Contact

Daria Pašalić
Editor-in-Chief
Department of Medical Chemistry, Biochemistry and Clinical Chemistry
Zagreb University School of Medicine
Šalata ul 2.
10 000 Zagreb, Croatia
Phone +385 (1) 4590 205; +385 (1) 4566 940
E-mail: dariapasalic [at] gmail [dot] com

Useful links

ZR1-1

 

Samošćanec K.ZR1-1: Validation of qualitative chromatography methods. Biochemia Medica 2009;19(Suppl 1):S85-S86.
University Department of Chemistry, Sestre Milosrdnice University Hospital, Zagreb, Croatia
Corresponding author:ksamosc1 [at] gmail [dot] com
 
Abstract
 
Validation is a procedure of testing the reliability, accuracy and precision of a method. Being closely related to the quality of the results obtained, it is an important element of all methods of measurement. These methods are characterized by parameters evaluating analytical acceptability, which should be in agreement with the previously defined requirements. All these parameters depend on the type and characteristics of the method. Methods are conventionally classified into qualitative and quantitative methods. By definition, qualitative methods include those analyses that point to the presence or absence of the analyte measured directly or indirectly in a strictly defined sample quantity. The result is obtained by reading off the measurement signal and comparing it with the reference or predetermined value. The result is binary (yes/no; present/absent). The reference material may be external or internal. External reference material is well characterized (certified reference material), whereas internal reference material may be a properly processed sample of known values or a commercial sample found in commercial test kits.
Qualitative parameters should be carefully identified and chosen according to the method requirements. They are substantial for the information to be obtained, e.g., verification of traceability, determination of result related inaccuracy (false positive/false negative), sensitivity and specificity, and selectivity: interferences, detection limits, unreliability area, and method robustness.
According to Eurachem definition, traceability is a characteristic of the measurement result or standard value related to reference material, usually national or international standards, through a continuing sequence of comparisons. This means that traceability can be assessed by comparison of results obtained by the reference method or by use of the reference material.
In qualitative analysis, sensitivity and specificity represent the test potential to differentiate true positive (sensitivity) from true negative (specificity) samples. High test sensitivity and specificity ensure false results to avoid. Selectivity should also be taken in consideration, as it is closely related to specificity and false positive results. Poor test selectivity implies the impact of reagent and sample matrix, i.e. the higher the selectivity, the greater the likelihood of analyte identification in the sample. The manufacturer is also obliged to provide a list of chemically similar substances that do not induce cross reactions.
As the result of a qualitative method is binary, the parameter of measuring uncertainty in quantitative methods is substituted by the area of unreliability in qualitative analyses because it better describes the area with a higher probability of error. In qualitative tests with a sample of known concentration (calibrator/control), the area around borderline value (cut-off, COV) is best defined. The area of unreliability is defined by the sensitivity and specificity parameters obtained. The number of false (+) positive and false (-) negative results can be evaluated by modifying COV, i.e. reducing the number of false (-) negative results by COV decrease, or reducing the number of false (+) positive results by COV increase. Robustness is a basic analytical characteristic describing the stability of binary result against slight changes in experimental conditions (change in sample quantity, pH, time of performance, temperature, solvent percent, etc.).
ZR1-2
Granić P. ZR1-2: Validation of quantitative chromatographic methods. Biochemia Medica 2009;19(Suppl 1):S86.
Division of pharmacokinetics and analytic toxicology, Zagreb University Hospital Center, Zagreb, Croatia
Corresponding author:pgranic [at] kbc-zagreb [dot] hr
 
