Safe to Eat, Not to Burn: A Fungicide, a Permitted List, and Who Was Doing the Testing

1. Introduction: a crop that is set on fire

Every residue question this journal has examined concerns something eaten, touched or breathed as it is. This one concerns something burned first.

1.0b And the regulator's own list makes it explicit

Products approved for food production are named in the recall notices as the source of the compounds found, which is not a coincidence but the whole mechanism.²

1.1 The same residue, two outcomes

Decided by whether the plant reaches a stomach or a flame.⁴

1.1a The plant is the variable, not the chemical

Nothing about the molecule changes between a food crop and this one. What changes is what the grower's customer does with the harvest.⁴

1.1b And the compound was legal on grapes

A mildew fungicide with ordinary food registrations, which is the fact that makes everything else in this article possible.⁵

1.2 What this article argues

That this is a hazard created by the method of consumption rather than present in the residue, that the regime relied on producers testing themselves, and that the contamination was found by somebody else's laboratory. Sections 10, 13 and 14 are the case.

2. The permitted list

The regulations are described as clear: licensed producers are permitted to use only the pesticides currently approved for use on cannabis under the federal pesticides act, and the use of any other pesticides, at any stage of cannabis production, is prohibited.¹

2.0b And the list is defined by registration rather than by hazard

A product is permitted because somebody sought and obtained a registration for this crop, which is the process our registration political economy article described.¹

2.1 At any stage is the strict part

Which rules out treating propagation material, treating a room before planting, or anything else a grower might consider separate from treating the crop.¹

3. A number that moved

Which we note in passing rather than as a finding.

Recall notices from early 2017 refer to thirteen authorised pesticides, and a clarification issued by the same department weeks later refers to fourteen.²³¹

3.0a The recall notices and the clarification disagree

Thirteen in the archived recall records, fourteen in the departmental statement, with no note in either acknowledging the difference.²¹

3.0b We report both because both are official

Two government documents from the same year giving two counts, which is the kind of small discrepancy our citation trail article was built around.¹²

3.1 Most likely an addition rather than an error

A list of that length changes when one product is registered, and the two documents are weeks apart.

That is our inference and neither document says.

4. Which is a very short list

Our comparison.

Thirteen or fourteen products for an entire crop is a smaller toolkit than any other production system this journal has examined, and smaller than the number of active ingredients our own structural work uses in a season.

4.0a The comparison with food crops is instructive

An ordinary field crop has hundreds of registered options across many modes of action. This one has a list a person can read in a minute.¹

4.0b And the pressure it creates is obvious

A grower facing mildew with few approved options is standing next to a shelf of products that work and are legal on other crops.

That is our inference and no source states it.

4.1 With the usual consequence

Our rotation articles found that a short list makes resistance management arithmetic difficult, since a rotation needs distinct modes of action to rotate between.

4.1b And the resistance consequence is not hypothetical

Mildew under glass is exactly the pressure our greenhouse articles describe, worked against a handful of approved options season after season.

4.2 We did not establish what is on it

None of our sources lists the approved products, only the compounds that are not on it.

5. What was found

Across several producers in late 2016 and 2017.

Myclobutanil, bifenazate and pyrethrins, described as prohibited in tobacco and marijuana.⁸ A later round also identified piperonyl butoxide, and one departmental statement referenced dodemorph.⁵⁴

5.0a The list grew as testing rounds continued

Pyrethrins in the first round at one producer, myclobutanil in a second, and further compounds at other producers later, which is what happens when the analytical panel widens.⁶²

5.0b And one is a botanical

Pyrethrins are the plant-derived compounds our botanical insecticides article examined, often assumed benign because they come from a flower, and prohibited here regardless.⁸

5.1 One of which this journal has covered

Our synergists article examined piperonyl butoxide as the compound added to insecticides to block metabolic resistance, which is a different role from the one it plays here.⁵

6. All approved for food

Stated in the recall notices themselves.

The two ingredients in one recall are not found in the thirteen pesticides authorized for use on cannabis plants, and are found in pest control products that are approved for use in food production.²

6.0b And that is stated by the regulator rather than the producers

Appearing in the official recall record, not in a defence offered by the companies involved.²³

6.1 Which is the sentence that reframes the whole episode

These were not exotic or illegal chemicals. They were ordinary agricultural products used on the wrong crop.²

7. And within food tolerances

In the first round.

