of Hcy showed the pattern of Hcy concentration with age and provide more personalized tools for clinical decisions.

Pandemic Response in the Clinical Laboratory: The Utility of Interactive Dashboards

The ability to access and analyze data is critical to manage a laboratory and respond and adapt to changes, particularly during a pandemic. Data analytic tools can not only improve laboratory operations, but also increase the visibility of the laboratory in the healthcare system and demonstrate the positive impact of the laboratory on patient care. In this article, we describe the creation and utility of laboratory dashboards. Several dashboards were designed to assist with pandemic response. For each dashboard, a stored procedure was created that performed a SQL query of our laboratory information system mirror database. We utilized the business analytics platform, Tableau, for data visualization. Users could modify the data by selecting a specific date range, time window, work shift, institution(s), specific test(s), and/or testing platform(s). Access was controlled by OKTA integration to the host server over the web, behind the hospital firewall. During the April 2020 surge, we saw an increase in blood gas testing and corresponding decrease in non-critical testing such as Vitamin D. At our institution, SARS-CoV-2 molecular testing was performed using four primary platforms, four in-house and one send-out. Weekly and hourly testing volumes as well as turnaround times fluctuated based on reagent availability, new testing requests, staffing, and operational changes. Productivity dashboards indicated that coagulation testing volumes were highest on the third shift and that all three analyzers may not be necessary. Further, specimen volumes and productivity of accessioning staff varied throughout the day. Phlebotomy venipuncture volumes and patient wait times also varied throughout the pandemic. A decrease in ambulatory draws was seen during the surge but after reopening draw volumes, particularly at offsite locations, surpassed prepandemic volumes. We demonstrate that data analytics and interactive dashboards are powerful tools, are helpful in response to a pandemic and lead to improved TAT, supply utilization, staffing and workflows. Furthermore, dashboards provide objective data to review with hospital leadership and promote collaboration.

Establishment of Reference Intervals for Thyroid-Associated Hormones Using refineR Algorithm in Chinese Population at High-Altitude Areas

Objectives

Diagnosis of thyroid disease among individuals dwelling at high altitude remains a challenge. Reference intervals (RIs) for thyroid-associated hormones among Tibetans living at various high altitudes were established to improve diagnosis.

Methods

One thousand two hundred eighty-one subjects were randomly recruited from Nyingchi, Shigatse/Lhasa, and Ali of Tibet. Thyroid-stimulating hormone (TSH), free triiodothyronine (FT3), and free thyroxine (FT4) were measured by the Cobas e601 electrochemiluminescence analyzer. We used multiple linear regression and variance component analysis to assess the effect of sex, age, and altitude on hormones. RIs were established by refineR algorithm and compared with those provided by the manufacturer.

Results

Serum TSH was significantly lower in males than in females, while FT3 and FT4 were higher in males. Both FT3 and FT4 decreased with increasing age. FT3 increased with altitude, while TSH and FT4 were less influenced by altitude. The RI for TSH was 0.764–5.784 μIU/ml, while for FT4, the RIs were 12.36–19.38 pmol/L in females and 14.84–20.18 pmol/L in males. The RIs for FT3 at Nyingchi, Shigatse/Lhasa, and Ali in females were 4.09–4.98, 4.31–5.45, and 4.82–5.58 pmol/L, while in males, the values were 4.82–5.41, 4.88–5.95, and 5.26–6.06 pmol/L, respectively. The obtained RIs for TSH and FT4 were generally higher, while that for FT3 was narrower than the RIs provided by Cobas.

Conclusions

Specific RIs were established for thyroid-associated hormones among Tibetans, which were significantly different from those provided by the manufacturer.

Big data and reference intervals

Although reference intervals (RIs) play an important role in clinical diagnosis, there remain significant differences with respect to race, gender, age and geographic location. Accordingly, the Clinical Laboratory Standards Institute (CLSI) EP28-A3c has recommended that clinical laboratories establish RIs appropriate to their subject population. Unfortunately, the traditional and direct approach to establish RIs relies on the recruitment of a sufficient number of healthy individuals of various age groups, collection and testing of large numbers of specimens and accurate data interpretation. The advent of the big data era has, however, created a unique opportunity to "mine" laboratory information. Unfortunately, this indirect method lacks standardization, consensus support and CLSI guidance. In this review we provide a historical perspective, comprehensively assess data processing and statistical methods, and post-verification analysis to validate this big data approach in establishing laboratory specific RIs.

Establishment of reference intervals of ten commonly used clinical chemistry analytes: a real-world study in China

Aim: This real-world study was aimed at establishing reference intervals (RIs) of ten commonly used clinical chemistry analytes (total cholesterol, triglycerides, Apo A1, Apo B, creatine kinase (CK), CK isoenzyme MB, glucose, alkaline phosphatase, γ-glutamyltransferase and blood urea nitrogen) in an apparently healthy population in China. Materials & methods: A total of 17,356 healthy participants aged 18-79 years who underwent check-up at MJ Health Check-up Center were included. The establishment of RIs was performed according to the Clinical and Laboratory Standards Institute EP28-A3c guideline. Roche Cobas c701 automatic analyzer (Roche Diagnostics, Mannheim, Germany) was employed to measure the concentrations of analytes. Results: Total cholesterol, triglycerides, Apo B, CK, alkaline phosphatase, glucose, γ-glutamyltransferase and blood urea nitrogen required gender and age-specific partitioning. Conclusion: The RIs established in this study were parallel to current national standards and previous RIs established in Chinese population. Real-world studies may play an important and practical role in the determination of RIs in the future.

Effect of sample size and the traditional parametric, nonparametric, and robust methods on the establishment of reference intervals: Evidence from real world data

Sample size and statistical methods are critical for establishing reference intervals (RIs) but they tend to be overlooked. In this study, we used R (3.6.3) to stratify the reference individuals by sex, and then stratified them using the random sampling method. Fourteen sub-data sets with a sample size of 40, 80, 120, 160, 200, 500, 800, 1000, 1500, 2000, 2500, 3000, 3500, and 4000 were extracted, respectively. The sex ratios of all sub-data sets were 1:1. Transformed parametric (using log transformation), nonparametric, and robust approaches as described in the Clinical and Laboratory Standards Institute guidelines were adopted to establish the RIs and the 90% confidence interval of the thyroid-stimulating hormone (TSH) using data from the sub-data sets. The Bland-Altman plot was used to evaluate the consistency of the upper and lower limits of the RIs established using the three methods. The upper and lower limits of TSH RI tended to be stable starting from the data set with a sample size of 1500. The RIs established using the three methods were more consistent when using a sample size greater than or equal to 2000.