File talk:Rational scale to assess the harm of drugs (mean physical harm and mean dependence).svg

 

 

critique

I know more of drugs in a practical, real-world sense and in a pharmacological scientific sense than most people and I would like to add that this graph is WAY off.

I would like to see a complete report on all of the studies used to collect this information.—the preceding unsigned comment was added by MagicJigPipe (talk • contribs) 03:57, 14 January 2008 (UTC)

I'd be fascinated to see caffeine added to this.—Preceding unsigned comment added by 216.27.182.83 (talk • contribs) 19:51, 15 December 2008 (UTC)

This graph doesn't make any sense, opiates do very little harm physically.—Preceding unsigned comment added by 91.156.108.123 (talk • contribs) 12:59, 28 December 2008 (UTC)

Agreed, the graph is extreme bullshit all the way through, anyone who knows a thing or two about drugs can see that. Highly suggesting removal. 14:46, 17 February 2014 (UTC)

"This is a tertiary source [..] as it summarizes experts' opinions on the matter (which are secondary sources) without any direct references to primary sources. " - Or in other words: the 'data' is just wild guesswork, the opinion of a few people without any real-life experiences. Which explains why it makes no sense at all. 14:54, 17 February 2014 (UTC)37.201.226.8

Fixed all the English wikipedia captions in order to clarify that it is just a poll result and not based on facts. C0NPAQ (talk) 16:02, 17 February 2014 (UTC)

Suggesting to put this tag to the file (whatever wiki politics you have, people have been misrepresenting stuff 'by accident' and that doesn't need to be so):

Please be cautious with captions and consider before usage: The graph data is not based on facts or true circumstances but on questionnaire results (of experts, like psychologists and policemen). It doesn't show the actual harm or dependency potential of the drugs, as it would be evident by e.g. studies about lifespan, relapse rates, crime rates, financial and social consequences, etc. That it was published in a medical journal further may mislead people to implicitly assume such an empiric origin of factual, medical and otherwise solid nature (rather than just the sheer opinions of a group of people).

C0NPAQ (talk) 16:28, 17 February 2014 (UTC)

make bold

make "mean" columns bold as i shown below.

Mean Physical harm ↓ Acute harm ↓ Chronic harm ↓ Intravenous harm ↓ Mean Dependence ↓ Pleasure ↓ Psychological ↓ Physical ↓ Mean Social harm ↓ Intoxication ↓ Social harm ↓ Health-care costs ↓

this is a minor but necessary edit. Is it really necessary to semi-protect this page????? Dancedto made (talk) 14:42, 4 January 2009 (UTC)

The druggies regularly play with the numbers in the harm table, usually making heroin/cocaine/cannabis more pleasurable and less harmful. Gandulf (talk) 15:09, 22 April 2009 (UTC)

False?

This article is extremely false and should be removed as soon as possible. Spreading wrong information can be very detrimental to people's education and thus affecting future decisions they might have to make. This graph should only be allowed to remain on all these different chemical Wikipedia pages if there are citations backing it up. People need to stop just reading wikipedia and listening to the news, and get information from their own research to from opinions. Ignorance is bliss.—Preceding unsigned comment added by 198.236.40.220 (talk • contribs) 18:35, 19 February 2009 (UTC)

The main thing I personally know about drugs is to avoid them completely (or at least all 20 listed, I use caffeine a lot - I wonder how addictive it is?). I certainly would not want anyone to try any drug just to do their "own research". Hopefully this chart will help kids learn that some drugs are far more harmful and far more addictive than others. Just by looking at the chart you can see why so many people find it so hard to quit smoking, for example, and why it is said that there are no former heroin addicts - once an addict always an addict. Plus it finally answers the question, is cocaine addictive - why is it that society considered coke a social drug and crack highly addictive? Wasn't it Richard Pryor who said he wasn't addicted to cocaine - he quit using it every day? Or just liked the way it smelled? Of course he also joked that "Marijuana's not addictive, I've been smoking it every day for years and I'm still not addicted." If you look at the data, the main thing that makes cannabis addictive at all is that it is pleasurable, so I wonder how addictive that makes chocolate? The two drugs I would love to see included the next go around are chocolate and caffeine. Gandulf (talk) 15:09, 22 April 2009 (UTC)

