This is a guest post by Lucy Taylor. Lucy is a pupil barrister at Coram Chambers.

When it is disputed whether a person uses drugs, the family court can – and often does – order hair strand testing to determine the issue. This article draws upon the work of lawyer Sarah Branson and drug tester Paul Hunter, who have explored the limitations of hair strand testing and highlighted the need for caution in relying on hair strand test results (See Recent scientific developments in hair strand testing and racial bias in current practices of hair strand testing. I’m going to call this ‘the Fam Law article’). Even so, hair strand testing continues to be the ’go-to’ when there is a question over drug use. Despite an increasing recognition of the possibility of mistakes and errors in the production and reporting of hair strand test evidence, this evidence is still regularly relied upon by the courts as evidence of substance use. These problems are often underestimated, which can have disastrous consequences.

In this blog post I will give a brief overview about what hair strand testing is and how it is used, before attempting to answer two questions:

• (i) why is hair strand testing problematic; and
• (ii) what is the solution?

Where I have referred to a study or statistics, these are sourced from the Fam Law Article in which full references are provided for those who wish to read in further.

How does hair strand testing work?

When a person uses drugs they enter the blood stream, producing metabolites, and the drug and its metabolites become incorporated into the hair as it grows. A metabolite is a small distinctive molecule which is produced during the digestion or break down of substances, including drugs, and which can be used to identify ingestion of a particular drug (as opposed to external contamination of the hair from exposure to drugs).

To test for drug use, a sample of hair is taken from the head and is cut into segments so that the assessor can identify drug use over time and estimate usage levels. For example, drugs and their metabolites found in hair close to the scalp are more likely to indicate recent drug use and those found further away from the scalp a more likely to represent more historic drug use. Once hair samples arrive at the laboratory, they are washed to remove any drug contamination from the environment and any residues from hair cosmetics. Once the hair has been tested, results are provided on the basis os ‘high’, ‘medium’ or ‘low’ levels of drug found in the hair. If the amount of drug found is above a standardised level (the ‘cut-off’ point), the test will be ‘positive’ for drug use, which will be reported as ‘chronic drug use’. If the cut-off point is not met, the tested person will receive a ‘negative’ result, which will be reported as ‘no evidence of drug use’. As will be explained, such labelling can be misleading.

How are hair strand test results used?

As Sarah Branson and Paul Hunter note in their article, a positive hair strand test result continues to be taken by the family court as an “almost irrefutable conclusion” of the use of drugs. For instance, in one case a parent “adamantly denie[d]” using drugs, and all other evidence (including 57 randomised urine tests taken over a period of 24 weeks) suggested that they did not use drugs. Agreement could not be achieved between the five hair strand test experts; the levels of drug found in the hair as reported by one laboratory were double that of another laboratory. Yet, the fact that the parent received a positive hair strand test was sufficient for the court’s finding that the parent was a drug user (see Re H (A Child: Hair Strand Testing [2017] EWFC 64– I’ll call this Re H below). Such conclusions are reached despite a growing body of scientific and medical opinion emphasising the need to exercise caution in interpreting hair strand test results, and an increasing recognition of the possibility of errors in the hair strand testing process (see, for example, Birmingham City Council v CD and SP and FW and X, Y & Z [2023] EWHC 748).

What are the problems?

As noted by Lord Justice Peter Jackson in Re H, hair strand test results can be affected by

“variables in relation to hair colour, race, hair condition (bleaching and straightening damages hair), pregnancy and body size. Then there are the variables inherent in the testing process” (para 41).

In short, there are difficulties at every stage of the hair strand process, from the very beginning at the sampling stage, through to the processing stage and the eventual recording of results.

Problems at the sampling stage

The hair sample taken for testing can itself determine the test result. According to Paul Hunter, drug levels found in respect of a hair strand from the side of the head will often be several times higher than the drug levels found in top/back of the head. Moreover, the tested person’s hair colour plays a large part in the results. As pointed out in the Fam Law Article, in one study, the level of drugs found in black-haired participants was over 15x higher than that found in those with ginger hair, although they were given the same amount and type of drug (opiates, in this case) over the same period. There are also further marked differences between those with same colour hair – for example, those with black Asian hair produce higher drug levels compared to those with Caucasian black hair.

