If the two methods yield identical results, then the plot should have an expected slope, β1, of 1.00 and an expected y-intercept, β0, of 0.0. In addition, however, the number of false-positives increases, which causes the specificity of this test to worsen since specificity = TN/(TN + FP), because any increase in the number of false-positives, a term in the denominator of this equation, results in a decrease in the value given by this equation. Lamb Report the concentration of Cd 2 + in the sample and its 95% confidence interval. $t_\ce{exp} = \dfrac{|β_0 − b_0|}{s_{b_0}} = \dfrac{|b_0|}{s_{b_0}}$. fun fact: you can also print this whole blog article as a sample PDF. Lastly, it is important to remember that knowing the sensitivity (ie, positivity in disease) and specificity (ie, negativity in health or non-disease) of a test is of limited value because these parameters represent the answer to the question: What is the probability of a patient having a positive test result if this patient has disease X? FH are incorporated into software programs that are relatively inexpensive to purchase and use. Example of a distribution of laboratory test values for an analyte (ie, the liver enzyme, gamma-glutamyl transferase [GGT]) for which the data are not Gaussian distributed. Miller Information and translations of analyte in the most comprehensive dictionary definitions resource on the web. Electrochemical test strips measure an electrical current when whole blood is applied . $s_{b_1}= \sqrt{\dfrac{6×(1.997×10^{−3})^2}{6×(1.378×10^{−4}) − (2.371×10^{−2})^2}} = 0.3007$, $s_{b_0}= \sqrt{\dfrac{(1.997×10^{−3})^2×(1.378×10^{−4})}{6×(1.378×10^{−4}) − (2.371×10^{−2})^2}} = 1.441×10^{−3}$, $β_1 = b_1 ± ts_{b_1} = \mathrm{29.57 ± (2.78×0.3007) = 29.57\: M^{-1} ± 0.85\: M^{-1}}$, $β_0 = b_0 ± ts_{b_0}= 0.0015 ± \left\{2.78 × (1.441×10^{−3})\right\} = 0.0015 ± 0.0040$, With an average Ssamp of 0.114, the concentration of analyte, CA, is, $C_\ce{A} = \dfrac{S_\ce{samp} − b_0}{b_1} = \mathrm{\dfrac{0.114− 0.0015}{29.57\: M^{-1}} = 3.80×10^{−3}\: M}$, $s_{C_\ce{A}}= \dfrac{1.997×10^{−3}}{29.57} \sqrt{\dfrac{1}{3} + \dfrac{1}{6} + \dfrac{(0.114−0.1183)^2}{(29.57)^2 × (4.408×10^{-5})}} = 4.778×10^{−5}$, \begin{align} TSH, thyroid-stimulating hormone; fT4, free thyroxine; NTI, nonthyroid illness; T3, trilodothyronine; HyperT, hyperthyroidism; HypoT, hypothyroidism. The magnitude of analytical variation is given by CVa, while biological variability is defined by CVb. Quantitative values for all analytes are affected by both imprecision (ie, lack of reproducibility) in the measurement of the analyte and intra-individual variation over time in the concentration of the analyte due to normal physiologic mechanisms (ie, biological variation) that are independent of any disease process. Fanke and co-workers evaluated a standard additions method for a voltammetric determination of Tl.16 A summary of their results is tabulated in the following table. Alternative approaches are found in Hartman, C.; Smeyers-Verbeke, J.; Penninckx, W.; Massart, D. L. Anal. How good is the test in discriminating between health versus disease? By appropriately selecting patients on whom the test is performed (ie, by maximizing the prevalence of disease in the population sampled). On the other hand, if a disease is potentially fatal and no therapy, other than supportive care, is available (eg, cystic fibrosis), false-positives would be unacceptable. Moreover, the diagnostic accuracy (ie, sensitivity and specificity) of PSA in providing this discrimination is higher (AUC = 0.86) in Figure 2.3B than in Figure 2.3A (AUC = 0.66), probably due to differences in the study designs represented by the data shown in each panel of Figure 2.3. From this data, the values for prevalence, sensitivity, specificity, efficiency, positive predictive value (PPV), and negative predictive value (NPV) can be determined: Prevalence (p) = No. Thus, both qualitative and quantitative ROC analysis demonstrates that PSA provides better discrimination than PAP in distinguishing men with prostate cancer from those with BPH or prostatitis. Legal. Most likely, FDA would consider such kits to be fin-ished IVD tests that require FDA clearance or approval. To calculate the 95% confidence intervals, we first need to determine the standard deviation about the regression. concentrations of more than one analyte in a single sample may be determined. Gray Common sources of preanalytical error; however, frequency decreasing with advent of better quality assurance (QA) procedures to ensure positive patient ID and labeling of specimen tubes. NA of true-positives + No. The principal approaches for establishing