Figure 1 above: Statistics from the Substance Abuse and Mental Health Services Administration (SAMHSA) details drug use across the 50 states based on the 2006 and 2007 National Surveys on Drug Use and Health, which involved interviews with over 135,000 people around the country.

(For more on the statistical breakdown of geography of non-prescription or illicit drug use by state please visit: http://mapscroll.blogspot.com/2009/07/more-on-geography-of-drugs-in-us.html)

In terms of drugs of abuse (DOA) analysis performed today in the US they fall into three basic chromatographic categories:

1 Immunoassay

1.1 A general description of this method can be found here: https://www.thetruthaboutforensicscience.com/a-screening-test-for/
1.2 This method is most decidedly of poor selective power and far from specific.
1.3 It is at best a screening device. (Remember that in a screening device you substitute qualitative and quantitative certainty for speed and ease of use so that untrained folks can conduct the testing.) A screening test is characterized by a high rate of false positive and a low rate of false negative. This is confirmed, as attendees of the recent webinar we conducted learned, through Dr. Valentine’s fine published research (finding low false positive on MJ at or about 5% when subsequently analyzed by GC-MS, but really high false positive approaching 50% in the case of amphetamines when subsequently testing by GC-MS). At worst it is a deceptive source of conviction.
1.4 At best the results are termed as semi-qualitative and in some case semi-quantitative (meaning it exceeded the limit of detection-i.e., the cutoff)
1.5 A motion in limine should always be filed to keep this method out of court as the probative value s low, but its misleading prejudicial effect is high. This should be the case even if secondary testing is performed to confirm.

2 Gas Chromatography (GC) based

2.1 GC requires the analytes to be analyzed in the gaseous state as they elute through the column. This is the limitation of this assay.  This happens in one of two ways:

2.1.1 Flash vaporization at the time of injection in the injector port, or
2.1.2 Headspace sampling

2.2 Gas Chromatography with Flame Ionization Detector (GC-FID)

2.2.1     As mentioned previously (Fool’s Gold or Real Gold? HS-GC-FID is it specific or selective or neither?) the simple fact that there are so very many compounds in the world that to say that any given column is truly specific and not merely selective is scientifically irresponsible… Bottom line, any peak identification on FID is characteristic or consistent with an analyte of interest, but is not necessarily unique or exclusive to the analyte of interest.  This is especially so in the case of DOA analysis. Sometimes the difference between a pharmacologically active compound and in inactive one is so very small. A lot of published manufacturer chromatograms using GC-FID in DOA analysis reveals some really bad chromatography (poor resolution, peak tailing, fronting, elevated baseline error). The likelihood of co-eluting substances in a DOA analysis is very, very real or even a chiral one for that matter. This is why FID use in DOA analysis is disfavored.

2.3 Gas Chromatography with Mass Spectrometry (GC-MS) or Gas Chromatography with Mass Spectrometry/Mass Spectrometry (GC-MS-MS)

2.3.1 We lose stereo-selectivity between chiral compounds unless special columns are used.
2.3.2 It still comes down to the GC part where the column is as that is where separation occurs. To have good MS, one must be certain that there is total and complete separation first. Again, even as the manufacturer published chromatograms  show this is not a given.
2.3.3 GC-MS is further limited by the method of interpretation of the spectra produced. In essence the spectra produced of the unknown that is analyzed is compared against standards that are kept in the computer library. Unfortunately GC-MS work has devolved into simply computer-assisted pattern recognition by undertrained folks who have no idea of acid-base chemistry and way spectra appear as they do (logical loss analysis) where nowadays the computer takes the unknown and compares it against the known in the library to arrive at probable matches that the analyst has to judge based upon the computer’s algorithm derived match factor score and probability match to make a judgment call of the unknown as a specific drug. This is particularly difficult and subject to interpretative error by the analyst in the case of non-pharmaceutical street drugs such as methamphetamine, heroin or cocaine in that they are not part of a formal industrial practice where QC processes are in place to minimization variations between presentations of the drug as they are produced.

2.3.3.1 The strength of any spectral library is in its source. The source of the standard that the unknown is compared against is in generally two categories:

2.3.3.1.1 A traceable reputable source such as NIST or Sigma Aldridge, or
2.3.3.1.2 A local source (the lab folks just pick some street sample seized off someone and call it heroin). This is very dangerous for obvious reasons as it has not be adjudged and confirmed as actually being the drug, but rather is assumed to be so.

3 Liquid Chromatography (LC) based

3.1 Liquid Chromatography requires that it be soluble to be analyzed. Again separation is key and essential. The advantage is that there is typically less preparation necessary in LC work, especially for DOA analysis.
3.2 LC-UV (Vis or DAD) is disfavored in general because of the lack of specificity.
3.3 LC-MS (MS) is great but it requires a lot of understanding of the spectra produced and requires a lot of elucidation to be performed by the analyst as it is truly a judgment call as there can be no spectral library comparison such as the case in GC-MS. It requires thought and true understanding of acid-base chemistry as there are a lot of complicated re-arrangements that can occur.