All tested lots in one recall contained low levels of myclobutanil or pyrethrins that did not exceed any of the levels permitted in food production for these two pesticides.²

7.0b Which is a strange sentence to read in a recall notice

An official document recalling a product while stating that the amount found would have been lawful on a different crop.²

7.1 Which is why the first recall was the lowest class

A type in which use of or exposure to a product is not likely to cause any adverse health consequences.²

7.1b Which is why the recall class changed

The first classification rested on the food comparison, and once the levels exceeded it the producer escalated to a class implying possible adverse consequences.³

7.2 And a later round exceeded them

With subsequent testing at another producer identifying lots containing residual levels that exceeded the levels permitted in food production, prompting an escalation to a higher recall class.³

8. So why are they banned here

The question the whole article turns on.

Not because of the quantity, which in the first instance was within a standard set for food. The product is not food.

8.1 The prohibition is categorical rather than numerical

Not a lower tolerance for this crop but no permitted level at all, which is what a list of approved products means.¹

8.2 Which removes any question of how much

Detection is the violation, and that is why a level lawful on food still produced a recall.

9. The combustion argument

Myclobutanil is permitted on eaten crops because the compounds are metabolised by the digestive system and rendered non-toxic, and prohibited on combusted plants because it emits hydrogen cyanide when heated, a substance described as interfering with how oxygen is used in the body.⁴⁸

9.1 The effects described are dose-dependent

Headaches, dizziness, nausea and vomiting at lower concentrations, with gasping, irregular heartbeats, seizures, fainting and death given for larger ones.⁸

9.1b And the conversion is the only step that matters

Whether a fungicide is toxic to a person eating it is a question nobody needs to answer here, because the compound reaching the lungs is a different substance from the one applied to the plant.⁴

9.2 Which tells you nothing about a trace residue

The range describes the gas, not the quantity a burned residue would produce.

10. Which is a route-of-use hazard

Our framing.

The residue is the same molecule at the same concentration whichever plant it is on. What differs is that one route destroys it and the other converts it into something else.

10.0a The registration decision has to anticipate the flame

Which means a regulator approving a product for this crop is reasoning about pyrolysis rather than about metabolism, a different science from the one most residue limits rest on.

10.0b And the food tolerance is therefore the wrong standard

A limit derived from eating says nothing about a route that converts the compound rather than digesting it, which is why §7's reassurance about food levels does not settle anything.²

10.1 So the toxicology attaches to the consumer's behaviour

A registration decision here is a prediction about what somebody will do with the harvested material.

10.1b And it makes the label a statement about the end user

Most labels restrict where a product goes. This restriction is about what happens afterwards, to somebody the applicator will never meet.

10.2 And the prediction is unusually safe

Since the crop has essentially one method of consumption.

11. And it is a category we have not met

Across this journal's chemistry articles.

We have covered pro-insecticides activated inside the target, a photoproduct more potent than its parent, and a metabolite that inverts a selectivity margin. Each of those transformations happens in an organism or on a surface. This one happens in the flame.

11.0b And it is the only one the user performs

The other transformations happen whether anybody intends them or not. This one requires somebody to apply heat deliberately.

11.0a And the temperature is not incidental

A cigarette or a joint reaches temperatures no storage, shipping or cooking step would, which is the condition the prohibition assumes and the one no food tolerance contemplates.

11.1 Which makes the consumer the reaction vessel

Rather than an exposure route.

Sections 10 and 11 are ours.

12. Who was doing the testing

At the time, the regulator was not testing the cannabis produced by licensed companies, and the companies were trusted to monitor themselves on that front.⁶

12.0b And the answer for two years was nobody

Between the start of one recall period in 2014 and the first detection in late 2016, no analytical check stood between a producer's own laboratory and a patient.²

12.1 Which is the question this article would have asked first

Since a permitted list is only as good as the check that something outside it was not used.

13. Taking companies at their word

Acknowledged directly.

After the banned chemicals were discovered, officials acknowledged they hadn't been testing for such pesticides, and were therefore taking companies at their word.⁵

13.0b And it is a defensible design on its face

Licensing, good production practices and inspection are how most regulated production is overseen, and analysis of every lot is expensive.

