Risk Assessment Methods

PSA #8: Patients can now almost fill a stadium
August 12, 2017
The Hempfest Bump – Retail Sales Increase
August 22, 2017
Show all

Risk Assessment Methods

Over the past 10 days, I have posted on HI-Blog relating to the recent suspension of the certification of PEAK Analytics to test State-legal Cannabis in Washington. Those posts and subsequent communications have alerted the WSLCB, the public, and industry participants to the exposure risk stemming from the continued availability of PEAK-tested product in Retail Access Points. They also allude to the non-zero risk of morbidity and mortality associated with such exposure.

Some of the comments I’ve seen relating to my work regarding PEAK’s inadequacy in testing for “Total Coliforms” and “Enterobacteriacea” suggest to me that a brief explanation of the method and logic I used to derive my estimates is in order. I will do so below in as concise a manner as I can. Hopefully, this will let people make better judgements regarding their willingness to expose themselves to PEAK-tested product.

The risk associated with exposure to things consists of two general parts: estimating the risk of being exposed and then estimating the danger (morbidity and/or mortality) associated with such exposures.

The bulk of my work focused on estimating the number of consumers that have been exposed to “tainted” product. Below, I summarize the methods I used to estimate that, on average, 540 people per day are being exposed to PEAK-tested product likely to contain unacceptable levels of one or more of the microorganisms in question.

But first, the risk associated with such exposures is also important. Not everyone being exposed will develop illness and/or be otherwise harmed by such exposures. It is also almost certain that the majority of those so exposed will not die from such exposure. Media reports likely would have picked up on such deaths if they were, indeed, an inevitable (or highly likely) result of such exposure.

The specifics of what the LCB does not allow in Cannabis and cannabinoid-containing products for sale in the marketplace they regulate are contained in WAC 314-55-102. The portion below shows some of the verbiage they use to describe unacceptable levels of microbiological organisms.

Note their use of the word “pathogenic”. That adjective means something which relates to a bacterium, virus, or other microorganism capable of causing disease

As far as I am aware, something that can cause disease can cause harm. How likely such exposures are to cause disease and how likely that disease is to progress to the point that harm is done to any person so diseased is not known in this case. It is, however, NOT ZERO.

I base that last statement on the fact that the LCB has decided to put into place rules that are supposed to limit exposure to pathogenic organisms in consumers utilizing the State-legal Cannabis system they are responsible for regulating.

I’ll get back to what type of harm (morbidity) and whether death (mortality) is likely to result from exposure to tainted PEAK-tested product at the end of this post.

Estimate of Exposure:

In summary, here is how I estimated that about 540 people per day are being exposed to “tainted” product tested by PEAK.

RJ-Lee, in auditing PEAK’s operation, found that their performance was unacceptable when it came to detecting the two classes of microorganisms mentioned above.

I looked at the rate of detection by all other certified labs testing State-legal Cannabis in Washington and compared it to the rate of detection by PEAK (which, it should be noted, was not zero). For this analysis, I used data on all microbiological QA testing that was done in the first 5 months of 2017.

The difference in the rate of failing product during QA tests associated with these two classes of microorganisms ended up being about 1.1% — with PEAK failing 0.43% of samples for one or both of these classes of microorganisms, while the rest of the lab industry failed samples at a rate of 1.53% for one or both of these classes of microorganisms

This difference in failure rates served as one foundation of my 540-person-per-day estimate. The other foundation was knowing how many people, per day, were being exposed to PEAK-tested product.

To estimate the number of people being exposed to tainted PEAK-tested product, I took my estimate of the number of people being exposed to PEAK-tested product and multiplied it by 1% (instead of 1.1%, as I wanted to err a little on the side of being conservative).

To do this, I looked at the number of transactions that were occurring at retail access points over the last couple of weeks in May. It turns out that there are about 1.2 million transactions occurring per week. Given the percent of the testing market that PEAK represents (over 30%), about 380,000 people PER WEEK (or about 54,000 people PER DAY) are likely being exposed to PEAK-tested product. My primary assumption here is that each purchase is consumed by only one person (no sharing). If there is sharing, the overall risk of exposure would increase from what I report here.

A very important point is that NOT ALL PEAK-TESTED PRODUCT will be tainted (by one or both of the classes of microorganisms in question). We do not know if any given package contains tainted product. We do, however, know that PEAK’s failure rates were 1.1% (absolute) below those of the rest of their peers. Hence, somewhere around 1.1% of PEAK-tested product can be assumed to contain levels of one or more of these microorganisms that would normally stop a product from being sold, as-is, in Washington’s State-legal Cannabis market.

1.1 % is not a large number, but it is both NOT ZERO, and it represents about 71% of the failure rate reported by the other active labs for these two classes of microorganisms. Put another way, PEAK reported less than 30% of the failure rate expected, given the normative failure rate for the rest of the industry.

Data from the other labs testing during this time suggest that PEAK was FAILING to FAIL product tainted by unacceptable levels of one or both of these two classes of microorganisms in approximately 1.1% of the relevant microbial screens that they conducted.

Multiplying the number of people that are buying PEAK-tested product daily (about 54,000 people) by this differential risk (which I decremented from 1.1% to 1.0% to err on the lower side), results in my estimate that about 540 people are being exposed to PEAK-tested product every day that SHOULD have failed one or both of the screens for these two classes of microorganisms, but were passed and allowed to continue on to market.

That is, put simply, how I estimated the number of exposures to PEAK-tested product likely tainted by at least one of the microbiological entities for which the auditor found PEAK’s processes unacceptable.

