Estimation Of Reference Intervals Research Articles

Accounting for differences between serum and plasma: An indirect approach to derive reference intervals for total protein, albumin, and globulin

BackgroundObservable quantitative variations exist between plasma and serum in routine protein measurements, often not reflected in standard reference intervals. In this study, we describe an indirect approach for estimating a combined reference interval (RI) (i.e., serum and plasma), for commonly ordered protein measurands: total protein, albumin, and globulin. MethodsWe applied an indirect reference interval estimation for protein measurements in serum and plasma using data from July 2018 to February 2024. The data were divided into three Epochs based on a period of plasma separator tube shortage during the COVID-19 pandemic. Bootstrap resampling was used to calculate RIs and corresponding 95% confidence intervals for each month. ResultsOur results demonstrate notable changes in RI limits for total protein, albumin, and globulin between Epochs, reflecting the influence of changing sample matrix. A combined RI was identified for all components and verified using plasma and serum samples from 20 healthy individuals and retrospective analysis of flagging rates on our outpatient population using new and historical RIs. ConclusionThe study demonstrates notable differences in the RIs for total protein, albumin, and globulin when container type changes. In addition, the results demonstrate the effectiveness of big data analytics in deriving RIs and highlights the necessity of continuous RI assessment and adjustment based on the patient population and acceptable specimen types.

The performance of unsupervised machine learning method in the indirect reference interval estimation of selected biochemical analytes

The performance of unsupervised machine learning method in the indirect reference interval estimation of selected biochemical analytes

An Objective Function-Based Clustering Algorithm with a Closed-Form Solution and Application to Reference Interval Estimation in Laboratory Medicine

Clustering algorithms are usually iterative procedures. In particular, when the clustering algorithm aims to optimise an objective function like in k-means clustering or Gaussian mixture models, iterative heuristics are required due to the high non-linearity of the objective function. This implies higher computational costs and the risk of finding only a local optimum and not the global optimum of the objective function. In this paper, we demonstrate that in the case of one-dimensional clustering with one main and one noise cluster, one can formulate an objective function, which permits a closed-form solution with no need for an iteration scheme and the guarantee of finding the global optimum. We demonstrate how such an algorithm can be applied in the context of laboratory medicine as a method to estimate reference intervals that represent the range of “normal” values.

Estimation of reference interval for neutrophil activity evaluation systems: A interim report

Estimation of reference interval for neutrophil activity evaluation systems: A interim report

Indirect estimation of pediatric reference interval via density graph deep embedded clustering

Establishing reference intervals (RIs) for pediatric patients is crucial in clinical decision-making, and there is a critical gap of pediatric RIs in China. However, the direct sampling technique for establishing RIs is resource-intensive and ethically challenging. Indirect estimation methods, such as unsupervised clustering algorithms, have emerged as potential alternatives for predicting reference intervals. This study introduces deep graph clustering methods into indirect estimation of pediatric reference intervals. Specifically, we propose a Density Graph Deep Embedded Clustering (DGDEC) algorithm, which incorporates a density feature extractor to enhance sample representation and provides additional perspectives for distinguishing different levels of health status among populations. Additionally, we construct an adjacency matrix by computing the similarity between samples after feature enhancement. The DGDEC algorithm leverages the adjacency matrix to capture the interrelationships between patients and divides patients into different groups, thereby estimating reference intervals for the potential healthy population. The experimental results demonstrate that when compared to other indirect estimation techniques, our method ensures the predicted pediatric reference intervals in different age and gender groups are closer to the true values while maintaining good generalization performance. Additionally, through ablation experiments, our study confirms that the similarity between patients and the multi-scale density features of samples can effectively describe the potential health status of patients.

