In the legal cannabis industry there is much ado about THC. Provincial distributors use THC potency to determine the value of products sold by licensed growers, retailers stock high THC products as a result of perceived consumer demand, and regulators propose using THC levels to determine the intoxication of cannabis consumers while driving. With so much focus and value put on this one molecule, accurately quantifying the THC values within legal cannabis products has never been more important.
So, how can you determine if the results you’re getting from your laboratory are accurate and represent the true value of THC in your products? First, you have to know a few things about variability.
Variability, in general, is the extent to which data points in a data set diverge—vary—from the average value, as well as the extent to which these data points differ from each other. In analytical chemistry, variability is defined as the extent to which the measurement will produce slightly different outcomes on the same test item, where the conditions of the measurement are either stable, or change over time, temperature, operator, etc. This definition has many layers to it, so let’s unpack it and explore how variability can impact the cannabinoid content measured in your sample.
How measurement conditions and processes impact variability
The first part of the definition of variability refers to the tendency of the measurement process to produce slightly different outcomes. On the surface this seems to be a straightforward statement, however, we first need to define what a measurement process is to truly understand what the definition is outlining.
A measurement process is the entire collection of procedural steps taken to execute a measurement. In some cases, the process is simple; like when we measure the mass of a sample. This is considered a direct measurement where the entire measurement process consists of one step – placing the sample on the balance.
In contrast, some measurement processes are very complex; as is the case when we measure the cannabinoid content in a sample of dried cannabis flower. This measurement process consists of five steps; selecting the sample or sub-sample, homogenizing the sample, extracting the cannabinoids from the plant material, purification of the extract, and finally analysis by HPLC. Given the complexity of this process, there is a higher tendency for it to give slightly different measurement outcomes when the analysis is repeated over time.
It’s akin to baking a loaf of bread exactly the same day after day; even experienced bakers will tell you that achieving consistent results is one of the most difficult aspects of baking. The freshness and temperature of the ingredients, the humidity in the room, the age of your oven, all of these factors influence the rate of variability of your bread making processes. Similarly, achieving precise and accurate results from complex measurement processes is one of the greatest challenges of analytical chemistry.
The second part of the variability definition describes the conditions that can affect the results obtained, such as; time, temperature, and operator. When we examine variability from an analytical chemistry perspective, the spread of outcomes can be impacted by each of these conditions and processes in an interdependent manner.
To establish how much variability a particular analytical measurement process has, requires a series of deliberate and controlled experiments. These experiments are designed to identify how each factor contributes to the overall spread in the measurement outcomes. Analyzing the results of these experiments will identify a range of measurements which include the true value.
Example: HPLC variability and THC potency results
Suppose an HPLC method is used to measure the cannabinoid content in a sample of dried cannabis flower. The method has been determined to have a 2% variability; meaning the sample’s true cannabinoid content is somewhere between +/- 2% of the measured number.
The sample submitted shows a final measured result of 20% THC content by weight. Given the HPLC method has a 2% variability as mentioned above, there is a high degree of certainty that the true THC content in the sample is a value between 19.6% and 20.4%.
From this example, you can see how the variability of a measurement process produces a result that occurs not in a single definitive point, but in a range; with the true value falling somewhere inside that range.
This poses an important question; why don’t analytical laboratories report this range rather than the exact number measured? Generally, laboratories don’t report the full range as variability is somewhat presumed. Determining whether or not a product meets quality standards will be based on a set of specifications predetermined by the manufacturer. For instance, if you expect your sample to test at 20% THC and the lab reports the measured value as 19.8%, then the sample is considered to be within the range of acceptance.
That’s why knowing the variability of the methods used by your laboratory is key to understanding whether a product meets your quality expectations. Choose a lab that has verified and validated their test methods and can provide details on the variability for each process they use. They should be more than willing to share this information with you and explain how they conducted their experiments and interpreted the results.
When you understand the measurement process used to test your products you’ll have more confidence in the accuracy and specificity of the results reported on your Certificate of Analysis and thus the information you’re providing to your partners and end consumers.