A drug-testing company says 3.4% of its workplace screens found drugs last quarter. What that can and cannot tell an employer

What the report says

The Drug Detection Agency (TDDA) sells workplace drug and alcohol testing, education and policy services in Australia and New Zealand. Its Australia Imperans Report for the third quarter of 2026 draws on tests taken at 19 clinic locations and more than 50 mobile clinics between 1 July and 30 September. It says 3.4% of all screens conducted by TDDA across Australia indicated the presence of drugs, against 3.0% in the second quarter and 3.1% in the same quarter of 2025, and describes the result as an overall increase in workplace drug detections in the third quarter of 2026.

Among the positive results, the report gives each drug’s share for this quarter and the same quarter a year earlier. Stimulants rose and cannabis and opioids fell:

The company’s chief executive, Glenn Dobson, said in the release: Our latest Imperans data shows that cocaine continues to move in the wrong direction in Australian workplaces. He added that the increase is not as dramatic as some of the movements we are seeing internationally. The report recommends employers review their drug and alcohol policies and consider the types and frequency of testing they undertake, including the use of random testing programmes where appropriate. TDDA sells that testing. That is worth knowing when reading the recommendation; it is not a reason to think the figures are wrong.

The company’s case for publishing the report is a fair one. It says regulators publish incident reports, but they do not quantify when substances are a factor. A regular count from a large tester is useful for that reason. The question is how much weight it can bear.

Five things the numbers cannot tell you

1. “Up 21.4%” is a change in a share. The release says ATS were in 64.8% of positive tests, up 21.4% YoY. The rise from 53.4% is 11.4 percentage points; 21.4% is that rise expressed as a proportion of the starting share. It is also a share of positive results, not of everyone tested. On the rounded figures, ATS positives work out at about 2.2% of all screens this quarter (64.8% of 3.4%) against about 1.7% a year earlier (53.4% of 3.1%). That is our arithmetic, and it is a smaller-sounding number than the headline. The bases also shift within the release: every state line is year on year except Victorian cocaine, given as up 8.0% vs previous Qtr. The company’s own graphs put Victorian cocaine at 17.0% of positive screens in the third quarter of 2025 and 10.3% this year, so on the basis used for the other lines it fell. The release does not say why that line uses a different base; the graphs it links to carry the year-on-year figures.

2. The shares add up to more than 100%. 64.8 + 40.2 + 12.7 + 9.2 is 126.9, because one positive screen can show more than one drug. In Western Australia the five drugs in the company’s graphs add to 160.0%, with ATS in 95.8% of positive screens, up from 73.6%. Shares like these move against each other: a drug’s share can rise without more of it being found, if positives for other drugs fall.

3. There is no denominator. The methodology note says Total figures on testing volumes or testing results by industry and region are commercially sensitive. The release says TDDA provides over 250,000 tests every year. The post launching the report in November 2024 put that at over 250,000 drug tests every year in New Zealand and Australia, an average of about 62,500 a quarter across both countries. The Australian count for this quarter is not given, and each state figure rests on fewer screens still. That matters most for the 0.4-point step from 3.0% to 3.4%. As a rough guide, and treating screens as if they were a random sample, which they are not, a 0.4-point gap between two quarters of 10,000 screens each would be within what chance alone could produce; with 50,000 screens a quarter, chance alone would be an unlikely explanation. We cannot say which applies. A quarter ago the company described 3.0% as continuing the broadly stable workplace drug detection rate observed over the past four quarters.

4. Four kinds of test are pooled. Data from pre-employment, post-incident, regular and random testing has been combined. Those are different populations: pre-employment screens test applicants, and post-incident screens are taken because something went wrong. If the mix of tests or of clients changes from one quarter to the next, the positive rate can move with no change in drug use. The release does not give the mix, or the industries.

5. Screens, with no word on confirmation. Testing methods included urine and oral fluid screening. The release does not say how many non-negative screens went on to a laboratory confirmation, or whether its figures count confirmed results at all. Nor does it claim a detection means a worker was impaired; it describes itself as a snapshot of drug trends across Australian workplaces and industries.

