In a recent narrative review, Scott & Scott concluded that on-site pill testing lacks evidence of efficacy and proposed a cut on public funding until efficacy is properly demonstrated. Here's why I dislike the paper. #harmreduction #pilltesting 1/ https://onlinelibrary.wiley.com/doi/abs/10.1111/imj.14742

I'll present rebuttals for each of their arguments, some of which have already been presented by others (see below). 2/

"On-site pill testing cannot account for the context and environment of the music festival."
"Pill testing cannot predict inter-individual variability in drug metabolism and toxicity or the effects of concomitant use of alcohol or other substances." 3/

The two previous arguments are irrelevant and even dishonest: A service cannot be evaluated based on goals and objectives that were never assigned to it. They both show a total lack of understanding of what pill testing is. 4/

Pill testing cannot be equated with quality control. It is (or should) always (be) provided in conjunction with counselling, based on the drug, set and setting model. Drug use is neither promoted nor discouraged but risks are thouroughly discussed with users. 5/

Pill testing is unreliable and "is a qualitative analysis and cannot establish drug concentrations or ratios of various isomers of racemic compounds." 6/

Barratt et al. challenged this argument. Not only can pill testing provide quantitative analysis but it quite often offers a wide range of tests to increase reliability. 7/ https://onlinelibrary.wiley.com/doi/10.1111/imj.14954

"Among the 155 samples that did not match what the user was 'sold/acquired', only 8 pills (5%) were discarded".

I see two problems with the previous argument. First, small effects can still have large and profound impacts. 8/

Second, once again, the goal of pill testing is not abstinence but harm reduction. A drug user will likely get rid of a pill only when actual content doesn't match expectations, something that usually doesn't happen very often. 9/

"Only a small portion of patrons use pill testing anyway".

This argument speaks in favor of pill testing. Indeed, one can claim that the service would be more often used and be more efficacious if it was more widespread and if drug use was legalized or decriminalized. 10/

Moreover, it underestimates the reach of pill testing. When a particularly dangerous sample is identified, an alert is issued on-site (in the festival) and sometimes nationally that reaches patrons who didn't use the service. 11/

"Although pill testing can arguably provide information about substance use trends and emergence of new dangerous agents, more comprehensive information is obtainable from full analysis of confiscated whole pills and those seized during drug raids or by undercover police". 12/

Here again Barratt et al. offer a rebuttal. Drug checking has the potential to monitor the drug market more rapidly and it "provides key information not available through police seizures, including the nature of discrepancy between expected and actual content of drugs". 13/

To this, I'll add that, especially in the case of Belgium, a major drugs market hub, drug checking provides information about what people use, whereas police seizures tell us more about which drugs are manufactured or shipped through Belgium. This is not the same. 14/

"No studies have definitely tested, in a controlled way, whether pill testing and/or associated brief interventions decrease verifiable harms associated with taking pills".

This is imo the only indisputable argument but... 15/

First, although desirable such studies are almost impossible to conduct given the usually repressive national policies on drugs and the lack of public funding. 16/

Second, until such studies are conducted, available observational studies suggest that pill testing is efficacious (e.g., Measham, 2018 ; Valente et al., 2019). 17/

Third, when tentative evidence suggests the efficacy of a service, shouldn't opponents first demonstrate the lack of efficacy or worse the harmfulness before advocating for a cut in public funding?! 18/

Scott & Scott end their paper with two studies in support of the argument that pill testing promotes drugs use.

First, I was unable to find the reported findings in one of the two cited papers (i.e., State Coroner's Court of New South Wales, 2019). 19/

Second, I find it almost amusing but dishonest to discard any evidence of the efficacy of pill testing because it's based on intentions and not behaviors but to use such evidence when it suggests the harmfulness of pill testing. 20/

Third, and further substantiating the accusation of selective use of references made by Barratt et al., I find it dishonest not to mention the studies (e.g., Benshop et al., 2002) showing that pill testing doesn't promote drug use. 21/

I strongly recommend reading the other responses to Scott & Scott written by Smit-Ritger & Van der Gouwe, on the one hand, and Measham et al., on the other hand. 22/

@FMeasham @WeAreTheLoopUK @DIMSinfo

In short, Scott & Scott's paper is incomplete, biased and of little use in the debate on the efficacy of drug checking. /23