Estimation of reference intervals for physical fitness indicators in healthy community-dwelling older adults

Estimation of reference intervals for physical fitness indicators in healthy community-dwelling older adults

Unsupervised machine learning method for indirect estimation of reference intervals for chronic kidney disease in the Puerto Rican population

Reference intervals (RIs) for clinical laboratory values are extremely important for diagnostics and treatment of patients. However, the determination of these ranges is costly and time-consuming. As a result, often different unverified RIs are used in practice for the same analyte and the same range is used for all patients despite evidence that the values are gender, age, and ethnicity dependent. Moreover, the abnormal flags are rudimentary, merely indicating if a value is within the RI. At the same time, clinical lab data generated in the everyday medical practice contains a wealth of information, that given the correct methodology, can help determine the RIs for each specific segment of the population, including populations that suffer from health disparities. In this work, we develop unsupervised machine learning methods, based on Gaussian mixtures, to determine RIs of analytes related to chronic kidney disease, using millions of routine lab results for the Puerto Rican population. We show that the measures are both gender and age dependent and we find evidence for normal age-related organ function deterioration and failure. We also show that the joint distribution of measures improves the diagnostic value of the lab results.

B-152 Estimation of Creatinine Reference Interval (RI) by the Indirect Method using an Unsupervised Machine Learning Tool and R Programming Language

Abstract Background Creatinine is widely used in diagnostic medicine as a primary renal marker. Clinical laboratory measurement of serum creatinine is not standardized, and the RI varies slightly between different manufacturers. Roche brings as reference ranges 0.5–0.9 mg/DL for women and 0.7–1.2 mg/dL for men. After clinical manifestations of elevated female results without clinical correspondence, the laboratory proceeded with reevaluation of RI. Methods We retrospectively studied the creatinine results of the laboratory database of samples run from January to October 2022 in two different technical areas in Brazil (São Paulo and Rio de Janeiro). The laboratory platform was Roche® 502/702, Jaffe method to creatinine. The LabRI tool was used with parametric, non-parametric, and robust statistical treatments to exclude outliers and algorithms in the R Language and latent abnormal values of outpatient patients. The partitions proposed were gender (male and female), age (18–60 years old and over 60 years old). The proposed RI for verification were those proposed by Rosenfeld, LG et al in REV BRAS EPIDEMIOL 2019; 22 (SUPPL2):E190002.SUPL.2 - F: 0.5–0.9 and M: 0.7–1.2. Exclusion criteria were patients with urea >48.5 mg/dL, glycemia >200 mg/dL, 24 h proteinuria >150 mg/24 h, LDL > 130 mg/dL, uric acid > 130 mg/dL, anti-nucleus positive factor, as well as clinical data compatible with chronic renal failure or dialysis, hypertension and diabetes. Results After exclusion, retrieved 225 181 individuals 52% female. Between the different sites, the results of the partitions did not undergo significant changes. For women in both age partitions, the values did not vary significantly. For men, there was no significant variation in the lower limit of IR, and we had a variation of 0.20 (18–60: 0.69–1.31 and >60: 0.69–1.41) in the upper limit, mainly for the age partition >60 years old, below the 95% confidence interval of the proposed references. Conclusions The adoption of the Roche IR and the Brazilian article was approved for the female gender (0.5–0.9) with 118 100 individuals, and a new IR was defined for the male gender (0.7–1.2) with 107 081 adult individuals, including the elderly over 60 years old, for both genders.

B-151 Estimation of the Uric Acid Reference Interval by the Indirect Method Using an Unsupervised Machine Learning Tool and R Programming Language