13.1 Which was the design rather than a lapse

A licensing regime with good production practices and producer responsibility, verified by inspection rather than by analysis. More than 300 inspections of 30 licensed producers are reported for one fiscal year.¹

13.1a And the inspection volume was substantial

Three hundred inspections across thirty producers in a single fiscal year is not a token regime, which makes the detection gap a matter of method rather than of effort.¹

13.1b Inspection looks for the wrong thing

It can find a labelled container on a shelf or an undeclared ingredient in a declared product. It cannot find a residue in dried flower.

That distinction is ours.

13.2 And inspection did find one case

An inspection at one site observed that plants had been treated with a product containing an undeclared pesticide ingredient, which began the first recall.²

14. How it was actually found

In the case that expanded the problem.

One company's testing revealed that products it had bought from another licensed producer contained pesticides not federally approved for the crop. It sent the product back, and the producer contacted the regulator and issued a recall.¹⁰

14.0a The buyer was itself a licensed producer

Selling material grown by others alongside its own, which is why it had both the reason and the equipment to check.¹⁰

14.0b And the supplier then reported itself

Contacting the regulator and issuing the recall once its buyer returned the material, which is the system working from the point of discovery onward.¹⁰

14.1 A customer with a laboratory

Which is what a licensed producer buying from another licensed producer happens to be.¹⁰

15. Which is the verification problem again

In the form this journal keeps finding.

Our articles on third-party food safety audits and on credence goods both concerned situations where the party with the information is the party being assessed. Here the assessment was performed by a buyer who happened to have the capability, which is not a system.

15.0b The discovery route was commercial rather than regulatory

A company protecting its own product line, which is a reliable motive but not a system.¹⁰

15.1 And it only works where the buyer is sophisticated

A patient receiving the same product has no laboratory and no reason to suspect anything.

15.1b And the sophistication was accidental

Nothing required the buyer to test what it purchased. It did so for its own reasons, and the regulator learned of the problem as a consequence.¹⁰

15.2 And the patient is the party with the least information

Buying a regulated medical product from a licensed producer, with every assurance in the chain resting on testing that was not being done.⁵

16. The scale of the recalls

One producer recalled 115 lots of dried product and oil sold over two years for one compound, then expanded to a further 185 lots for another. Another recalled 69 lots and lost its organic certification.²⁷

16.0b Lots rather than units is the unit of recall

Which means a batch is withdrawn whether or not any particular package in it was ever tested.²

16.1 And the recall periods run back years

One covering product sold from September 2014, which means material had been in patients' hands for two years before anybody looked for it.²

17. What the regulator's own testing found

160 unannounced tests in one year found pesticides in nearly forty of them, and the fungicide at the centre of this article in 28 samples of dry flower.⁷

17.0b And the tests were unannounced

Which removes the preparation problem our announced-inspection articles identified.⁷

17.1 Which is a quarter of samples

Taken unannounced, after the recalls, after the regime had been publicly embarrassed, which is when compliance should be at its highest.⁷

17.1b And the other compounds were still present

Pesticides generally in nearly forty samples, with the fungicide accounting for most but not all of them.⁷

17.2 That framing is ours

The source reports the counts without comment.

18. The concealment allegation

Which we report as reported.

A national newspaper reported that one licensed producer had deliberately applied the toxic fungicide beginning as early as 2014, hiding containers of the banned chemical in ceiling tiles.⁷

18.0b And the company had been acquired by then

Operating under a larger producer's ownership and a new brand at the time of the reporting.⁷

18.1 The word deliberately is doing all the work

Everything else in this article describes residues of unknown origin, including one producer stating it did not know how the compound reached its crop.⁴

19. Which is a different problem entirely

Our distinction.

Contamination from an unknown source is a quality control failure. Applying a prohibited compound and hiding the containers is not a failure of anything; it is a decision.

19.0b And it is alleged to predate the recalls by two years

Beginning as early as 2014, against inspections running throughout that period.⁷¹

19.1 And it explains why the response was verification rather than guidance

Better production practices do not address somebody who knows the rule and breaks it.