And it’s not just colour than can impact results. As explained above, hair strands are sectioned by testing companies to identify the client’s use of drugs over the past months. This done by reference to a standardised growth rate of 1cm per month, as per the guidance of the Society of Hair Testing (SoHT), yet hair growth rates vary between people, and in particular, as between people of different ethnicities. Branson and Hunt state that hair strand testing companies therefore currently fail to acknowledge that Asian hair grows at a rate of about 1.4cm a month, whilst Caucasian hair grows at around 1.2cm a month, and African hair grows at around 0.9cm a month.

All this goes to highlight the difficulties with standardised hair strand testing if results are not sufficiently examined with respect to the contextual background. With the present outdated reporting methods used today by the significant majority of hair testing laboratories, someone with ginger hair may receive a ‘negative’ test whilst someone with black hair may receive a ‘positive’ result, when the two have in fact taken the same amount of drugs at the same frequency. Similarly, someone with Asian hair may be reported to have achieved abstinence, whilst someone with African hair may not, even though both have in fact stopped using drugs a significant period of time ago. Even those with the exact same type of hair may have differing results if one person’s hair was taken from the side of their head and the other’s hair was taken from the top.

The impact of heat treatments on hair also needs to be considered. This is particularly problematic when the court is testing for the use of crack cocaine. Cocaine users who heat their hair (for example with thermal hair straighteners or curling tongs) may test positive for crack cocaine even when they have not ingested crack, since the compound used to test for crack (AEME) is one which is formed when cocaine is heated and then detected in the hair strand testing process. Since AEME is not a metabolite of crack cocaine only a by-product from heating cocaine, it should not be used to confirm crack cocaine use. The result is the possibility of misreporting crack cocaine in many cocaine users (see Fam Law Article). In context, this is an issue since crack cocaine is considered more harmful than cocaine, and so is considered to have a greater impact on parenting capacity. This means that someone who uses heat on their hair could have their children taken into care on the basis of their crack cocaine usage, when in fact they only use cocaine.

More generally, dye, bleach and chemical treatments can all lead to false negatives and positives. According to a study, hair dye or bleach removes around 60% of the drug from the hair sample and can remove even more depending on usage (see Fam Law Article). Other difficulties pertaining to the sampling include – but aren’t limited to – pregnancy, body size, UV exposure, hair washing and conditioning frequency, brands of shampoo and conditioner, swimming, shaving, hair style and living environments. Clearly, therefore, there are a range of factors which can influence a person’s test result. Until those scrutinising hair strand testing results consider each of these factors, it is incredibly difficult to justify the amount of weight placed on hair strand test results in court proceedings. Paul Hunter says therefore that reporting ‘the use of treatments’ is not enough, and the impact of treatments on the test is often not explored.

The Process and Results

Mistakes can be made at several stages of the hair strand testing process. One example is the stage at which samples are washed. The SoHT is clear that “several approaches” can be used to distinguish between external contamination and drugs incorporated into the hair through consumption (SoHT Guidelines PDF). It explains that “organic solvent such as methylene chloride or acetone will remove only surface contamination whereas aqueous solutions or methanol will swell the hair and extract drugs from within the hair matrix”, yet some testing companies [https://lextox.co.uk/blog-posts/testing-the-wash/#] still use methanol on the basis that “the use of methanol to wash hair samples is considered to be the most effective for decontamination”.

With different companies preferring different techniques, a standardised test procedure is impossible, and those having their hair tested may be more or less likely to achieve a negative result, depending on the company they (or the lawyers) choose to carry out the test. This was demonstrated in Re H, wherein three different testing companies provided results, and the results from one of the companies was significantly higher than those from other companies for the matching hair sample (see para 14).

