In a comprehensive study entitled: ‘Implications of the use of cannabis in the South African workplace’, I recommended South Africa’s corporate environment to move from urine to saliva drug testing, as a rational and scientifically validated testing for cannabis (T Lentsoe Implications of the use of cannabis in the South African workplace (LLM dissertation, University of Johannesburg, 2019)). In my previous article, I construed on the fundamental flaw of urine drug testing characterised by the prolonged detection period. In other words, ‘urinalysis’ shows cannabis used in the past, as opposed to accurate impairment. Therefore, employees who use cannabis at home run a risk of potential disciplinary actions and dismissals at work (T Lentsoe ‘Is workplace drug testing subject to a blanket ban? Unpacking Enever v Barloworld Equipment’ 2024 (Aug) DR 56). In response to this ramification, the study I conducted recommended South African employers to shift from urine to a saliva testing system, as it better indicates impairment by showing recent cannabis use (Lentsoe (op cit) at 48). In addition, a positive cannabis test from saliva samples must be interpreted according to the cut-off limit to establish accurate impairment (Endeavour Energy v Communications, Electrical, Electronic, Energy, Information, Postal, Plumbing and Allied Services Union of Australia [2012] FWAFB 4998 (14 August 2012) at para 40).
The long battle for cannabis community was no doubt a blasting fire following twist and turns in various courts. When one recalls the events of 2018, former Chief Justice of the Constitutional Court Raymond Zondo, was televised reading the judgment of Minister of Justice and Constitutional Development and Others v Prince (Clarke and Others as Intervening Parties, Doctors for Life International Inc as Amicus Curiae) and related matters 2018 (10) BCLR 1220 (CC), where the Constitutional Cout legalised the use, possession, and cultivation of cannabis for ‘private’ use (para 63). Put another way, this means that cannabis can only be used in private since the 2018 judgment. Although, this seemingly was a victory for the cannabis community, the aftermath of the Prince judgment left many questions unanswered for companies that have put in place drug testing policies. Among other challenges, workplace drug testing took the centre stage, becoming a leading topic in employment law. While employees maintain the right to use cannabis in private, it was not clear whether employers may subject employees to disciplinary actions or dismissals for testing positive for cannabis. As stated previously, drug testing methods used by corporate environments are flawed in that they are characterised by prolonged window detections of cannabis. If, to give a practical example, an employee consumes cannabis within a private realm, and a few days later reports for duty, such employee will likely test positive for cannabis when using urine testing (Lentsoe (op cit) at 3). Scientific studies, when investigated with extra caution, reveal that urine testing methods, often used by the South African employers, show the presence of drug metabolites in the specimen, and not necessarily the likelihood of drug impairment at the time of the screening. In essence, urinalysis does not detect whether the employee’s ability to perform tasks is inhibited (L Miike and M Hewitt ‘Accuracy and reliability of urine drug tests’ (1988) 36 Kansas Law Review 641 at 642).
Despite the Prince judgment, misfortunes unfolded as employees continued to be victims of dismissals merely for smoking at home. Tracing back to the events of 2020, Ms Bernadette Enever, employed as a category analyst (para 4), used cannabis before bedtime for health and religious purposes. She later presented to work (para 9-10), and was dismissed for testing positive for cannabis. The matter was referred to the Labour Court in Enever v Barloworld Equipment, a division of Barloworld SA (Pty) Ltd (2022) 43 ILJ 2025 (LC) (Enever I) for review. Enever I, as the court of first instance, ruled that the dismissal was substantively fair (Enever I at para 48). Unsatisfied with the decision of the Labour Court, Ms Enever appealed the ruling in Enever v Barloworld Equipment South Africa, a division of Barloworld South Africa (Pty) Ltd [2024] 6 BLLR 562 (LAC) (Enever II). The subsequent judgment of Enever II in the Labour Appeal Court (LAC) came with good news for Ms Enever and cannabis users. In essence, the LAC stated that the drug testing policy of Barloworld is discriminatory on arbitrary grounds. As a result, the LAC concluded that the dismissal is automatically unfair in terms of s 187(1)(f) of the Labour Relations Act 66 of 1995. The LAC ruled that the drug testing policy of the employer, Barloworld was ‘irrational’ for its prolonged window detection (Enever II para 63).