19.1b And it changes how the low-risk finding reads

A trace residue from an unknown source and a deliberate application over two years are different exposure histories, even if the measured levels at recall were the same.¹

19.2 We have this from one report at one remove

Which §29 records.⁷

20. The harm question

Where the sources diverge.

One producer's recall is associated with ten adverse reaction reports received by the regulator.²

20.0b And an adverse reaction report is a weak instrument here

Requiring the patient to notice, to attribute, and to file, against symptoms that overlap with ordinary illness.

20.1 Which is a small number against the exposure

Ten reports against roughly 25,000 customers notified, though an adverse reaction report requires somebody to connect an illness to a product and then file it.²⁸

21. Two counts that do not match

A newspaper reports that hundreds of patients exposed through the December recalls have come forward complaining of serious health issues, ranging from breathing problems to persistent headaches and nausea, blistering rashes and muscle pain.⁵

21.0b And a recall itself generates reports

Once people are told their medicine was contaminated, symptoms already present acquire a candidate explanation, which is a known feature of post-recall reporting and not a reason to dismiss anybody.

That caution is ours.

21.1 Ten and hundreds are different numbers

And they are not necessarily inconsistent, since one counts formal adverse reaction reports to a regulator and the other counts people speaking to journalists or lawyers.²⁵

21.2 Which is a reporting-channel difference

Of exactly the kind our sighting metric and complaint data articles describe.

21.2b And the symptoms overlap with something else

One reported patient had already been diagnosed with a cannabis-related syndrome before the recall became public, and a laboratory director considered her symptoms consistent with possible hydrogen cyanide exposure.⁸

21.3 And a class action followed

Against one of the producers.⁷

22. The regulator's position

Stated plainly.

The department determined that the products affected represented a low health risk, and that the risk of serious adverse health consequences resulting from the inhalation of combusted myclobutanil in the recalled cannabis products was determined to be low.⁴¹

22.0a And it is specific about the route

Naming inhalation of the combusted compound rather than exposure to the residue, which is the correct question even where the answer is reassuring.¹

22.0b And it cites an occupational safety institute

Which implies a comparison against an exposure limit for hydrogen cyanide rather than against anything specific to this route.¹

22.1 Which is a statement about these quantities

Not about the compound, and not a retraction of the prohibition that made the recalls necessary in the first place.

22.2 And it sits alongside the enforcement decision

The response took account of the low risk posed by trace amounts and of the companies' full cooperation during the recall and subsequent investigations.¹

23. What changed afterwards

In sequence.

Random testing was announced in February 2017; screening by an independent laboratory before product reached market was announced in May; mandatory screening for all banned pesticides followed after spot tests found two further producers with residues; and fines of up to a million dollars were introduced by the end of the year.⁸⁶⁵

23.0b And the first step was announced within months

Recalls beginning November and December 2016, random testing announced February 2017.⁶⁸

23.1 Each step followed a further detection

Random testing after the recalls, mandatory testing after the random tests found two more, which is an escalation driven by results rather than by design.⁵

24. Verification replacing trust

Our summary of that sequence.

Each step moves the checking further from the producer and closer to a party with no interest in the result, ending with independent laboratories testing before sale and a penalty large enough to matter.

24.0b And the penalty arrived last

Testing first, then the fine, which suggests the problem was understood as detection before it was understood as deterrence.⁷

24.1 Which is the remedy our credence good article identified

Separating the party who benefits from a finding from the party who makes it.

24.1b And the cost falls on every producer

Including those who never applied anything outside the list, which is the ordinary consequence of a verification regime replacing a trust one.

24.2 And it was adopted only after the failure

Which is the ordinary sequence and worth noting rather than criticising.

25. The organic certification detail

Which is the strangest thing in this story.

One producer stated it did not know how the substance entered its crops, noting that it is a certified organic grower and does not use pesticides in its production processes.⁴ It subsequently lost that certification.⁷

25.0b And the certification was suspended rather than defended

Which suggests the certifying body treated the residue as decisive whatever the process records showed.⁶

25.1 Which our certification articles would predict

A certification is an audit of a process and not a measurement of a product, so a certified process can produce an uncertified result.

25.2 And the producer's denial may be true

Cross-contamination from purchased inputs, shared equipment or a supplied growing medium would produce a residue nobody at the company applied, and none of our sources establishes what happened.⁴

26. What we take from it

Three things.