a diagnosis based on laboratory test results include:4. This user guide includes all the information that you need to run POC assays using the CardioChek PA test system. Average the replicate signals for each standard before completing the linear regression analysis. of false-negatives) or TP/(TP + FN) = 160/(160 + 40) = 160/200 = 0.800 = 80%, Specificity = percentage of individuals without disease who have a negative test result = No. More importantly, our scheme reports class membership probabilities for each and every test sample. If uncorrected, the presence of the interference will result in biased estimates of analyte concentration. of true-negatives + No. Inaccurate quantification due to poor analyte recovery can be overcome by matrix matching, which involves the preparation of calibration standards in a sample similar in composition to the sample under study (Akkaya et al., 2017). Adapted from Ref 1. The analyte-coated microparticles are rapidly agglutinated in the presence of antibody reactive to the analyte and without competing analyte in the test sample. Quantitatively, the AUC values for PSA and PAP are 0.86 and 0.67, respectively. In the current climate of business-oriented medicine, the answer should not be: Any test for which reimbursement by a payer (eg, Medicare) is likely to be denied. When using population-based reference intervals, however, it is critical that members of the reference population be free of any obvious or overt disease, especially diseases likely to affect the analyte for which the reference interval is being determined. An example of a qualitative assay is an immunoassay test for pregnancy. Tests with a combined sensitivity and specificity total = 170 or greater are likely to prove clinically useful. What is the analyte’s concentration in the original sample? Inappropriate specimen collection tube/container. This article demonstrates how to filter which tests are run. The cost of the test and/or the technical difficulty in performing the test may be so high that its availability is limited. of 0.114, CA, \[C_\ce{A} = \dfrac{S_\ce{samp} − b_0}{b_1} = \mathrm{\dfrac{0.114− 0.0014}{29.59\: M^{-1}} = 3.80×10^{−3}\: M}, $s_{C_\ce{A}}= \dfrac{1.996×10^{−3}}{29.59} \sqrt{\dfrac{1}{3} + \dfrac{1}{6} + \dfrac{(0.114−0.1183)^2}{(29.59)^2 × (4.408×10^{-5})}} = 4.772×10^{−5}$, \begin{align} In addition, a change in values for an analyte that exceeds the change (ie, reference change value [RCV]) expected due to the combined effects of analytical and biological variation alone is due most likely to a disease process or to the affect of any therapy on the disease. Algorithm for identifying individuals with thyroid disorders based on TSH level. As private health maintenance organizations (HMOs) and government-sponsored agencies (eg, Department of Health and Human Services [DHHS] and the Centers for Medicare and Medicaid Services [CMS]) seek to provide quality medicine cost effectively, reduction in the ordering of “unnecessary” laboratory tests has become a favorite target of these efforts. (c) Plot each data set along with the regression line and comment on your results. We estimate the cost of the test to be less than 500, which is comparable or lower than other screening tests for single cancers, such as colonoscopy. of true-negatives/(No. Thus, it is incumbent upon physicians and laboratorians to understand which laboratory tests are appropriate to order in the diagnosis and follow up of a patient’s medical condition. relevant to the patient. The matrix is everything in the sample except the analyte. 5. A 50.00 mL sample containing an analyte gives a signal of 11.5 (arbitrary units). The clinical usefulness or practical value of the information provided by ROC curves in patient care may vary, however, even for tests that have good discriminating ability (ie, high sensitivity and specificity at a particular decision threshold). This cycle consists of 3 phases: preanalytic, analytic, and post-analytic (Figure 1). With an Ssamp. Frontotemporal dementia, music perception and social cognition share neurobiological circuits: A meta-analysis. There are 4 major legitimate reasons for ordering a laboratory test:4. Acceptance Criteria ≥3 for each replicate . Learning needs in clinical biochemistry for doctors in foundation, © American Society of Clinical Pathologists, Significant Operational Improvements with Implementation of Next Generation Laboratory Automation, Point-of-Care Testing Effectiveness on Blood Donor Hemoglobin Testing, MicroRNA-138 Regulates T-Cell Function by Targeting PD-1 in Patients with Hepatitis B Virus–Related Liver Diseases, A Hemolytic Transfusion Reaction Caused