In December, New Brunswick-based producer Zenabis recalled a line of “CBD Light” gel caps, labelled as containing six milligrams of CBD per capsule, with no THC. It turns out, they were actually THC gel caps with no CBD whatsoever. 

Zenabis isn’t the only producer to run into labelling issues. Last September, Aurora Cannabis recalled it’s Blue Dream Sativa which had been sold through the Société québécoise du cannabis (SQDC). The label said the flower contained 19.1 per cent total THC, but the correct value was 15.7 per cent. More than 3,000 units were sold before the product was recalled. 

More recently, in April, consumers who picked up a package of MedReleaf Corp.’s San Rafael ’71 Island Sweet Skunk were in for a surprise. A batch of the strain that was labelled as having a total THC content of 17.99 mg/g actually clocked in at 179.88 mg/g.

These cases are not as uncommon as you might believe. Safety and quality assurance were supposed to be two of the hallmarks of legalization. But in the world of highly-regulated, legal cannabis, can consumers trust what the labels say? 

No universal approach

Like most things, the answer depends on who you ask. 

Cannabis testing is required by law — every batch needs to be third-party tested by one of the 122 labs across the country that have been certified to work with cannabis by the federal government — and the labels provide critical information for consumers.

However, testing cannabis is not as straightforward as other pharmaceuticals. The plant contains thousands of compounds, it can be consumed in a variety of different ways and there is an endless range of products, often containing other additives, that can be made from it. Further complicating matters is that testing methods vary from lab to lab, and there is no standardized, universal approach. 

Nowhere is this more apparent than in THC percentages, which is one of the biggest drivers in cannabis purchasing. Sources who spoke with The GrowthOp for this story said it’s not uncommon for producers to send samples to various labs and partner with whichever lab sends back the highest score. 

THC inflation impacts everyone from producers, particularly small scale growers who might not have the means to shop around for a high score, to consumers, who are left with an inaccurate understanding of what products work for them and which do not. 

The issue is further complicated by new product categories, like beverages and edibles. Health Canada allows for 10 percent variance on dried flower potency labels, but for edibles, depending on the cannabinoid content, as much as a 25 percent variance is permitted. This is because it’s harder to get accurate results in edibles, which contain a wide range of ingredients.

And while some variance is normal and expected, the mislabelling is not always accidental, says Nick Jikomes, Leafly’s principal research scientist. 

“What has been a persistent problem in the industry is labs producing results that are not fully accurate, and that often manifests itself as inflating the THC numbers in the plant,” he says. “The packaging and labelling are, unfortunately, not always going to be accurate signifiers of the cannabinoid content of the plant.”

Elevating standards

In 2018, Jikomes dug into the issue for a study that was published in Nature. Focusing on his home state of Washington, Jikomes found that there were systematic differences in the cannabinoid content reported by different laboratories. 

“Importantly, interlab differences in cannabinoid reporting persisted even after controlling for plausible confounds,” the paper reads, before concluding that there is a need for standardized laboratory methodology and a framework for assessing lab quality.

Jikomes says that that study influenced the way Leafly works with labs. To become “Leafly certified,” labs are required to share a large number of anonymized data samples, timestamps, anonymized producer ID’s and more. Leafly also hunts for systematic bias in the data before the partnerships are formed.

“We can see very easily when a lab is showing clear indications of producing biased results,” Jikomes says. “And they have to continually share data with us on an ongoing basis so that we can continue to monitor it.”

Jikomes says there are a number of labs that Leafly looked at partnering with before ultimately choosing to walk away for one reason or another. 

“Unfortunately, we chose not to work with a lot of labs out there. From a data science perspective, I want the data to be accurate so that I can trust what I’m looking at, but the entire spirit of Leafly is that we’re a trusted consumer resource. We very much care about making sure all of the data is accurate, and that it’s always coming from a licensed facility.”

What it all adds up to, Jikome says, is that simply being licensed to work with cannabis is not enough. Additional processes and measures need to be implemented to maintain higher standards and accuracy. After all, it’s not just THC and CBD that labs are testing, they are also being trusted to find pesticides, bacteria, mould, heavy metals, mycotoxins and more. 

Until the standards for validation and proficiency improve, the best test for knowing what cannabis products work for you might just come down to something more traditional: trial and error.