A hazard can be created by the method of consumption. Section 9.⁴

Self-testing is not testing. Section 13.⁵

And a short permitted list creates pressure toward the products next to it. Section 4, which is our inference.

26.0b And a fourth, about where the standard came from

The comparison used in §7 was a food tolerance, because that is the number that existed, which is how measurement borrowed from a neighbouring domain becomes the default answer.

26.1 And one thing we cannot take from it

Whether the health outcome was real, which §§20 to 22 leave open and which no source we found resolves.

27. Our own position

The disclosure.

We hold a structural applicator licence and this is an agricultural matter we have no part in. Our interest is that §13 describes a regime in which the licensed party is trusted to comply and checked by inspection rather than by analysis, which is the regime we operate under.

27.0b And our incentive is the same shape

A product that works and is registered for a use next to ours is the same temptation described in §4, in a trade where nothing is sampled afterwards.

27.1 And nobody tests what we leave behind

No residue sampling follows a structural treatment in any jurisdiction our articles have examined, which our use reporting article approached from a different direction.

28. The Manitoba position

28.1 The regime is federal

The permitted list, the recalls and the enforcement all sit with the national regulator, so the provincial position is the same as everywhere else.¹

28.1b And the crop is grown here

Which makes the permitted list a live constraint for producers in this province rather than a distant regulatory curiosity.

28.2 What we could not find

The current permitted list, whether it has grown since 2017, and any Manitoba production facility appearing in the enforcement record.

28.2b And the product reaches consumers through provincial retail

Which places the sale under one authority and the residue rules under another.

28.3 And the growing conditions here are indoor

Which concentrates the pest problems our greenhouse and stored product articles describe, in a setting with the shortest legal product list of any crop.

29. Limitations and open questions

Most of this is news reporting. Six of our ten sources are newspapers or commercial cannabis publications, with only the recall notices and one departmental clarification being official documents.¹²³

That is the most important limitation for §§18 and 21, where the concealment allegation and the count of affected patients both come from reporting we could not corroborate.⁷⁵

The chemistry reaches us as a summary. That myclobutanil is metabolised by digestion and emits hydrogen cyanide when heated is stated by newspapers and by the regulator's communications, and we read no study of pyrolysis products at the temperatures involved.⁴

And that gap matters because the quantity of hydrogen cyanide produced from a trace residue is the number that would determine whether §22's low risk assessment is right.

The events are from 2016 and 2017. The legal framework has changed substantially since, and we did not establish the current list, the current testing regime or whether the enforcement approach has moved again.⁶

One source is a commercial cannabis site. Used for customer counts and a comparison with an American laboratory's findings, neither of which we could check elsewhere.⁹

Sections 3.1, 4, 8, 10, 11, 15, 19, 21.2, 24 and 25.1 are our reasoning. The account of a route-of-use hazard, the comparison with the transformations in our other chemistry articles, the reading of the verification failure and the point about certification are ours rather than sourced positions.

30. Conclusion

Licensed Canadian cannabis producers may use only the pesticides approved for that crop under the federal pesticides act, given as thirteen in the recall notices of early 2017 and fourteen in a departmental clarification weeks later, with everything else prohibited at any stage of production.¹² The compounds found in the recalls, myclobutanil, bifenazate and pyrethrins, are all found in products approved for food production, and the first round of tested lots contained levels that did not exceed food tolerances.² They are prohibited here for a reason that has nothing to do with quantity. Myclobutanil is permitted on some eaten crops because digestion breaks it down, and is banned on combusted plants because heating it emits hydrogen cyanide.⁴

That is a category this journal has not met. We have written about compounds activated inside a target insect, a photoproduct more potent than its parent, and a metabolite that reverses a selectivity margin, and every one of those transformations happens in an organism or on a treated surface. This one happens in the flame, which makes the consumer's method of use the reaction that creates the hazard, and makes a registration decision a prediction about what somebody will do with the harvest. The prediction is unusually safe, because this crop has essentially one route of consumption.