Abstract Background Determining the Reference Intervals (RI) of quantitative clinical laboratory tests is crucial for patient diagnosis, treatment, and monitoring. It is also a concern for laboratories seeking to meet accreditation requirements. Our study aimed to evaluate the manufacturer’s RI for uric acid (3.7–9.2 mg/dL) through indirect methods using a programming tool in R language in a Brazilian laboratory. Methods We performed a data mining study for uric acid using the laboratory’s LIS database from January to December 2021. Outpatients >18 years old, analyzed on the Siemens® platform with results within linearity were included. Exclusion criteria were total cholesterol >200 mg/dL, triglycerides >150 mg/dL, glycemia >126 mg/dL, ferritin >300 ng/mL and creatinine >1.5 mg/dL at any time in the patient history and CRP >10 mg/dL in the same month. We used a tool that integrates different algorithms including preliminary treating, grouping, and statistical reduction of test results. The Expectation-Maximization (EM) algorithm calculates the mean, standard deviation, and weight of the distribution for each cluster. Additionally, it integrates algorithms that determine the global mean/median, modes, antimodes, kurtosis, and skewness, to find values that best represent the reference population’s distribution (Figure). Manufacturer's and estimated RI for Male and Female partitions. Results The final database included 49.489 subjects, partitioned in 26.439 women and 23.050 men. Statistically estimated RI were 2.8–6.7 mg/dL for women and 3.8–8.6 mg/dL for men. Considering the RI proposed by the manufacturer, 23.8% of women were outside the RI provided by the manufacturer, all of them below the lower reference value. Conclusions Our results confirm the importance of verification/establishment of reference intervals by clinical laboratories and suggest that estimated RI using this software is appropriate for the assisted population, demonstrating consistency with physicians’ concerns. After specialist analysis, the manufacturer´s RI has been replaced by the calculated population-specific RI.

Critical appraisal of two Box-Cox formulae for their utility in determining reference intervals by realistic simulation and extensive real-world data analyses

BackgroundThe reference interval (RI) is defined as the central 95 % range of reference values (RVs) from healthy individuals. The ideal method for determining RIs is to transform RV distribution into Gaussian and estimate its 95 % range parametrically. One-parameter Box-Cox formula (1pBC) is widely used for correcting skewness (Sk) or kurtosis (Kt) in data distribution. However, 1pBC is not popular for computing RIs due to its unreliability in Gaussian transformation. While its two-parameter version (2pBC) is not used due to a challenge in fitting power (λ) and shift (α) parameters simultaneously. In this study, technical issues in fitting both formulae are assessed, and an alternative algorithm for successful use of 2pBC is proposed. MethodsFor fitting 1pBC, optimal λ was determined by stepwise linear search. For 2pBC, optimal [λ, α] combination was pursued in two ways: by grid search of λ and α (2pBCgrid) or by using the grid search but keeping α-range close to the reference distribution (2pBCopt). Their accuracy and precision in determining RIs were compared by generating power-normal distributions simulating RVs of 23 major chemistry analytes. Additionally, their practical utilities were compared by analyzing 776 real-world datasets comprising test results of 25 analytes that were obtained from the global multicenter RV study of IFCC. For comparison, the performance of nonparametric method was evaluated in both settings. ResultsFor analytes with not-much-skewed distributions, unbiased estimation of RIs was accomplished by all methods. Nevertheless, when reference distributions are located far from zero, λ estimated by1pBC and 2pBCgrid fluctuated widely, which was attributable to virtually flat goodness-of-fit profile for [λ, α]. For highly skewed distributions, 1pBC caused bias in estimating RI and λ due to remotely peaked goodness-of-fit profile. Real-world data analyses revealed that 2pBCopt and 1pBC achieved Gaussian transformation (|Sk|<0.1 and |Kt|<0.3) in 82.4 % and 66.9 % among 776 datasets, respectively. Fitting bias signified by Kt<−0.4 was more common to 1pBC. The practical utility of 2pBCopt was unbiased prediction of analyte-specific distribution-shape (λ). Whereas nonparametric method gave highly variable RIs for both simulated and real-world datasets. Conclusions2pBCopt is suitable for calculating RIs and grasping distribution-shape from the estimated λ.

A pipeline for the fully automated estimation of continuous reference intervals using real-world data

Reference intervals are essential for interpreting laboratory test results. Continuous reference intervals precisely capture physiological age-specific dynamics that occur throughout life, and thus have the potential to improve clinical decision-making. However, established approaches for estimating continuous reference intervals require samples from healthy individuals, and are therefore substantially restricted. Indirect methods operating on routine measurements enable the estimation of one-dimensional reference intervals, however, no automated approach exists that integrates the dependency on a continuous covariate like age. We propose an integrated pipeline for the fully automated estimation of continuous reference intervals expressed as a generalized additive model for location, scale and shape based on discrete model estimates using an indirect method (refineR). The results are free of subjective user-input, enable conversion of test results into z-scores and can be integrated into laboratory information systems. Comparison of our results to established and validated reference intervals from the CALIPER and PEDREF studies and manufacturers’ package inserts shows good agreement of reference limits, indicating that the proposed pipeline generates high-quality results. In conclusion, the developed pipeline enables the generation of high-precision percentile charts and continuous reference intervals. It represents the first parameter-less and fully automated solution for the indirect estimation of continuous reference intervals.

