The article is based on a purely descriptive and rather anecdotal study, more sociological than biomedical in nature, which, based on a series of surveys conducted in Denmark (a choice made by the authors given the lack of regulation of veterinary cannabis products in this country). It shows that there is a high percentage (38%) of dog owners who use non-standardised cannabis preparations to treat their pets for different types of ailments. Although the idea is interesting and some conclusions can be drawn, the reality is that the study has limitations that, undoubtedly, may have generated biases that complicate the validity of the conclusions. Among these limitations, I will highlight the three most important ones. 

The first limitation has to do with the procedure for obtaining the surveys. It is true that the number of surveys cannot be considered small, but the way they are obtained (online with advertising through social networks) may have influenced the decision of pet owners to answer the surveys or not, favouring more answers from owners who use cannabis, which makes it difficult to consider the 38% figure as representative of the situation in Denmark. 

The second problem has to do with heterogeneity, which is typical of this type of study, and which undoubtedly must have significantly affected the results, complicating the validity of the study's possible conclusions. This heterogeneity is evident in relation to: 

• The types of pet ailments for which cannabis preparations are used: pain is the main reason, but also allergies, seizures, dermatological problems, behavioural alterations, etc. 
• The variability of the preparations used: cannabidiol oil (CBD) is the most commonly used, but cannabidiol is also used in other formulations, most of them uncontrolled (there is no indication of the presence of other cannabinoids and no indication of possible contaminants), and also including THC in some cases. Also to be considered within this heterogeneity is the way in which the preparation was acquired and the origin of each one, which must be assumed to have been highly variable in this study. 
• The administration patterns that were followed, i.e. what mode of administration was used, what dose and what duration of treatment, and how was the treatment distributed over time? This information is not included in the article, so it is understood that the variability must have been very large. 

The last issue has to do with the determination of the efficacy of the treatment, which in this study is based exclusively on the impression of the owners, not on any kind of analysis or examination carried out by a veterinary service, and which could certainly be assimilated to the so-called 'placebo effect' which, in this case, would be related to the owners' desire for their pets to benefit from the treatment. 

In any case, the study is of interest insofar as it reproduces results and identifies problems that are common in humans, with patients self-medicating using uncontrolled cannabis preparations. The solution to these problems will have to come from regulating of the use of medical cannabis, which is still pending in our country, and from the assumption that the development of cannabis/cannabinoid-based medicines is not something that only has to do with human pathologies, but that the potential benefits should also be applied in the veterinary field.  

It is important to note that there are canine pathologies that are relatively similar to human pathologies, for example, the case of degenerative myelopathy which is relatively similar to amyotrophic lateral sclerosis. This is an important fact, as the study of these canine pathologies can serve as a fundamental tool in the research of their equivalent pathologies in humans and vice versa, as veterinary medicine should benefit from treatments that are developed in the human field. This study could contribute to strengthening this relationship.

The study is an Internet survey—you take the responses and report them. It’s not hard to do and is typically a bit biased, [depending on the] methods of seeking respondents, but solid enough to provide the information [the authors] report. 

I expect [the result] is the same or higher here [in the United States]. There are surveys suggesting similar use in the US, where there is no need for a script to give low-THC varietal hemp extracts to pets.  It is not surprising, and this idea is opening up in many countries where low-THC hemp is becoming legal to distribute to the general population. 

McGrath, Kogan and other researchers have done US survey research which appears to be similar [to this study], as a lot of people have used [cannabinoids] for pain, anxiety, dermatological issues and seizures. There is now ample evidence in dogs that it can help with osteoarthritis pain, seizures, and atopic dermatitis; the data is less convincing for anxiety. I think the novelty is that people are doing it in both Europe and the US now; it appears to be safe but people have no idea what the dose should be, and many veterinarians are unaware as well. There really needs to be education for vets on how to best advise clients who want to use this nutraceutical treatment for dogs with specific issues.   

The [study’s] biggest limitation is that the dose was not disclosed for any survey questions. The dose that is commonly recommended is around 2 mg/kg body weight for many issues I mentioned above, and maybe higher for anxiety and behaviour[al issues].  We have no idea whether the perceived improvements by owners are due to the placebo effect or true treatment effect. Dosing information would have been useful, to see if clients are dosing anywhere near recommendations by veterinary professionals from the peer-reviewed literature where it has been shown to be useful. 

I think [this study] might open the conversation for clients to veterinarians, to better understand how to use cannabinoids in the daily life of dogs where clients want to use it.