Using YouTube to study effects of hallucinogen salvia
I do prevention work with alcohol and drugs, so I am a consumer and a researcher in the field, and I need to keep up with the drug use fashion.
Drugs come into fashion and leave fashion, and sometimes there are new ones on the scene, and those new ones can really cause some of us in the prevention and research fields some trouble.
If users know more about the drugs than we do, it can make us look, well, like stiffs who really do not know what they are talking about.
A few years ago one of our surveys of students started to show more and more students listing a drug salvia divinorum as one of their "other drugs".
I needed to get up to speed very quickly on what this salvia divinorum really was and what it could do, so we could see whether or not we needed to develop prevention measures.
I found that salvia divinorum is part of a large family of plants, including garden varieties like mint and sage.
However, this specific species also contains a very potent hallucinogen. It was originally used for ritualistic purposes by the shamans in Oaxaca, Mexico, but over the past few years had started to seep into the recreational drug scene in the US and the UK.
This drug was not illegal, so while more and more people were trying it, people like me whose job is to understand drug use really did not know this was happening until much later.
Clearly I needed to learn more about it, and like most people my first instinct was to use Google to see what was out there.
Not too surprisingly there were sites describing it, but more surprising was that a lot of YouTube videos were present of people smoking Salvia and documenting the effects of it.
It did not take me too long to get a sense of what salvia divinorum could do, but the research literature was more frustrating.
There were no substantial studies documenting the effects of salvia, so it was difficult to know which aspects of the drug we needed to be worried about.
Normally what we like to do when we are interested in a drug is take it into the lab, have subjects use it and document the effects.
But this is time consuming and expensive, and for salvia divinorum I could not do even this. We did not know the effects, and we did not know how safe it was.
So I started thinking of what happened when I wanted to research alcohol. Then, my solution then was go out into the field. There are drunk people all over the place! It is certainly not hard to find a drunk person.
So I would get naturally drunk people and run them through the tests. Since I was not responsible for them being drunk, I was able to do a lot of research quickly.
Could I do the same sort of thing with salvia? The problem, from what I had seen in the videos, was that it is a very short-acting drug.
Although it is one of the most potent hallucinogens - people experience very profound impairments - it is gone very fast, in seven to ten minutes.
This meant both that people did not do it in public, and that even if I did find someone, the effects would likely be gone.
So I was in trouble - I could not get them in the lab and I could not go and find them. But then it occurred to me. How was it I knew so much about the effects of salvia?
I had seen it, I had observed it on YouTube - the website had already served as my database. Could I do that in a more systematic way?
And that is what we did. We randomly sampled YouTube videos and had researchers watch lots and lots of videos in 30 second segments.
From YouTube to Yale
They were looking for specific impairments like co-ordination loss, loss of sense of self, speech impairment and sweating. And we timed how long they kept the smoke in their lungs as a proxy for dose effects.
Our research was able to demonstrate a clear dose response within about 30 seconds, which tapered off at around five to seven minutes.
We published the study and very soon afterwards I got a call from a researcher at Yale University who was doing the lab study.
He was very excited because he was starting to find the same things that we had published. That lab study will help fill in the blanks, doing things I could not do like cognitive testing and longer-term impairments.
But using the YouTube videos allowed me to get very quickly into the scientific literature and the public domain a scientific view of what the observable effects of using salvia divinorum were.
And we could assess it in a natural setting, where people were actually using the drugs. And drugs do have setting effects: it may affect people differently to take a drug like salvia in a lab with someone in a lab coat standing over them, than when they take it with a bunch of friends who are also drinking and doing other things.
So I am hoping that this breakthrough can lead us towards more observational kinds of research.
YouTube is an incredible archive of social behaviour. It is a place where people feel very comfortable putting even what we would think of as personal activities, for everyone to see.
If we take those videos and start systematically to look at them - I think of it as systematic voyeurism - it can get us very quickly towards understanding some very complicated behaviours.
In my field, knowing the people around the user, understanding their behaviours and their impact is very hard to do in a lab. But if they have been videoed, we can code it, and we can code it on a level of thousands of people.
And in future, if we can use computers to look for physiological signals like heart or respiratory rate, we will be able to move through these videos even more quickly and even more objectively.
But in my field and beyond, these videos are more than just people sharing. They are extraordinary windows into social and biological behaviour that we really did not have just a few short years ago.
This is an edited version of Dr Jim Lange's Four Thought, which is on BBC Radio 4 at 2045 GMT on 21 December.Podcasts from the series are also available.