The rest is a verification story of a kind this journal has told before. The regulator was not testing licensed producers' cannabis and has acknowledged it was taking companies at their word, relying on inspections and good production practices rather than analysis.⁶⁵ One inspection did find an undeclared ingredient. The larger discovery came from a competitor: a licensed producer tested material it had bought from another and found compounds not approved for the crop.¹⁰ Recalls reached roughly 25,000 customers, unannounced testing the following year found pesticides in nearly forty of 160 samples, and a national newspaper reported that one producer had applied the banned fungicide deliberately from 2014 and hidden the containers in ceiling tiles.⁹⁷ What followed was random testing, then independent screening before market, then mandatory testing, then fines of up to a million dollars: each step moving the checking away from the party with an interest in the answer. The regulator assessed the health risk from the recalled products as low. Ten adverse reaction reports were received, and hundreds of patients were reported to have come forward. Both numbers can be right, and neither is the one that would settle it.

References

1. Clarification from the federal health department on myclobutanil and cannabis, published as a departmental statement in 2017. Official government communication. Source for the statement that licensed producers are permitted to use only the fourteen pesticides currently approved for use on cannabis under the federal pesticides act and that the use of any other pesticides, at any stage of cannabis production, is prohibited; for the statement that the regulations are clear and that licensed producers are responsible for ensuring their products comply; for the figure of more than 300 inspections of 30 licensed producers in one fiscal year, with results published online; for the list of enforcement options including education, recalls, adding terms and conditions to a licence, licence suspension or revocation; for the determination that the risk of serious adverse health consequences from inhalation of combusted myclobutanil in the recalled products was low, referencing an occupational safety institute; and for the addition of new terms and conditions requiring testing for unauthorized pesticides at the affected producers. https://www.canada.ca/en/health-canada/news/2017/03/clarification_fromhealthcanadaonmyclobutanilandcannabis.html
2. Recall notice for cannabis for medical purposes issued by one licensed producer, published in the federal recalls database and since archived. Official recall record. Source for the account that a departmental inspection in October 2016 observed plants treated with a product containing an undeclared pesticide ingredient not authorized for use on cannabis, initiating a voluntary recall; for the recall of 115 lots of dried product and oil sold between September 2014 and October 2016, expanded to a further 185 lots produced between January and November 2016 following subsequent testing; for the definition of the recall class used as one in which use of or exposure to a product is not likely to cause any adverse health consequences; for the statement that the two ingredients are not among the thirteen pesticides authorized for use on cannabis plants under the federal pesticides act and are found in pest control products approved for use in food production; for the statement that all tested lots contained low levels that did not exceed any of the levels permitted in food production for these two pesticides; and for the report of ten adverse reaction reports received in relation to products sold during the recall period. https://recalls-rappels.canada.ca/en/alert-recall/recall-cannabis-medical-purposes-mettrum-ltd
3. Recall notice for cannabis for medical purposes issued by a second licensed producer, published in the federal recalls database and since archived. Official recall record. Source for the account that a voluntary recall of the lowest class was initiated in December 2016 after five lots tested positive for low levels of two named compounds not among the thirteen authorized pesticides, both found in products approved for food production; and for the escalation in January 2017 to a higher recall class after subsequent testing identified additional lots containing residual levels that exceeded the levels permitted in food production. https://recalls-rappels.canada.ca/en/alert-recall/recall-cannabis-medical-purposes-organigram-inc
4. Newspaper report on the expansion of a medical cannabis recall after a banned pesticide was found, published by a national daily in January 2017. News material. Source for the statement that myclobutanil is permitted in small doses on certain crops that are eaten since the chemical compounds are metabolized by the digestive system and rendered non-toxic, and is also approved for crops that do not retain high levels of residue as they grow; for the statement that the pesticide is not approved for use on plants that are combusted, such as tobacco or cannabis, and is known to emit hydrogen cyanide when heated; for the note that lawmakers in three American states moved quickly to ban it, in some cases by emergency legislation; for the departmental reference to two banned pesticides including dodemorph and the range of possible steps including detention of product, recalls or revoking a licence; for one producer's statement that it did not know how the substance entered its crops, noting it is a certified organic grower that does not use pesticides in its production processes; and for the departmental determination that the recalled products represent a low health risk. https://www.theglobeandmail.com/news/british-columbia/medical-marijuana-recall-expanded-due-to-banned-pesticide-being-found/article33643174/