Even after the testing process is completed, results are arbitrarily presented as falling either above or below a standardised ‘cut-off level’. Where the level of drugs found in the hair is below the cut-off level, the client will receive a negative result and as such, will be taken to not use drugs. If the client receives a positive result, since the level of drugs found is above the cut-off level, the client will be taken to be using drugs, and is at risk of being deemed unable to parent sufficiently. The issue here is that cut-off levels have been standardised by the SoTH, which, in deciding upon cut-off levels, did not take into account the individual factors and influences discussed above. Instead, the population was taken as a collective. This sparks problems where two different people fall either side of the cut-off, because of their hair colour, or chosen testing company, or hair style, despite using the same type and amount of drug. One could have their children taken into care, whilst the other may not.

Conclusions: the solutions?

It can be seen, therefore, that the whole hair strand testing process is littered with difficulties and potential injustices, from the taking of the hair to the distribution of results. Yet, this issue is paid insufficient attention both within the industry, and within court proceedings.

In the long term, we must work towards a greater understanding of the importance of contextualising hair strand test results and treating them with caution, at the very least. All prevailing influences and relevant factors should be established and considered in each case to ensure that test results are contextualised, so their interpretation is more accurate. This is not, however, a new realisation within the science world. For years, experts have been clear that “toxicologists reporting hair strand analysis results should move away from simply providing results by the application of cut-offs, to a process of assisting the Courts as experts by providing evidence-based opinions” (see Fam Law Article), and this approach has, albeit only to a limited extent, found its way into some courts. Nonetheless, it remains clear that any recognition of the fact that hair strand test results should – to put it plainly – be taken with a pinch of salt, is far from widespread. Until there is more consistent acknowledgement and understanding of the limitations of hair strand testing, we must push to develop the approach to analysing test results.

The authors of the Fam Law article point out that: “one hair testing company in the UK [forensic-testing.co.uk] has now introduced a test… which provides an objective assessment of the amount of hair damage in each segment of hair”. This could assist more in-depth analysis of the reliability of hair strand test results in relation to drug use over time. Also, as emphasised by Paul Hunter, the courts should be instructing experts to assess the evidence, to provide evidence-based opinions rather than binary answers based on standardised ‘cut-offs’. Hunter considers that relying on experts who are working more closely as part of the process, rather than some distanced laboratory, will move the courts towards greater understanding of hair strand testing and its results. Whilst the instruction of experts is time consuming and expensive, the far greater cost is the injustice otherwise caused to those whose children are taken away from them on a false basis.

In the short term, it should be remembered that hair strand testing is not the only form of testing available. It is often favoured over, for example, nail testing, since it can be segmented and therefore provide an indication of time scales, but, as explored above, this can often be unreliable anyway considering, for example, the lack of acknowledgement of people’s different hair growth rates. For this reason, it may be argued that nail testing should be used instead. Another option is urine sampling. It is arguably not enough to order alternative testing methods alongside the well-established use of hair strand testing, since courts favour hair strand test results in reaching their decisions, even where urine and nail samples point to an alternative conclusion, and Re H supports this proposition.

Perhaps most importantly, professionals (including those within testing companies and legal professionals working in the courts) have a responsibility to remind each other and the court of the limitations in hair strand testing. Lawyers should not be afraid to query or challenge results, and to point out the limitations of the evidence or complexities of interpretation. A more cautious approach to interpreting test results should reduce the potential injustices that may otherwise occur. Therefore, Paul Hunter advocates for an approach similar to that of Lord Justice Peter Jackson in Re H, wherein he reminded us that:

• “It would be artificial to require valid data to be struck from the record because it falls below a cut-off level when it may be significant in the context of other findings” (para. 47)
• Results should be accompanied by “A clear statement that the description is of the level of substance found and not of the level of use, though there may a broad correlation” (para. 59)
• “Descriptions such as “positive”, “negative”, “indicates that” and “not detected” can be used and understood vaguely or incorrectly… reporters should define their terms precisely so that they can be accurately understood” (para. 59)
• “The Family Court works on the civil standard of proof, namely the balance of probabilities. It would therefore help if opinions about testing results could be expressed in that way” (para. 59)
• “Where there is reason to believe that environmental contamination may be an issue, this should be fully described, together with an analysis of any factors that may help the reader to distinguish between the possibilities” (para. 59)