by an Unexpected Le, Daratumumab Interference in Flow Cytometry Producing a False Kappa Light Chain Restriction in Plasma Cells, About the American Society for Clinical Pathology, Questions to Ask Before Ordering a Laboratory Test, Approaches for Establishing a Diagnosis Based on Laboratory Test Results, Clinical Performance Characteristics of Laboratory Tests, Receiver-Operator Characteristic (ROC) Curves, Reference Interval for Interpreting Laboratory Test Results, Critical Difference Between Consecutive Laboratory Test Results, Receive exclusive offers and updates from Oxford Academic, Laboratory Procedures Available at Your Fingertips: CLSI’s, Update on Laboratory Testing to Identify Individuals With Inhalant Allergies, Validating New Reagents: Roadmaps Through the Wilderness, STAT Turnaround Time: Satellite and Point-to-Point Testing. A “perfect” test would have both 100% diagnostic sensitivity and specificity, which seldom occurs in practice and if it does, the population of diseased and non-diseased patients studied was probably not large and varied enough to demonstrate that the test was not perfect. The signal for the reagent blank is 0.002. \end{align}. 2 . The following table contains the relevant information. of individuals with disease/No. M Dramatic representation of diagnostic sensitivity and specificity using the analyte prostate-specific antigen (PSA) as an example. In practice, control samples are most commonly used for spiking. . Examples of MLL FISH signal pattern results from a test set of 4 normal cases, 3 cases with 11q23 rearrangements, 2 with loss of chromosome 11, and 1 with gain of chromosome 11 are presented in Table 3. The hardship (eg, financial and/or physical) associated with the test may cause patients to be unwilling to submit to the test. Bacterial, viral, and/or fungal cultures. If the test portion is a particulate solid, it may be necessary to convert it to a solution. In this example, the test is assigned 95% diagnostic specificity and 95% diagnostic sensitivity. Specimen collection tube not filled properly. Since physicians frequently order the same test at multiple time points during the course of the patients’ management, they are faced with the challenge of interpreting when the magnitude of the change in values for an analyte constitutes a significant change (or critical difference [CD]) that may (or should) affect medical decision making (eg, trigger a change in therapy, such as increasing or decreasing a drug dosage). The days of a “shotgun approach” to ordering laboratory tests has, of necessity, been replaced by a “rifle” (or targeted) approach based on an understanding of the test’s diagnostic performance and the major “legitimate” reasons for ordering a laboratory test. Before considering this aspect of the selection of laboratory tests, we must first understand the terms that describe their diagnostic performance. All quantitative assays have a finite lower limit of detection (LLD), distinct from 0, that more precisely constitutes the lower limit of the reference interval when this lower limit encompasses 0. In the example cited above, performing PSA testing on men over age 50 years improves the PPV of PSA since the prevalence of prostate cancer increases from <1% in Caucasian men aged less than 50 years to 16% in men aged 50 to 64 years and to 83% in men over 64 years of age. Oral contraceptives with high [progestin], Hemoconcentration (eg, use of a tourniquet), Specimen storage (@ 0–4 °C for up to 4 days), Copyright © 2021 American Society for Clinical Pathology. The purpose of this CE Update is to discuss the laboratory testing cycle and its importance in diagnostic decision making. Laboratory data is never a substitute for a good physical exam and patient history (clinicians should treat the patient, not the laboratory results). Using LC or LC-MS/MS for detection, this test produces precise numerical results in micrograms per kilogram (μg/kg) or parts per billion (ppb). of true-positives/(No. Assume that the uncertainty in the stock solution’s molarity is ±0.002. Calculate the concentration of analyte in a sample with a signal of 0.118. Screening (eg, for congenital hypothyroidism via neonatal thyroxine testing). Therefore, any PSA value less than 0.05 ng/mL would be reported appropriately as “less than 0.05 ng/mL” and not as 0.0 ng/mL. Frank H. Wians, Clinical Laboratory Tests: Which, Why, and What Do The Results Mean?, Laboratory Medicine, Volume 40, Issue 2, February 2009, Pages 105–113, https://doi.org/10.1309/LM4O4L0HHUTWWUDD. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide, This PDF is available to Subscribers Only. To determine the concentration of analyte in a sample, a standard additions was performed. 