Forensic identification and confirmation
The forensic toxicology profession and the courts have increasingly demanded that the identification of a substance be beyond reason able scientific doubt. The principle has long been established that forensic identification of an analyte requires the use of two techniques that employ different physical and chemical principles (SOFT/AAFS Guidelines Committee 2006). This approach has the advantage that two completely different scientific techniques are used, which are supportive in arriving at a positive result. It has been argued that forensically acceptable detection and identification of an analyte can be achieved by a single extraction of a postmortem sample followed by GC-MS analysis of that extract. The argument is that GC-MS is a combination of two very different analytical methods – separation of the mixture and determination of retention time being one, and production of the mass spectrum being the other. While this approach is reasonable, it produces a forensically acceptable confirmation only if laboratory contamination and, if possible, contamination of the original specimen can be ruled out. Therefore, at a minimum, the drug should be detected using two different extracts of the same specimen. This is often accomplished incidentally, because separate extracts may be prepared for the initial drug screen and for a subsequent quantitative analysis. Even better is detection and identification of the substance in two different specimens. An example might be detection and identification in a urine specimen, followed by quantification in blood and one or more other tissues. This latter approach increases confidence in the result by ruling out a false-positive finding through contamination from glassware or, indeed, one of the specimens. Identification of a drug by the use of two similar methods, such as two different immunoassays, is not acceptable, even though such tests may employ different endpoint reactions (e.g. fluorescence polarisation immunoassay (FPIA) and enzyme immunoassay (EIA)). The reason is that the antibodies used may have similar cross-reactivities, even though the designs of the immunoassays as a whole differ. Similarly, identification of a substance on the basis of different retention times (or different relative retention) on two different GC columns is rarely acceptable unless it can be shown clearly that the columns differ markedly in their retention and discrimination characteristics. It should be borne in mind that retention time in GC is dependent on the distribution constant (K) of a substance, that is how it partitions between the stationary phase and the mobile phase, which in the case of GC is a gas. Hence if K is the same for a drug and another substance present in the sample, even if a longer GC column with the same stationary phase is used for the analysis, the two substances will still have the same retention time on the longer column. Thus the A caveat to this approach involves simultaneous detection and quantification of ethanol and other specific analytes, such as carbon monoxide. Although two independent methods may be used to identify ethanol, it is still generally accepted that a single GC method is forensic ally acceptable. Analysis of alcohol using two different columns with non-correlating stationary phases and internal standards provides much greater confidence. The quality of GC methods used in most laboratories, and the fact that few other compounds are likely to be present at such a high concentration, means that erroneous identification of ethanol is emphasis is on selecting a second GC column for which K is different for the drug and the interfering substance.
unlikely. The few other compounds include methanol, isopropanol and acetone, for which separation from ethanol in the analytical system must be demonstrated. Similar arguments can be made for carbon monoxide, especially when the history clearly indicates that carbon monoxide poisoning is likely. However, in rare circumstances grossly elevated carbon monoxide concentrations may be detected where there is not an obvious source of the poison. In such instances, confirmation by an independent technique is highly desirable (e.g. headspace GC or palladium chloride via a Conway diffusion cell).