Abstract
 
Chromatographic methods are commonly used for the quantitative and qualitative analysis of drug substances, drug products and compounds in biological fluids.
Validation of a method is the process by wich a method is tested by the developer or user. Methods shuold be reproducible when used by other analyssts, on other equivalentequipment, on other days or locations. The process of validation should be early in the development cycle, but validation should be continued form of revalidation with method changes.
Parameters for validation of quantitative chromatographic method are: accuracy, detection limit and quantitation limit, linearity, precision, reproducibility range, recovery, sample solution stability, specificity/selectivity.
ZR1-3
Gašljević V.ZR1-3: Measurement uncertainty and method validation. Biochemia Medica 2009;19(Suppl 1):S87.
Croatian Metrology Society, Zagreb, Croatia
Corresponding author:visnja [dot] gasljevic [at] hmd [dot] t-com [dot] hr
 
Abstract
 
In today global world accurate, reliable and traceable measurement results are of vital importance. Therefore, the standards for laboratories like ISO/IEC 17025 and ISO 15189 demand a metrological approach to measurement processes requiring validation/verification of methods, assurance of metrological traceability, knowledge of measurement uncertainty, monitoring of trends etc.
Making the right decision on whether or not a method is adequate for a stated use (validation), as well as a realistic estimation of measurement uncertainty are only possible when a measurement process is well known. Its ‘goodness’ is characterized in terms of the random and systematic errors that affect the measurements. Experiments conducted in order to validate a method can give good insight into their magnitude and also into their sources and therefore can be used for measurement uncertainty estimation. This article describes how target measurement uncertainty can affect the establishment of the validation criteria on method parameters like: repeatability, intermediate precision, bias and how to use validation experiments for measurement uncertainty estimation.
ZR1-4
Nikolac N.ZR1-4:Statistical analysis in validation of methods in analytical toxicology. Biochemia Medica 2009;19(Suppl 1):S87-S88.
University Department of Chemistry, Sestre Milosrdnice University Hospital, Zagreb, Croatia
Corresponding author:nora [dot] nikolac [at] gmail [dot] com
 
Abstract
 
Laboratory validation of analytical procedures includes confirmation of inaccuracy, imprecision, linearity, interferences and method comparison, if a parallel system is already present in the laboratory.
Inaccuracy is expressed as percentage of deviation (bias) and includes comparison of results with known target values. Bias is a measure of systematic error. Recovery test discovers proportional systematic errors by measuring samples spiked with increasing known amount of analyte.
Imprecision is determined by repeated within-run and between run measurements. For the quantitative methods, mean, standard deviation and coefficient of variation are calculated and for the qualitative methods, percentage of results that deviate from the most common result. Imprecision measurement reveals the existence of random errors.
Linearity confirmation (for quantitative methods) includes dilution measurements in order to determine concentration range with linear dependence. Results are analyzed with linear regression and coefficients of regression and determination are calculated. However, in some cases, very high coefficients of determination (0.98) could be obtained, but the dependence is not linear but curved, which could be established by calculating residuals.
Comparison of two parallel analytical procedures for quantitative methods is analyzed with Bland-Altman graph and Passing-Bablok regression.
There are several ways of presenting Bland-Altman graph. In the most commonly used approach mean values of two measurements on x-axis are plotted against the differences of two measurements on y-axis. If all measurements are within ± 1.96 SD, there is a good conformance between methods. This type of graph is suitable for detecting systematic errors and outliers.
Passing-Bablok regression is a statistical procedure for comparison of two variables in cases when it is not known which variable is dependent and which one is independent. As a result, line equation is calculated, with slope and intercept and their corresponding confidence intervals (CI). Confidence intervals are extremely important for interpreting results: if CI for intercept doesn’t include value 0, there is a constant difference between measurements. Also, if CI for slope doesn’t include value 1, there is a proportional difference between measurements.
Also, Pearson’s coefficient of correlation should be calculated. Interpretation of correlation coefficients in analytical validation is different than in biological systems. For acceptable correlation, coefficient of correlation should be at least 0.95.
Comparison of two qualitative methods is much simpler: result is expressed as a percentage of measurements in which the same result is obtained.
Validation in analytical toxicology also includes some specific parameters: selectivity, range of measurements, limit of quantitation, limit of detection and robustness.