Hence, my statement that 540 consumers of State-legal Cannabis products are likely being exposed per day.

I am quite confident in the accuracy of that number. It is an estimate (which, by definition, is not precise), but it is a good estimate. It is also a conservative estimate.

About 540 people per day (or about 3,780 people per week) are being exposed .

Estimate of Harm (morbidity and/or mortality):

How many people, then, are likely to become sick (be harmed) or die (be very much harmed) from exposure to product passed by PEAK, yet likely containing one or both of these classes of microorganisms? I do not know.

I do know, however, that it will likely be greater than zero.

That comes from the basic assumption that 540 people times a non-zero risk of morbidity and/or mortality will result in a non-zero number of people likely getting sick or getting dead as a result of such exposure.

The WSLCB considers these microorganisms to be pathogenic or to be otherwise indicative of pathogenic potential.

They, therefore, have the potential to cause disease (or to be indicative of something that does).

It follows that approximately 540 people per day are being exposed to pathogens that have the potential to cause disease. That, again, is a non-zero risk.

What are the diseases these microorganisms cause? …. I do not know (other than they likely fall into the class of “coliform- and/or enterobacteriacea-related diseases).

How many people will experience disease from such exposures? … I do not know, other than it is likely to be more than zero (otherwise, why even do microbiological screening of products?).

How many people will experience death from such exposures? …. I do not know, other than it is possible that it will be more than zero.

The MORE people that get exposed, the more that are likely to become diseased (to be harmed) and, by extension, the more that are at risk of dying from those microorganism-related diseases. That number may be small, but IT IS NOT ZERO. At least, it is highly unlikely to be ZERO.

It is even more unlikely to be zero now that the only legal way to purchase cannabis in Washington is at a licensed Retail Access Point. Now that Patients depending on cannabinoids as part of their medication / nutrition / supplement regime are more likely to be buying cannabis at Retail Access Points, we must remember that some of the 540 daily exposures likely involve exposures of Patients to microorganisms in tainted PEAK-tested product.

Per my exposure estimate work, PEAK’s inability to accurately detect these two classes of microorganisms exposes almost 3,800 people per week to one or both of these classes of microorganisms (on average).

The LCB would not have set QA thresholds for these classes of microorganisms if harm from exposure was not assumed to exist. I do not know the specific harm that can be expected to happen to these 3,800 people, but I assume that the LCB’s requirements indicate that harm CAN result from such exposures.

Please remember that some Patients are likely to be differentially susceptible to pathogenic organisms (immuno-compromised Patients, for example). With such increased susceptibility, there is an increased potential for harm in such Patients when exposed to pathogenic microorganisms.

In such people, the risk of morbidity from such exposures is increased. Similarly, the risk of resultant mortality is also increased.

Once the risk has been quantified:

On this basis, I raised an alert with the authorities, with many of the Retailers selling Cannabis products to consumers, and with interested members of the industry and public that have access to either my Facebook posts, Twitter announcements, or the 502Cannabis Google Groups forum.

The WSLCB responded to my e-mails last Friday, and they currently do not view the situation as requiring a recall of PEAK-tested product currently on store shelves.

This post, hopefully, gives you more insight into why I believe that the risk of disease and death amongst Cannabis consumers in Washington today is increased by the presence of PEAK-tested product on the shelves of Washington’s State-legal Cannabis Retail Access Points.

I still believe that to be true. The WSLCB seems to disagree.

Net result – use your best judgement and, please, do not “freak out” over PEAK-tested product. Any given exposure is unlikely to cause harm in and of itself.

As time goes by, more and more people are being exposed (or so my models suggest). The risk of harm may be small, but a small risk multiplied by lots of exposures has a tendency to manifest itself in harm.

Be safe, be informed, and live as healthy and civil a life as is possible for you.

In closing, as of the end of business yesterday (midnight for Retail Access Points), my models suggest that approximately 11,373 consumers have likely been exposed to tainted PEAK-tested product since PEAK was informed of RJ-Lee’s audit results on July 24.

If one in 100 such exposures led to harm, then just shy of 114 people would be expected to have been harmed since July 24.

If one in 1,000 such exposures led to harm, then around 11 people would be expected to have been harmed during that time.

If one in 10,000 such exposures led to harm, then it would be about 1 person that I’d expect to have been harmed during that time.

If only one in every 100,000 such exposures led to harm, then we would be at a fraction of a person being harmed thusfar (perhaps a child?).

I do not know which, if any, of these 4 scenarios best describes the degree of harm that has been experienced by those ingesting product tested and passed by PEAK Analytics that likely should have failed microbiological screens. I am, however, fairly certain that many thousands of consumers have been put at increased risk by the availability of PEAK-tested product on the shelves of Washington’s Retail Access Points
What I do know is that I am unlikely to be one of them.

That at-risk number increases every day.

I would like to think that none of those likely exposed are harmed as a result.

I also would like to think that none so exposed die as a result.

PLEASE ask to see the Certificate of Analysis (CoA) when you shop at any of the State’s active licensed Retail Access Points. What you do is always your choice, but if the shop is unable to supply the test sheet corresponding to the product in question, please know that the shop is not living up to the standards and expectations of the LCB (and rule) regarding how they deal with customer enquiries relating to product testing.

Please choose wisely.

Life has taught me that denying the presence of a problem rarely fixes that problem.

Even if the risk of harm is low from exposure to product tested by PEAK prior to the suspension of their laboratory certification, it is not zero (given the auditor’s report and the rules governing QA testing).

My work and efforts over the past 10 days have, I hope, shed further light on the extent of that risk.

Leave a Reply

Your email address will not be published. Required fields are marked *