Celebrations were cut short when Barloworld further appealed to the Constitutional Court, in a bid to enforce its workplace drug testing policies. In turn, the apex court dismissed the employer’s leave to appeal, thus ending a long battle for cannabis users. This was no doubt a victory for cannabis users within drug testing jurisprudence. The apex court in Barloworld Equipment (Pty) Limited v Enever (CC) (unreported case no CCT139/24, 22-7-2024) (Maya DCJ, Gamble AJ, Kollapen J, Madlanga J, Majiedt J, Mathopo J, Mhlantla J, Rogers J, Theron J and Tshiqi J) (Enever III) rejected the appeal because there were no reasonable prospects of success.
Item 7(b)(i) of the Code of Good Practice: Dismissal of the Labour Relations Act stipulates that a workplace policy must be reasonable before a dismissal is rendered to be substantively fair (A Van Niekerk and N Smit Law@work (Durban: LexisNexis 2016) at 281-282). Importantly, ‘reasonableness’ is a broad standard which also comprises of ‘rationality’. In the context drug testing, the methods used by the employer must be rationally connected to the legitimate purpose of maintaining workplace safety. The following section provides a discussion on how South African courts failed to construe on the reasonableness and rationality of urine testing system according to the Code.
Since Prince endorsed constitutional privacy to use cannabis in private, employees were nevertheless subjected to dismissals merely for testing positive. Importantly, this was seen in a series of events since 2018, where arbitration rulings and labour court judgments approved dismissals for urine samples reflecting tetrahydrocannabinol (THC) (S Sibiya ‘The legal dilemma of managing and regulating private consumption of cannabis in relation to the workplace’ (2024) 45.1 Obiter 107 at 111-115). In Mthembu v Commission for Conciliation, Mediation and Arbitration and Others (2020) 41 ILJ 1168 (LC), for example, the employee was dismissed for testing positive for cannabis. Surprisingly, the Commission for Conciliation, Mediation and Arbitration (CCMA) failed to scrutinise on the reasonableness of the urine policy according to item 7(b)(i) of the Code. Mthembu failed to construe on the prolonged detection window associated with urinalysis. Similarly, in Moodley and Clover SA (Pty) Ltd (2019) 40 ILJ 2857 (CCMA), the employee was dismissed for testing positive for cannabis following the collection of urine samples (paras 22 and 42). Moodley took a different approach in that the laboratory report relied on a cut-off concentration in urine samples. In other words, a positive drug test was interpreted as the amount of THC above the cut-off limits of 15 per ng/mL. The initial outcome from the clinic was inconclusive because the nurse merely concluded that 200 per ng/mL indicates a high presence of THC without analysing the cut-off limit (para 44 and 53). Although Moodley scrutinised the cut-off concentration in urine specimens – precisely the relevant factor for interpreting a positive drug test – urine testing is unreliable, as it can detect substances consumed weeks prior. The CCMA failed to address the irrationality of the urine testing.