11. For many PSA assays, the LLD is typically 0.05 ng/mL. The menu of laboratory tests available to clinicians constitutes an impressive array that has expanded exponentially since 1920 when Folin and Wu devised the first useful test for the quantification of serum glucose concentration.3 The current list of tests offered by one major reference laboratory includes nearly 3,000 analytes, which does not include the additional array of more commonly ordered tests (eg, complete blood count [CBC], electrolytes [sodium, potassium, chloride, carbon dioxide], thyroid stimulating hormone [TSH], glucose, etc.) Thus, reference intervals are intended to serve as a guideline for evaluating individual values and, for many analytes, information on the limits of an analyte for a population of individuals with the disease or diseases the test was designed to detect is even more informative. . ), determined on a defined population of “healthy” individuals, that lie between the lower and the upper limits that constitute 95% of all values. (d) A single sample is buffered to a pH of 3.7 and analyzed for cadmium, yielding a signal of 66.3. Estimate the precision and reproducibility of serological assays that test human plasma from donations of whole blood and blood components with the Thermo Scientific™ AcroMetrix Serology Single Analyte Controls. This test was performed on 10,000 men, 200 of whom have biopsy-proven prostate cancer. $\sum_{i}x_i = 2.371×10^{-2} \hspace{20px} \sum_{i}y_i = 0.710$, $\sum_{i}x_iy_i= 4.110×10^{–3} \hspace{20px} \sum_{i}x_i^2 = 1.278×10^{–4}$, Substituting these values into equation 5.17 and equation 5.18, we find that the slope and the y-intercept are, $b_1 = \dfrac{6×(4.110×10^{−3}) − (2.371×10^{−2})×(0.710)}{(6×1.378×10^{−4}) − (2.371×10^{−2})^2} = 29.57$, $b_0 = \dfrac{0.710 − 29.57×(2.371×10^{−2})}{6} = 0.0015$, $S_\ce{std} = 29.57 × C_\ce{std} + 0.0015$. as single- and multiple-analyte test strips. Simplified algorithm for the diagnosis of a monoclonal gammopthy versus asymptomatic multiple myeloma versus active multiple myeloma (Source: Mayo Communique. Readers should also be able to describe the general principles for selecting the most appropriate laboratory test based on its diagnostic performance characteristics. 7. In general, laboratory tests with both high sensitivity and high specificity are desirable since both false-negatives and false-positives are equally unacceptable under most clinical circumstances. $S_\ce{spike} = k_\ce{A}\left(C_\ce{A}\dfrac{V_\ce{o}}{V_\ce{f}} + C_\ce{std}\dfrac{V_\ce{std}}{V_\ce{f}}\right)$, $0 = \dfrac{k_\ce{A}C_\ce{A}V_\ce{o}}{V_\ce{f}} + k_\ce{A} × \left\{C_\ce{std} \dfrac{V_\ce{std}}{V_\ce{f}}\right\}$, which is in the form of the linear equation, $Y = y\textrm{-intercept} + \ce{slope} × X$, where Y is Sspike and X is Cstd×Vstd/Vf. More recently, neural networks, a branch of artificial intelligence, have been used to evaluate and interpret laboratory data.7,8 These computerized networks mimic the processes performed by the human brain and can learn by example and generalize. The pattern of laboratory test results shown for the pregnant “Patient” in Table 2 most closely match those consistent with a diagnosis of idiopathic thrombocytopenic purpura (ITP), rather than other possible causes of pregnancy-associated thrombocytopenia: gestational thrombocytopenic (GTP); thrombotic thrombocytopenia (TTP); hemolytic uremic syndrome (HUS); disseminated intravascular coagulation (DIC); or, (syndrome of) hemolysis, elevated liver enzymes, and low platelet count (HELLP). Chim. To three significant figures, each data set has a slope of 0.500 and a y-intercept of 3.00. The purest substances are referred to as analytes. Such an understanding is critical to good laboratory practice and patient outcomes. Each solution was analyzed twice. Less invasive or less expensive tests may provide similar information. Highly purified … The effect of prevalence on predictive value is demonstrated in Table 2. Example 2 The laboratory test, prostate-specific antigen (PSA), was studied with regard to its ability to discriminate patients with prostate cancer (PCa) from those without PCa. Our study lays the conceptual and practical foundation for a single, multi-analyte blood test for cancers of many types. Chaudhari (b) Based on your results explain why the authors concluded that the internal standardization was inappropriate. Because Excel’s Data Analysis tools provide most of the information we need, we will use it here. $t_\ce{exp} = \dfrac{|β_1 − b_1|}{s_{b_1}} = \dfrac{|1.00 − b_1|}{s_{b_1}}$. Most clinicians can achieve