Application of statistical design of experiments to assess pre-treatment parameters in forensic hair analysis for amphetamine

• NCJ Number: 300958
• Journal: Forensic Chemistry Volume: 20 Dated: 2020
• Author(s): Jennett Chenevert Aijala; Wensong Wu; Anthony P. DeCaprio
• Date Published: August 2020
• Grant Number(s): 2018-75-CX-0037; 2017-IJ-CX-0005; 2014-DNR-4918
• Agencies: NIJ-Sponsored
• Publication Type: Research (Applied/Empirical)

Annotation

This project used a 2 4 factorial block DoE design to compare the efficacy of parameters in the decontamination and extraction of amphetamine from human head hair.

Abstract

Toxicological analysis of human head hair for detecting drugs of abuse requires multiple pre-treatment steps before analysis, such as washing to remove external contamination, pulverization/homogenization of the sample matrix, and subsequent extraction of the drug from matrix components. The predominant approach to developing and assessing the efficacy of hair pre-treatment protocols uses a traditional one factor at a time (OFAT) approach, in which one independent variable at a time is changed to observe the effect on the dependent variable. An alternative approach to assessing pre-treatment protocols is statistical design of experiments (DoE), which involves the systematic variation of all independent variables (i.e., pre-treatment conditions) simultaneously, allowing for the variation in the dependent variable (i.e., extraction efficiency) associated with each independent variable and combinations thereof to be observed. In the current project, DoE studies found that the most effective pre-treatment conditions included 1-percent SDS washes followed by methanol washes for 30 s each time, milling the hair into a powder, and final extraction of drug with 12.5 μL of 2 mM proteinase K solution/mg of hair. The most notable result of these DoE studies was that the combinatorial effect of all parameters was consistently significant, confirming that it is necessary to consider pre-treatment protocols as complete sets of procedures, as is done in DoE, compared to optimizing individual factors using an OFAT approach. (publisher abstract modified)