Could an employee lose their job by consuming dagga outside working hours, especially in safety-sensitive industries where workers are prohibited from being under the influence of intoxicating liquor or drugs?

These are some of the legal and ethical dilemmas raised by medico-legal experts from the University of Pretoria (UP) in a recent research article in the South African Medical Journal (SAMJ).

Dr JB (Tim) Laurens, until recently a senior forensic toxicology researcher and lecturer at UP, and Professor Pieter Carstens, a professor of criminal and medical law, note that the 2018 Constitutional Court decision to legalise the consumption of cannabis “in private” has created a number of challenges for employers and employees.

This was because labour laws such as the Occupational Health and Safety Act and the Machinery and Occupational Safety Act make it an offence for staff to be under the influence of intoxicating liquor or drugs at work.

To ensure that constitutional rights to privacy, freedom of religion, autonomy and other rights were properly balanced against the right of other members of the public to safety and health, Dr Laurens and Prof Carstens call for a “legally defensible” approach to the regulation of dagga – by adopting a similar approach to regulation of legal alcohol use.

They note, for example, that the psychoactive component of dagga – Delta (9)-tetrahydrocannabinol (THC) – is excreted slowly from the human body and can still be detected several days after having been consumed.

THC is absorbed quickly when smoked and produces a rapid euphoric effect on the brain, with peak concentrations within nine minutes of the first puff. Effects can last for two to four hours, potentially impairing psychometric tasks, memory, sense of time, motor coordination and reaction speed.

However, these peak concentrations (as high as 160ng/ml in blood) then decline rapidly to about 10% within one to two hours. This is followed by a second phase of slow redistribution of THC into the bloodstream and hepatic system.

They note that the elimination half-life of THC is approximately one day in casual smokers and three to five days in chronic smokers, but it is also possible for measurable levels to be detected in passive smokers.

Research in other parts of the world suggested that people with THC/blood concentrations above 5ng/ml were up to six times more likely to be responsible for driving accidents, but those with levels below 2ng/ml showed no signs of impairment.

The methods used to test employees for dagga usage could also create difficulty. While blood testing was seen as the best way to estimate impairment, collecting specimens of blood was very invasive and also required specialist staff.

While saliva and urine testing were seen as scientifically accurate and less invasive, the concentration levels could not necessarily be used to reach accurate conclusions about the level of impairment or intoxication.

As a result, it was essential that company policies on prohibited substance testing respect human rights to freedom and autonomy, and that employees provide consent for such testing.

They suggest that the use of cannabis on work premises should be prohibited in total and that testing policy should indicate a threshold concentration for THC specific to the type of tests used (urine or saliva).

If a person was found to be above the urine threshold, it did not necessarily mean that they were "under the influence". “It merely means that he or she did not comply with the threshold agreed upon in the policy and [this] should then be regarded as a breach of contract,” Prof Carstens and Dr Laurens note.

It was therefore essential to set a threshold concentration that was low enough for an individual to operate in a risk-sensitive environment, while also respecting their autonomy to use cannabis in private.

They also make the point that using a "zero" concentration as a threshold is not scientifically correct and could be problematic to defend in a court or labour hearing, since the "zero" depends on the analytical detection method. If the organisation insisted on a “zero” concentration threshold, then the detection method should be specified, and the limit of detection be provided in the policy.