Comparison of three indirect methods for verification and validation of reference intervals at eight medical laboratories: a European multicenter study

Abstract Objectives Indirect methods for the indirect estimation of reference intervals are increasingly being used, especially for validation of reference intervals, as they can be applied to routine patient data. In this study, we compare three statistically different indirect methods for the verification and validation of reference intervals in eight laboratories distributed throughout Europe. Methods The RefLim method is a fast and simple approach which calculates the reference intervals by extrapolating the theoretical 95 % of non-pathological values from the central linear part of a quantile-quantile plot. The Truncated Maximum Likelihood (TML) method estimates a smoothed kernel density function for the distribution of the mixed data, for which it is assumed that the ‘‘central’’ part of the distribution represents the healthy population. The refineR utilizes an inverse modelling approach. This algorithm identifies a model that best explains the observed data before transforming the data with the Box-Cox transformation. Results We show that the different indirect methods each have their advantages but can also lead to inaccurate or ambiguous results depending on the approximation of the mathematical model to real-world data. A combination of different methodologies can improve the informative value and thus the reliability of results. Conclusions Based on routine measurements of four enzymes alkaline phosphatase (ALP), total amylase (AMY), cholinesterase (CHE) and gamma-glutamyl transferase (GGT) in adult women and men, we demonstrate that some reference limits taken from the literature need to be adapted to the laboratory’s particular local and population characteristics.

Age-and sex-specific reference intervals of total cholesterol, LDL cholesterol, HDL cholesterol and non-HDL cholesterol. Comparison of two algorithms for the indirect estimation of reference intervals

Abstract Objectives Reference intervals of total cholesterol concentrations in plasma and of their fractions low-density lipoprotein (LDL)-, high-density lipoprotein (HDL)-and non-HDL concentrations are seldom studied with respect to the relevance of age and sex. Therefore, the effect of age and sex on the reference intervals was reinvestigated with 2 indirect procedures. Methods As an indirect approach, the truncated minimum chi-square method was applied. All analyses were performed by computer programs available. The script published on the homepage of the German Society of Clinical Chemistry and Laboratory Medicine (DGKL) allows to derive a continuous age dependency of reference intervals together with their confidence and equivalence limits. The results of this approach were compared with those obtained by an indirect method developed more recently, the refineR algorithm. Results In the present study, the upper reference limits of total cholesterol varied from 5.1 to 7.8 mmol/L (197–302 mg/dL) depending on various biological variables (as age, sex, inpatients versus outpatients). These upper limits increased with age. Differences between sexes can be neglected except for the age above 80 years. The pattern of reference limits of LDL cholesterol and non-HDL cholesterol paralleled those of total cholesterol. The reference limits of HDL cholesterol were higher in women than in men but were independent of age. Conclusions Reference limits for the concentrations of total cholesterol and their fractions LDL-, HDL-and non-HDL concentrations should be stratified for age and sex.

Non-parametric Clinical Laboratory Reference Interval Estimation in Volunteer Blood Donors: An Example for Prothrombin Time and Partial Thromboplastin Time