The independent check: what sewage says

The Australian Criminal Intelligence Commission’s National Wastewater Drug Monitoring Program measures drugs in sewage. Its Report 25, published on 30 April 2026, analysed samples collected from December 2024 to October 2025; in August 2025 the program covered 57% of the population, about 14.5 million people. Its findings point the same way as the company’s:

• Methylamphetamine consumption in August 2025 was the highest recorded over the life of the Program for both capital cities and regional areas.
• Cocaine consumption reached the highest levels recorded by the Program in August 2025 in regional areas and October 2025 in the capital cities.
• Regional areas of Western Australia currently have the highest overall methylamphetamine consumption, which sits alongside the company’s 95.8% ATS share for that state.

So the vendor’s numbers are not out on their own. Two limits keep the comparison to direction only. First, timing: the wastewater data ends in October 2025, which is the company’s comparison quarter, not this one. It says stimulant use was at record levels a year ago; it says nothing yet about July to September 2026. Second, population: wastewater measures everyone connected to a treatment plant, working or not, while the company measures the people its clients chose to test, for the four reasons above. The two numbers are not comparable and we have not compared them. When we checked the program’s report index on 16 October 2026, Report 25 was still the latest; Report 26 had not been published. We will check the direction again when it is.

What the law asks of an employer

Safe Work Australia’s page on drugs and alcohol opens: Alcohol and drugs can affect a person’s ability to work safely. This includes medicines that are prescribed or over-the-counter. For a business, it says As a PCBU, you must manage health and safety risks. This might include setting specific policies for the use of drugs and alcohol. It lists a duty to consult with workers who carry out work for the business or undertaking and who are (or are likely to be) directly affected by a health and safety matter. It says workers must be fit and well enough to do their job, not be under the influence of alcohol or drugs and not use alcohol or illegal drugs while at work.

On testing itself the page is short: Some companies have policies to test their workers for alcohol and illicit substances, particularly if a worker could kill or seriously injure themselves or someone else. It adds that There’s a legal blood alcohol level in some jobs, including road and rail transport, maritime and mining occupations. Neither the page nor the model Act says that a business must drug test.

The duties sit in the model Work Health and Safety Act. Section 19 says a person conducting a business or undertaking must ensure, so far as is reasonably practicable, the health and safety of its workers while at work (what “reasonably practicable” means). Section 28 says a worker must take reasonable care for his or her own health and safety and co-operate with any reasonable policy or procedure of the person conducting the business or undertaking relating to health or safety at the workplace that has been notified to workers. The words “drug” and “alcohol” do not appear in the model Act; the duties are general, and Safe Work Australia’s page applies them to drug and alcohol risk. The model law is a template. The law that binds a business is its own jurisdiction’s, and Safe Work Australia notes that SWA is not a regulator and cannot advise you about WHS issues in the workplace; it sends readers to their regulator. We are not giving legal advice either.

Our read, labelled as opinion

Take the direction from a report like this, not the decimals. The company’s stimulant and cocaine figures point the same way as the national wastewater program, an independent measure, and that is a reasonable prompt for an employer in a safety-sensitive industry to look again at its own approach. But a vendor that withholds its volumes, pools four kinds of test and reports screens without saying whether any were confirmed has published a signal, not a measurement, and a headline percentage built on a share of an unknown number deserves to be discounted accordingly. The company is entitled to keep its volumes private; a reader is entitled to give the precision less weight as a result.

Before relying on any vendor’s trend figures, an employer can fairly ask four questions: how many tests, for what reasons, screened or confirmed, and in which industries. And the more useful response to a report like this is usually not more tests but a check of the basics the law and the regulator’s guidance point to: whether the business has a drug and alcohol policy at all, whether workers were consulted on it, and whether it has been notified to them, since section 28’s duty to co-operate attaches to a reasonable policy that has been notified to workers.