Replicating and Validating Lab Results
Best practices for analytical method validation
A critical requirement of all laboratories is the need to replicate results. While some industries may not require method validation or method transfer between laboratories, an effective method validation process demonstrates reliability and repeatability of the analytical method and establishes the controls needed to replicate results. Quantitative methods generally present more replication challenges. This paper uses chromatography as a proxy, but the principles apply to all quantitative methods.
Prior to result generation or method validation, a method must be designed to meet the desired need (fit for purpose). This requires an understanding of how the results will be used. Effectively, what type of samples will be tested, what is the expected concentration to be found in the sample, and what is the acceptable error in the quantitative value? Answers to these questions may be provided by the individual who requested the testing or may require you to seek guidance from regulators or standards in your industry1,2. The outcome of this process is a written method that can be followed in a consistent manner. The written method should include the scope of the method, requirements for reagents, standards, equipment, and personnel, and example calculations, including all conversion factors and reporting requirements. The method may also include preventive measures, such as direction to flush a column prior to use or a requirement to clean all laboratory glassware in a certain manner. The method should be treated as a living document and updated at the end of the validation, and as needed throughout the lifetime of the method.
Method validation steps
Method validation is composed of sets of experiments to assess different method performance parameters, such as specificity, linearity, range, accuracy, precision, quantitation limit, detection limit, robustness, and system suitability. A risk assessment should be performed to determine if all or a portion of these experiments are required to adequately understand method performance. However, specificity must be evaluated in all cases, as it is useless to validate any method if it is not specific for the targeted analyte. Before starting any quantitative performance or validation activity, confirm that authentic reference material is available with the required purity.
Specificity is the ability of the method to distinguish the analyte of interest from other analytes present in the sample or in the matrix1. Specificity should be differentiated from selectivity of the method. Methods intended to quantitate multiple analytes must be able to differentiate each of the components of interest. Ideally, a sample matrix free of the analyte of interest is available and analyzed as part of the validation. Any detectable amount of the analyte of interest present in the target-free matrix is addressed as matrix related interference. In addition, compounds similar to the analyte of interest should be spiked into the sample matrix and analyzed to assess potential interference with accurate determination. For example, a method designed to assess the quantity of isopropanol in hand sanitizer may need to distinguish isopropanol from n-propanol, ethanol, and propanal. A method designed to quantitate the amount of limonene in a citrus-scented shampoo may need to distinguish between limonene and limonene oxide, or between the D and L limonene enantiomers.
“Accuracy experiments provide an important source of information to provide error estimates of reported results.”
Accuracy experiments assess agreement between the accepted or known amount of analyte in a sample and the measured amount using the method. Ideally, a clean matrix is spiked with the analyte of interest and the amount of analyte is measured. It is important to determine the amount of analyte in the clean sample and use this value to correct the measurement in the spiked sample prior to calculating accuracy results. The uncertainty in replicate measurements provides an understanding of the error associated with the employed method. If the error is greater than the acceptable level for the intended use, the method must be optimized, and the experiment repeated. Accuracy experiments provide an important source of information to provide error estimates of reported results.
Range is the demonstrated upper and lower concentration over which the method has been demonstrated to show acceptable accuracy, linearity, and precision. Range should cover the expected range of results. For example, if a product is produced to contain aluminum and must contain three to five percent by weight when released by the manufacturer, an appropriate range for the quantitative method might be two to six percent. Results outside the demonstrated method range may not be accurate and should not be reported. Rather, results should be reported as less than the lower concentration limit if the results are below the lower limit, or dilutions performed and samples reanalyzed if above the upper limit.
Precision assesses the degree of scatter from multiple series of measurements. This is achieved by measurement of multiple replicates of the same sample on multiple days, by multiple analysts, on multiple instruments and at multiple physical sites or locations. If your company has a single instrument at a single site, you may be able to adequately establish precision from two analyses (day one analyst one, day two analyst two). However, if you have several instruments at multiple sites, a matrix should be designed to achieve adequate evaluation. The precision experiment can be repeated to qualify new analysts, new analytical systems from the same or different vendors, and new laboratory sites.
Robustness evaluates the impact of small but deliberate changes to the method. Appropriate design is achieved by understanding the critical aspects of the method and analytical system. Some traditional evaluations, such as altering inlet temperature for a GC-based method, are generally not useful. However, changes to sample preparation, such as the amount of time for derivatization or evaluating separation of a critical pair of analytes on multiple columns, is useful. Robustness experiments provide understanding of which parameters require control and the acceptable range for each parameter. The outcome of robustness experiments is incorporated into the written method. For example, derivatize each sample at 75 +/- 5C for 60 +/- 10 minutes.
System suitability is the last aspect to be addressed here, but it may be the most important. System suitability is designed to assess critical method aspects and is repeated each time the method is performed. Examples of suitability may include the preparation and analysis of multiple standards to ensure solutions are prepared correctly, the evaluation of solutions which do not contain the analyte being tested to ensure the system is free of interference, the evaluation of samples with a known amount of analyte to assess measurement error, as well as the measurement and comparison of diagnostically relevant, technology-specific attributes.
“System suitability is designed to assess critical method aspects and is repeated each time the method is performed.”
Examples of diagnostically relevant, technology-specific attributes for chromatographic methods include retention time, peak shape, column efficiency, critical resolution from a known marker (typically a compound related to the analyte being tested), and injection repeatability. System suitability results should be collected to provide historical reference and to understand when the analytical system and method are performing as expected, or if an error or issue must be addressed prior to release of results.
Validation of an analytical method will ultimately improve laboratory efficiency by identifying issues at the first incident, providing a framework for qualification of new analysts, instruments, and laboratory installations, as well as ensuring control of critical method parameters.
Gyuri Vas is an analytical chemist at Intertek Pharmaceutical Services in Whitehouse, NJ. He has more than 20 years of experience in research, development, validation, and laboratory management. Gyuri is an industry recognized expert for extractable/leachable and genotoxic impurity testing, and trace level method development and validation.
Louis Fleck is the trace organic analysis manager at Intertek Pharmaceutical Services in Whitehouse, NJ. He has more than 20 years of experience performing and managing activities that directly support the pharmaceutical industry, including chromatographic method development and validation, and extractable and leachable testing.
1. International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use. ICH Harmonised Tripartitate Guideline; Validation of Analytical Procedures Q2 (R1), 1994
2. Validation and Peer Review of US Environmental Protection Agency Chemical Methods of Analysis, US Environmental Protection Agency, FEM Document Number 2005-01 October 14, 2005 REVISION: February 3, 2016