In Rankeng / Signature Cosmetics and Fragrance (Pty) Ltd [2020] 10 BALR 1128 (CCMA) (Rankeng) the employee was dismissed for reporting for duty under the influence of cannabis. He admitted having used cannabis earlier that morning before reporting to work (para 4-6). Rangkeng took a different approach from the previous arbitration rulings because the employee was reinstated on the basis that drug testing methods do not indicate whether the employee is impaired to the extent that his capacity to perform duties is inhibited. Rankeng, compared to the dicta of Mthembu and Moodley, applied a somewhat less stringent approach when issuing a sanction, ultimately lashing dismissal as a harsh sanction. Eventually, the CCMA replaced the dismissal with a final warning as a proportional and reasonable sanction (para 21). In SGB Cape Octorex (Pty) Ltd v Metal & Engineering Industries Bargaining Council & Others (2023) 44 ILJ 179 (LAC) the employee was dismissed for testing positive for THC. However, since Prince’s narrative, the employer’s substance policy in SGB required the use of a ‘saliva test’ for confirmation, as a legitimate mechanism through which accurate impairment may be scrutinised. After the examination of urine samples, and the subsequent saliva screening, both testing results reflected positive outcomes (para 3-4). Although the employer utilised saliva testing, which has the advantage of detecting recent use, the amount of THC in the saliva specimen were not construed according to the appropriate definition of a positive drug test correlating to the cut-off limits as addressed in Moodley. Even when using a reliable drug test of saliva, regard must be had to the extent of such impairment. In Marasi v Petroleum Oil and Gas Corporation of South Africa (LC) (unreported case no C219/2020, 27-6-2023) (Rabkin-Naicker J), the court analysed the ‘cut-off limits’, with extra circumspection, as the relevant factor for determining a positive test for drugs. The employee was dismissed after his urine samples exceeded the cut-limit of 50ng/mL, set by the employer’s policy (para 2). This approach is seen in Moodley but similarly, the employer used a urine testing method which is unreliable.
The common feature of the arbitration rulings and Labour Court judgments relates to their failure in addressing the unreliability of urinalysis and its impetuous violation of item 7(b)(i) of the Code.
Section 39(1)(c) of the Constitution paves a gateway for our courts to rely on foreign law where necessary (I Currie and J De Waal The Bill of Rights Handbook (Cape Town: Juta 2017) at 146). While urine testing has the fundamental flaw of showing inaccurate impairment, South Africa’s corporate environment may draw solutions from foreign jurisdiction such as Australia, where saliva testing is widely used. In the case of Endeavour Energy, the Australian court endorsed saliva testing which shows recent impairment of cannabis, consumed from 24 hours prior.
Endeavour Energy established a rational link between the ‘goal’ of preserving workplace safety and the ‘purpose’ of saliva drug testing policy (para 40). Put differently, saliva system is rationally connected to its aim of maintaining workplace safety, because it shows recent cannabis use. The court chose to scrap the urine testing system, and advocated saliva testing as a rational testing. The reasoning of the judgment derived from the prolonged detection window of urinalysis. Although Endeavour Energy correctly recommended saliva testing, the full bench however acknowledged this methodology, similarly to urine, as susceptible to false results if not approached with caution. Australian judicial precedents suggested swab testing, and further called for the cut-off concentration threshold test. In other words, although saliva testing detects recent cannabis use, it is important to consider the cut-off limits when interpreting a ‘positive’ drug test.
In Australian jurisdiction, what constitutes a positive drug test from saliva, is construed as the amount of THC above the cut-off limits of 15 ng/mL from the laboratory, set by the Joint Australian/New Zealand Standard (AS/NZS 4760:2019) procedure for specimen collection and the detection and quantification of drugs in oral fluid. If the urine samples contain THC below 15 ng/mL, the employee is declared to have tested negative. Such levels of THC do not affect the employee’s capacity to perform a task. In comparison, saliva samples reflecting THC above the 15 ng/mL are deemed positive.
Conclusion
While South African drug testing jurisprudence is still undeveloped, the discourse of this study recommends South African employers move from urine and adopt saliva testing system as a legitimate mechanism through which accurate impairment may be established. Although saliva testing has the advantage to detect recent impairment of cannabis, this study recommends that the levels of THC in saliva samples must be interpreted according to the cut-off limit, using saliva drug tests such as the CT Oral Scan product. In New Zealand, the CT Oral Scan was scientifically verified as a testing tool that conforms with the cut-off limits stipulated in the Joint Standard. South African drug testing jurisprudence must be developed in light of foreign positions to develop rationally grounded drug testing policies.
Thato Lentsoe LLB PG Dip LLM (UJ) is a law lecturer at the Independent Institute of Education’s Varsity College.
This article was first published in De Rebus in 2024 (December) DR 57.