Introduction. To date, there are no reference intervals for prothrombin time (PT) and activated partial thromboplastin time (APTT) based on “normal” Filipino adults. The common practice in most laboratories is to adopt manufacturer provided values or foreign literature even if the importance of establishing or at least verifying laboratory reference intervals has been stressed by Clinical Laboratory Standards Institute (CLSI). Objectives. Here we aim to describe our experience in using a simple non-parametric method to generate reference intervals for PT and APTT, from healthy Filipino volunteer blood donors. Methodology. We used a de novo, a priori non-parametric estimation method following the CLSI guidelines on establishing reference intervals. Results. The non-parametric lower reference limit for PT is 12.55 seconds, with 90% confidence interval of 12.3 to 12.75 seconds. While the non-parametric upper reference limit for PT is 16.15 seconds, with 90% confidence interval of 15.55 to 16.55 seconds. The non-parametric lower reference limit for activated partial thromboplastin time is 26.12 seconds, with 90% confidence interval of 22.95 to 27.1 seconds, and the non-parametric upper reference limit for activated partial thromboplastin time is 37.44 seconds, with 90% confidence interval of 36.75 to 38.65 seconds. The PT and APTT reference intervals were different from foreign sources and manufacturer provided values in terms of interval width and values of the reference limits by 2 to 4 seconds. Conclusion and Recommendations. Estimation of coagulation reference intervals from volunteer health blood donors is doable, simple, and practical. Collaborative multi-center efforts may be done to expand the pool of reference individuals that are included and increase the representativeness of the reference intervals generated. This simple method can also be used to generate reference intervals for other clinical laboratory assays and may also be extended to at least verify reference intervals in special populations like pregnant women, the elderly, and the pediatric population.

Verification of sex- and age-specific reference intervals for 13 serum steroids determined by mass spectrometry: evaluation of an indirect statistical approach.

Conventionally, reference intervals are established by direct methods, which require a well-characterized, obviously healthy study population. This elaborate approach is time consuming, costly and has rarely been applied to steroid hormones measured by mass spectrometry. In this feasibility study, we investigate whether indirect methods based on routine laboratory results can be used to verify reference intervals from external sources. A total of 11,259 serum samples were used to quantify 13 steroid hormones by mass spectrometry. For indirect estimation of reference intervals, we applied a "modified Hoffmann approach", and verified the results with a more sophisticated statistical method (refineR). We compared our results with those of four recent studies using direct approaches. We evaluated a total of 81 sex- and age-specific reference intervals, for which at least 120 measurements were available. The overall agreement between indirectly and directly determined reference intervals was surprisingly good as nearly every fourth reference limit could be confirmed by narrow tolerance limits. Furthermore, lower reference limits could be provided for some low concentrated hormones by the indirect method. In cases of substantial deviations, our results matched the underlying data better than reference intervals from external studies. Our study shows for the first time that indirect methods are a valuable tool to verify existing reference intervals for steroid hormones. A simple "modified Hoffmann approach" based on the general assumption of a normal or lognormal distribution model is sufficient for screening purposes, while the refineR algorithm may be used for a more detailed analysis.

Reference intervals for thyroid disorders calculated by indirect method and comparison with reference change values.

The aim of the study was to calculate reference intervals (RIs) for thyroid stimulating hormone (TSH), free thyroxine (fT4) and free triiodothyronine (fT3) and evaluate the clinical significance of these intervals by use of reference change values (RCV) of the analytes. Laboratory patient data between August and December 2021 were evaluated for the study. A total of 188,912 patients with TSH, fT4, fT3, anti-thyroid peroxidase antibodies (Anti-TPO) and anti-thyroglobulin antibodies (Anti-Tg) results were evaluated. All measurements were performed on Cobas c801 (Roche Diagnostics, Penzberg, Germany) using electrochemiluminescence immunoassay technology. Estimated RIs were compared with manufacturer's by means of RCVs of analytes. Thyroid stimulating hormone values didn't differ significantly by gender and age. The combined RIs for whole group (N = 28,437) was found as 0.41-4.37 mIU/mL. Free T4 values (11.6-20.1 pmol/L, N = 13,479 in male; 10.5-19.5 pmol/L, N = 17,634 female) and fT3 values (3.38-6.35 pmol/L, N = 2,516 in male; 3.39-5.99 pmol/L, N = 3,348 pmol/L in female) significantly differed by gender (P < 0.050). Both fT4 and fT3 values also showed significant differences in age subgroups comparisons. So, male and female RIs were represented separately for age subgroups. When compared with manufacturer's RIs, TSH whole group and fT4 subgroups RIs didn't exceed the analytes' RCVs, but this difference was greater for fT3. Reference interval estimation by use of indirect method out of laboratory data may be more accurate than manufacturer provided RIs. This population based RIs evaluated using RCV of analytes may provide useful information in clinical interpretation of laboratory results.