this total with a good history and physical examination! n.a., not applicable. Because the clinical performance characteristics of all laboratory tests differ with respect to their diagnostic accuracy (ie, sensitivity and specificity), the selection of the appropriate laboratory test to order will vary depending on the purpose for which the test is to be used. Because indeterminate errors affect both analytical methods, the result of unweighted linear regression is biased. TA (B) Prostatic acid phosphatase (PAP) and PSA in differentiating prostate cancer from benign prostatic hyperplasia and prostatitis at various cutoff values (indicated adjacent to points on each of the curves). Thus, 95% of healthy men have a serum PSA concentration between these limits. Since reference intervals for all analytes are based typically on the limits for the analyte that include 95% of all values obtained on healthy individuals with the assumption that the distribution of these values is Gaussian (or “bell-shaped”), it is important to recognize that 5% (or 1 out of 20; ie, the 2.5% of healthy individuals with analyte values in the left tail of the data distribution and the 2.5% of healthy individuals with analyte values in the right tail of the distribution when the reference interval is defined as the limits of the 2.5th and 97.5th percentiles of the distribution of all analyte values obtained on healthy individuals) of healthy individuals will have values outside these limits, either low or high (Figure 2.4). Begin by entering the data into an Excel spreadsheet, following the format shown in Figure 5.15. Nevertheless, it is widely accepted that the judicious use of laboratory tests, coupled with thoughtful interpretation of the results of these tests, can contribute significantly to diagnostic decision making and patient management. 5.E: Standardizing Analytical Methods (Exercises), https://chem.libretexts.org/@app/auth/3/login?returnto=https%3A%2F%2Fchem.libretexts.org%2FBookshelves%2FAnalytical_Chemistry%2FExercises%253A_Analytical_Chemistry%2F05.E%253A_Standardizing_Analytical_Methods_(Exercises), 4.E: Evaluating Analytical Data (Exercises), http://www.rsc.org/images/Brief22_tcm18-51117.pdf, http://www.rsc.org/images/brief10_tcm18-25920.pdf, information contact us at [email protected], status page at https://status.libretexts.org, ACS Committee on Environmental Improvement “Guidelines for Data Acquisition and Data Quality Evaluation in Environmental Chemistry,”. Laboratory test results may influence up to 70 percent of medical decision making.9 However, one must wonder whether the test results are being interpreted correctly, and—if not—what the impact is of incorrect or inappropriate interpretation on the accuracy of diagnostic decision making based, in part, on laboratory test results. Fortunately, most assays for a wide variety of analytes have excellent precision (ie, <5% to 10% CVa), such that the principal component among these 2 sources of variation (ie, analytical or biological) is biological variation (CVb). Plebani This discussion will begin with some general comments about approaches to ordering clinical laboratory tests, followed by real-world examples to illustrate these approaches. When a 10.00 mL aliquot of the resulting solution is diluted to 25.00 mL it gives signal of 0.235 (arbitrary units). and then use this value of kA to solve for CA. With the dotnet test command in .NET Core, you can use a filter expression to run selective tests. The following examples use dotnet test… 12. single component—such as an analyte-specific reagent coupled to a solid surface, positive and negative control solutions, instructions for conduct-ing or validating specific test meth-ods, and so on. What does analyte mean? However, other data using ROC curves to assess the ability of the tumor markers, prostatic acid phosphatase (PAP) and prostate specific antigen (PSA), to differentiate prostate cancer from BPH and prostatitis at various cutoff values is illustrated in Figure 2.3B. How do physicians increase the predictive value of laboratory tests? Troost and Olavsesn investigated the application of an internal standardization to the quantitative analysis of polynuclear aromatic hydrocarbons.15 The following results were obtained for the analysis of phenanthrene using isotopically labeled phenanthrene as an internal standard. We estimate the cost of the test to be less than \$500, which is comparable or lower than other screening tests for single … . Medical algorithms (or “decision trees”) are particularly useful in establishing a diagnosis based, in part, on information obtained from ordering the most appropriate (ie, necessary) laboratory tests. Figure 5.29 R session for completing Practice Exercise 5.7. A 10.00-g sample containing an analyte is transferred to a 250-mL volumetric flask and diluted to volume. Cardone, M. J.; Palmero, P. J.; Sybrandt, L. B. This is okay because the calculation depends on the average signal and the number of replicates, and not on the individual measurements. For many of your sample testing and screening needs, PBL can help you measure your analyte(s) of interest. Jr (d) A single sample is buffered to a pH of 3.7 and analyzed for cadmium, yielding a signal of 66.3. The sensitive biological element, e.g. Moreover, microbial contamination can lead to the degradation of the analyte as it was observed for two biomarkers thymidine and 2'-deoxyuridine in urine samples. The scales for the x-axis and y-axis are the same for each pair. Because each channel has three Glass Nano Reactors (GNRs) coated with a capture antibody, you get triplicate answers for each sample. The ROC curve (AUC = 0.66; 95% confidence interval: 0.60–0.72) in Figure 2.3A demonstrates that PSA has only modest ability in discriminating BPH from organ-confined prostate cancer. Otherwise, enjoy the free sample PDF and have a nice day browsing the web! The degree of imprecision (ie, lack of reproducibility) in the quantitative measurement of any analyte is given by the magnitude of the coefficient of variation (CV), expressed usually as a percent, obtained from multiple measurements of the analyte using the formula: %CV = (SD/mean) × 100; where mean and SD are the mean and standard deviation of the values obtained from the multiple measurements of an analyte. We will then review the important diagnostic performance characteristics of laboratory tests, how they are calculated, and a principal tool (ie, receiver-operator characteristic [ROC] curves) used to assess the diagnostic accuracy of a laboratory test at specific cutoff values for the test. You will find tolerances for different types of volumetric glassware and digital pipets in Table 4.2 and Table 4.3. 13. Figure: Representative Sample Showing at the Limit of Detection . Construct a residual plot for each data set. The LibreTexts libraries are Powered by MindTouch® and are supported by the Department of Education Open Textbook Pilot Project, the UC Davis Office of the Provost, the UC Davis Library, the California State University Affordable Learning Solutions Program, and Merlot. Remove this apparent outlier and reanalyze the data using a linear regression. When using multiple standards, the indeterminate errors affecting the signal for one standard are partially compensated for by the indeterminate errors affecting the other standards. To create a residual plot, we need to calculate the residual error for each standard. The use of neural networks in clinical and anatomic pathology is likely to expand because of their ability to achieve a higher level of accuracy than that attained by manual processes. Once a clinical laboratory test with the appropriate diagnostic accuracy has been ordered, how are the results of the test interpreted? Thus, in qualitatively comparing 2 or more tests in their ability to discriminate between 2 alternative states of health using ROC curves, the test associated with the curve that is displaced further toward the upper left-hand corner of the ROC curve has better discriminating ability (ie, a cutoff value for the test can be chosen that yields higher sensitivity and/or specificity) than tests associated with curves that lie below this curve. P.S. Figure 2.5 provides an illustration of this point applicable to the analyte, gamma-glutamyl transferase (GGT), in which the data is positively skewed. Acta 1997, 338, 19–40, and Zwanziger, H. W.; Sârbu, C. Anal. Next, we need to calculate the standard deviations for the slope and the y-intercept using equation 5.20 and equation 5.21. Immunoassays are highly sensitive and specific. The FISH probe results correlated well with the cytogenetic results in all cases; in this example the assay sensitivity and specificity were each 100%. M Provides a visual statement of this limit and may represent one of the calibration curves. Spitzauer We also acknowledge previous National Science Foundation support under grant numbers 1246120, 1525057, and 1413739. We offer it all, via a full PDF tool suite. In reality, no substance has been found to be 100% pure in its quality, so we call a substance that is found to be most pure (for some metals, 99% after electrolysis) an analyte. An appropriate standard additions calibration curve based on equation 5.10 places Sspike×(Vo + Vstd) on the y-axis and Cstd×Vstd on the x-axis.