Expansion of the EZSTATSG1 Microsoft Excel Tool for ANSI/ASB Standard 036 Method Validation by the Addition of Weighted Quadratic External Calibration Models Utilizing Five-, Six- or Seven-Point Curves

Abstract

The first generation of this Microsoft (MS) Excel (Redmond, WA, USA) tool for method validation, EZSTATSG1, was designed for methods utilizing only linear calibration curves requiring seven calibration levels, and quadratic calibration models were not supported. This significantly improved version, EZSTATSG2, includes all of the features of the original template such as weighted linear calibration models, bias and precision data, dilution integrity and ion suppression. New features in this version include flexible five-, six- or seven-point calibration curves, six weighted quadratic calibration models, standardized residuals by use of frequency plots overlaid with the normal distribution function along with five-number summary data and processed sample stability. The implementation of Visual Basic for Applications in Excel UserForms prevents accidental alteration of existing formulas and also ensures that pertinent cells are relocked every time a file is reopened. The quadratic models feature the 95% confidence intervals for checking the significance of the second order term and are fully-characterized by providing the equations for the axis of symmetry, directrix, and coordinates for vertex and focus. Example data of α-hydroxymidazolam demonstrates that the quadratic calibration curves fit the data more adequately than the linear models for this method. This second-generation tool summarizes all of the validation parameters of a method for both linear and quadratic calibration models. Models with the lower average sum of relative errors and higher R2 values are color coded green, indicating the likelihood of a better fitting model. Like the EZSTATSG1, the redesigned EZSTATSG2.xltm MS Excel self-actuating validation tool and a completed PDF example are available to the scientific community for download as Supplementary data. Updates to the template can be found at https://www.EZSTATS4validation.com.

Strategies to detect trace methamphetamine contamination on hard surfaces: Assessing realistic performance of commercially available presumptive tests and a laboratory‐based LC–MS/MS method

“…Method validation calculations were performed using the EZSTATSG2 workbook by Sofalvi et al. [28], which includes a quadratic calibration model. Extraction recovery was 8.91% at 0.75 μg/sample and 70.72% at 7.50 μg/sample.…”

“…Within-and between-run CVs were also automatically calculated in EZSTATSG2 [28] from an ANOVA: single-factor analysis.…”

“…Method validation calculations were performed using the EZSTATSG2 workbook by Sofalvi et al. [28], which includes a quadratic calibration model. Unless otherwise noted, samples were prepared as described in Section 2.3.…”

“…Calculated concentrations from triplicate analysis of five extractions were used to find the calculated mean, percent bias, within‐run coefficient of variation (CV), and between‐run CV. Bias was automatically calculated in EZSTATSG2 [28] as the difference between the mean and expected concentrations, divided by the expected concentration. Within‐ and between‐run CVs were also automatically calculated in EZSTATSG2 [28] from an ANOVA: single‐factor analysis.…”

“…Method validation calculations were performed using the EZSTATSG2 workbook by Sofalvi et al [28], which includes a quadratic calibra- Method sensitivity was identified by assessing the LOD and LLOQ, which were set to 0.05 μg/sample and 0.10 μg/sample, respectively.…”