Definitions and major prerequisites of direct and indirect approaches for estimating reference limits.

Reference intervals are established either by direct or indirect approaches. Whereas the definition of direct is well established, the definition of indirect is still a matter of debate. In this paper, a general definition that covers all indirect models presently in use is proposed. With the upcoming popularity of indirect models, it has become evident that further partitioning strategies are required to minimize the risk of patients' false classifications. With indirect methods, such partitions are much easier to execute than with direct methods. The authors believe that the future of reference interval estimation belongs to indirect models with big data pools either from one laboratory or combined from several regional centres (if necessary). Independent of the approach applied, the quality assurance of the pre-analytical and analytical phase, considering biological variables and other confounding factors, is essential.

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

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. 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). 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. 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.

Indirect Method for Estimation of Reference Intervals of Inflammatory Markers.

The direct method for reference interval (RI) estimating is limited due to the requirement of resources, difficulties in defining a non-diseased population, or ethical problems in obtaining samples. We estimated the RI for inflammatory biomarkers using an indirect method (RII). C-reactive protein (CRP), erythrocyte sedimentation rate (ESR) and presepsin (PSEP) data of patients visiting a single hospital were retrieved from April 2009 to April 2021. Right-skewed data were transformed using the Box-Cox transformation method. A mixed population of non-diseased and diseased distributions was assumed, followed by latent profile analysis for the two classes. The intersection point of the distribution curve was estimated as the RI. The influence of measurement size was evaluated as the ratio of abnormal values and adjustment (n×bandwidth) of the distribution curve. The RIs estimated by the proposed RII method (existing method) were as follows: CRP, 0-4.1 (0-4.7) mg/L; ESR, 0-10.2 (0-15) mm/hr and PSEP, 0-411 (0-300) pg/mL. Measurement sizes ≥2,500 showed stable results. An abnormal-to-normal value ratio of 0.5 showed the most accurate result for CRP. Adjustment values ≤5 or >5 were applicable for a measurement size <25,000 or ≥25,000, respectively. The proposed RII method could provide additional information for RI verification or estimation with some limitations.

Mixture density networks for the indirect estimation of reference intervals

BackgroundReference intervals represent the expected range of physiological test results in a healthy population and are essential to support medical decision making. Particularly in the context of pediatric reference intervals, where recruitment regulations make prospective studies challenging to conduct, indirect estimation strategies are becoming increasingly important. Established indirect methods enable robust identification of the distribution of “healthy” samples from laboratory databases, which include unlabeled pathologic cases, but are currently severely limited when adjusting for essential patient characteristics such as age. Here, we propose the use of mixture density networks (MDN) to overcome this problem and model all parameters of the mixture distribution in a single step.ResultsEstimated reference intervals from varying settings with simulated data demonstrate the ability to accurately estimate latent distributions from unlabeled data using different implementations of MDNs. Comparing the performance with alternative estimation approaches further highlights the importance of modeling the mixture component weights as a function of the input in order to avoid biased estimates for all other parameters and the resulting reference intervals. We also provide a strategy to generate partially customized starting weights to improve proper identification of the latent components. Finally, the application on real-world hemoglobin samples provides results in line with current gold standard approaches, but also suggests further investigations with respect to adequate regularization strategies in order to prevent overfitting the data.ConclusionsMixture density networks provide a promising approach capable of extracting the distribution of healthy samples from unlabeled laboratory databases while simultaneously and explicitly estimating all parameters and component weights as non-linear functions of the covariate(s), thereby allowing the estimation of age-dependent reference intervals in a single step. Further studies on model regularization and asymmetric component distributions are warranted to consolidate our findings and expand the scope of applications.