5. Newspaper report on mandatory pesticide testing for medical cannabis producers, published by the same national daily in May 2017. News material. Source for the statement that officials acknowledged they had not been testing for such pesticides and were therefore taking companies at their word; for the requirement that licensed producers screen products for all banned pesticides after random spot tests of seven growers found two with traces of myclobutanil or piperonyl butoxide; for the description of myclobutanil as used to kill mildew, a known carcinogen strictly prohibited for use on plants that are smoked because it produces hydrogen cyanide when heated; and for the report that hundreds of patients exposed through the December recalls came forward complaining of serious health issues ranging from breathing problems to persistent headaches and nausea, blistering rashes and muscle pain. https://www.theglobeandmail.com/news/national/health-canada-mandates-pesticide-testing-for-medical-cannabis-producers/article34913001/
6. Article on random testing and penalties in the Canadian cannabis pesticide enforcement sequence, published by a commercial cannabis news outlet. Trade news material, cited as attributed material. Source for the statement that the regulator was not at that time testing cannabis produced by licensed companies and that the companies were trusted to monitor themselves; for the timeline in which one producer announced recalls in November 2016 and two others followed weeks later; for the announcement in February 2017 of random testing for banned pesticides; for the announcement in May that producers' cannabis would be screened by an independent laboratory before being put on the market; and for the order requiring two named producers to submit products to regular testing with one having its organic certification suspended. https://www.leafly.ca/news/strains-products/random-testing-million-dollar-fines-canadas-cannabis-pesticide-crackdown
7. Article reviewing pesticide detections in Canadian cannabis testing, published by the same commercial cannabis news outlet. Trade news material, cited as attributed material and flagged because the concealment allegation in this article rests on its account of a newspaper report. Source for the figures of 160 unannounced tests in 2017 finding pesticides in nearly forty samples and myclobutanil in 28 samples of dry flower; for the statement that one producer lost its organic certification and recalled 69 lots after testing positive, and was subsequently the target of a class action; for its report that a national newspaper found another licensed producer had deliberately applied the fungicide beginning as early as 2014 and hidden containers of the banned chemical in ceiling tiles; for the naming of three further producers whose products tested positive that year; and for the introduction of fines of up to one million dollars by the end of 2017. https://www.leafly.ca/news/health/pesticides-health-canada-2019
8. Broadcaster news report on the announcement of spot checks following the recalls, published in February 2017. News material. Source for the announcement of random testing of medical cannabis products after recalls affecting nearly 25,000 customers led to reports of illness and the possibility of a class action; for the identification of myclobutanil, bifenazate and pyrethrins as prohibited in tobacco and marijuana; for the description of myclobutanil as a fungicide permitted on food crops which produces hydrogen cyanide when burned, with the department's account that hydrogen cyanide interferes with how oxygen is used in the body and may cause headaches, dizziness, nausea and vomiting, with larger concentrations causing gasping, irregular heartbeats, seizures, fainting and death; and for an individual patient account and a laboratory director's comment that the reported symptoms are consistent with possible hydrogen cyanide exposure. https://www.cbc.ca/news/canada/nova-scotia/medical-marijuana-cannabis-health-canada-random-testing-1.3971601
9. Article on the regulator's spot-checking announcement, published by a commercial cannabis website. Commercial content material of uncertain editorial standard, cited as attributed material. Source for the figures that recall notices were sent to 3,895 customers of one producer and 21,054 of another, against roughly 119,709 people then holding valid medical licences, which the article characterises as leaving over twenty per cent of patients affected; and for its report that an American testing laboratory found a high proportion of samples positive for pesticides in October 2016. https://merryjane.com/health/health-canada-cannabis-recall-causes-pesticide-testing
10. Newspaper article on the pesticide-driven recall, reproduced on a press archive platform. News material. Source for the account that one company's own testing revealed that products it had purchased from another licensed producer contained pesticides not federally approved for cannabis production, that it returned the products, and that the supplying producer then contacted the regulator and issued a recall; for the statement that the affected product contained bifenazate or myclobutanil or both, with myclobutanil widely reported to produce hydrogen cyanide when burned; and for the note that myclobutanil is banned from medical marijuana production in a named American state. https://www.pressreader.com/canada/calgary-herald/20170117/281698319446637