Use of cut-off levels: the wellbeing of families hangs in the balance

Several inaccurate and misleading claims have been made recently in support of the use of cut-off levels in hair strand testing. “Cut-off levels always exist, whether we want them to or not,” was one position put forward. Another went so far as to contend that the scientific arguments made against cut-offs constitute “misinformation”. In this article, we continue to respond to claims made in support of The Society of Hair Testing (SoHT) cut-off levels. You can read our initial response here.

 

Cut-offs are guidance only – not legally defined thresholds

Generally speaking, a cut-off level is an independently defined threshold that predetermines how the results of a test should be interpreted. Where a legally defined threshold has been established, such as the analysis of blood alcohol concentration to determine if someone has been driving under the influence of alcohol, their use is entirely appropriate. This is because their purpose is to determine whether or not the person has broken the law.

In contrast, when it comes to hair strand testing, no such legally defined thresholds exist. The question to be answered in each case is, once all other factors that could have influenced the measurement are considered, whether the drug concentration found would be interpreted as consistent with active use, exposure, or a mixture of both.

The only thresholds that should be applied to hair strand testing results are the testing laboratory’s validated limits of detection and quantitation for the analyte in question, as defined below:

• LOD (Limit of Detection): The lowest level at which a drug can be detected with certainty, though not accurately quantified.
• LLOQ and ULOQ (Lower and Upper Limit of Quantification): The lowest and highest level at which a drug can be reliably quantified. In hair and nail testing, this is normally expressed as drug concentration per milligram of sample.

It is our view that any drug detected above the limit of detection should be reported as such and, if between the LLOQ and ULOQ, an accurate concentration should be provided. To do otherwise would be to withhold accredited results from the courts and to ignore UK case law that requires all results to be reported to the courts.

It is important to note that this is not the same as saying every drug concentration detected at or above the limit of detection would be ascribed to active drug use. Instead, a careful and considered interpretation should be carried out in each case. This interpretation should not only consider the measured drug concentration but also all other known factors that could have influenced the measurement.

Although their accredited limits of detection and quantitation are frequently well below the SoHT cut-offs, many other laboratories continue to report results that fall below the cut-off concentrations as “negative” or “not detected”. Such inaccurate terminology provides a misleading depiction of the true nature of the individual’s drug exposure to the court. Any drug concentration detected at or above the validated LOD used by the testing laboratory should be reported. What this concentration means is then a matter of interpretation once all other factors have been taken into consideration. This distinction is summarised nicely in this 2018 paper.

 

Flawed justification for SoHT cut-offs

One argument made in favour of SoHT cut-offs is that they prevent miscarriages of justice by reducing the number of cases reported as drug use that were actually the result of environmental or passive exposure, or so-called false positives. However, this claim only holds any weight if the sole factor considered when assessing the likelihood of drug use from hair strand testing is the quantitative drug concentration.

In reality, multiple factors can affect measured drug concentration levels, making each case unique. The idea of being able to apply a binary interpretation based on a predetermined cut-off level and that this approach somehow reduces the potential for miscarriages of justice is questionable.

Furthermore, this method introduces a different kind of miscarriage of justice: cases where a drug user’s concentration falls below the cut-off, leading to an incorrect “not consistent with drug use” finding. In other words, a false negative result. In family court proceedings, such errors can result in a child potentially being left in an unsafe environment – a tragedy that proponents of SoHT cut-offs seem willing to accept. The data highlighted in our previous article looking at the accuracy of cut-offs when applied to real cases puts the fallacy of cut-offs into context.

The flawed logic underpinning the use of cut-offs also becomes obvious when examining how hypothetical cases would be interpreted using cut-offs. Decades of research into the analysis of hair fibres for drugs has established that, amongst others, the following factors can significantly affect the drug concentration present within hair fibres:

 

• The use of permanent hair dyes or bleaching of hair fibres
• Frequent application of thermal hair straighteners
• The melanin content (pigmentation) of hair fibres
• The variance in drug metabolism between individuals
• Exposure levels to UV from sunlight

 

In addition to this, there is often an error margin of up to 20% in the analytical measurement at the cut-off concentration. Let us consider two hypothetical scenarios. The first is an individual with black hair who has denied actively using cocaine but has declared that their partner is a cocaine user, placing them at high risk of exposure. The second is an individual with bleached blonde hair who has declared no known contact with cocaine or cocaine users. Their hair strand testing results for cocaine came back at 505 pg/mg for the first individual and 495 pg/mg for the second.

Through application of the cut-off concentration for cocaine of 500 pg/mg, the first individual would be designated as a cocaine user and the second individual as “negative” for cocaine, despite the two concentrations being analytically the same when the error margin of the measurement is taken into consideration.

In addition, the first individual’s hair pigmentation and high exposure potential could provide an innocent explanation for the levels of cocaine detected. On the other hand, the bleaching of the hair by the second individual is likely to have artificially reduced the cocaine concentration to below the cut-off level. While these are hypothetical scenarios, they reflect cases commonly encountered by FTS and highlight why the application of such simplistic decision-making criteria to a field as complex as hair strand testing is scientifically unsound.

 

Cut-offs are too blunt an instrument for the family courts

FTS does not apply SoHT cut-offs to hair strand test results for children or adults. We believe forensic evidence should never be filtered before it is presented in court. It is not the job of a hair strand testing provider to pick and choose what data is relevant to judicial proceedings. When the future of a family is being decided, as is often the case when hair strand tests are instructed for the family courts, forensic toxicology data must be interpreted as accurately and with as much context and explanation as possible.

This requirement to fully and faithfully explain all findings has been made repeatedly by the courts. If cut-offs are applied without regard to the wider evidential picture, as judge Lord Justice Peter Jackson has concluded, “there is a risk that the results will acquire a pseudo-certainty, particularly because (unlike most other forms of information in this field) they appear as numbers.”

We know from our work with barristers and solicitors that miscarriages of justice have occurred when results based on SoHT-suggested cut-offs were relied on in court. We hope those currently using cut-off levels in their reports recognise the shortcomings of this practice and take steps to correct them. The wellbeing of families hangs in the balance.