Inaccuracies

This chart really is ridiculously inaccurate, it needs to be deleted. Andrew Nutter (talk) 23:39, 20 January 2009 (UTC)

I think if you read the article you would see that they went through a lengthy procedure to decide which factors to assess and how to assess them. Each expert first made their own independent assessment, which was then discussed as a group, a few drugs at a time, and then they could change their own assessment based on the group discussion. Finally a correlation was made to see how individual groups varied in their assessment. Most of the drugs received about the same assessment from each group, with only a few exceptions. All of this is in the article. Gandulf (talk) 15:09, 22 April 2009 (UTC)

Delete

As per the comments here, this image needs to be deleted. It is, first of all, not scientific in the least, and secondly, the actual qualifications of the "specialists" in question were weak. On top of that, "only 29 of these psychiatrists, out of 77, decided that the methodology presented in the paper was worthwhile" - that's terrible.
If this image isn't deleted, at least remove it from all Wikipedia articles. The image is a cruel joke to anyone with a clue about drugs (on a scientific level, not a armchair drug-user level). 98.150.161.251 15:37, 3 November 2009 (UTC)

Such an argument is ridiculous, how is it "not scientific"? It was published in a distinguished medical journal and written by some of most experienced drug experts in the UK in this paper. "only 29 of these psychiatrists, out of 77, decided that the methodology presented in the paper was worthwhile" is incorrect, quoting the paper "Replies were received and analysed from 29 of the 77 registered doctors who were asked to assess 14 compounds" that doesn't mean that the other 48 didn't think that it was worthwhile, just that they didn't respond. I'm not saying that the image is perfect, there are obviously huge difficulties trying to produce a quantifiable way to rate the harm caused by drugs. As far as I am aware however this is the only attempt to do so and therefore I think that it should definitely not be deleted. Smartse (talk) 18:30, 3 November 2009 (UTC)

Concur with Smartse. If we don't trust referreed journal articles as en:WP:RS on medical issues, we might as well just post "here's what some guy told me" information. Whether the data is misinterpretted in the article and whether others have published alternative analysis or done whole alternative surveys, those are reasonable points for editors to investigate. But the debate over this image has run for months or years, but nobody who thinks this diagram is crap has seen fit to find a better one from any source that is at least as reliable as the one cited for this one's data. DMacks (talk) 19:09, 3 November 2009 (UTC)

I also agree with Smartse and DMacks, the source appears reliable. If there are other methodoligies that produce a different chart, those should be included as well. Shanata (talk) 10:10, 11 November 2009 (UTC)

While I want to make it clear that I disagree with the data presented in the image there are a few things that need to be said about it.

One, it should be represented on Wikipedia, on a very limited basis, only because it is a novel attempt at a very difficult process: that is quantifying the danger of many different drugs on the same scale. In my opinion, under any appearance of the image should be a mention of its relative partiality - that fact that it is the result of a collection of opinions and controversial (as is somewhat the case on David Nutt's page here on the site). I personally believe David Nutt's page is about the only reasonable place for this image to reside unless another page is created that specifically deals with the quantization of harm done by controlled substances.

Two, it is not the original image from the paper. Perhaps there is some copyright reason for that, I do not know. However, it is very important to realize that the image does not even really resemble the original image from the journal article. Psycherhexic (talk) 07:46, 30 December 2009 (UTC)

Black border

This graph would probably be better without the black border. Jason Quinn (talk) 10:54, 19 September 2010 (UTC)

Vertical axis text-orientation

Some of the languages' versions of this file have the label for the vertical axis rotated. I think that looks nicer and is easier to read. This is an issue for the .en, .fr, .it, .pt ones; the .ko, .nl, .ru, and .tr are all rotated already. DMacks (talk) 12:17, 19 September 2010 (UTC)

Please Replace

The Lancet did an updated study with a new graph's such as (http://i.imgur.com/16Syj.png) and here is the Full Source: http://download.thelancet.com/pdfs/journals/lancet/PIIS0140673610614626.pdf?id=3d35b1b5aa0ec416:6e23b5e7:12f43d60c44:-52ad1302514347957 — Preceding unsigned comment added by 24.21.229.185 (talk • contribs) 10:34, 11 April 2011 (UTC) (UTC)

It doesn't picture dependence though and it is less flashy and eye-catchy. Quite frankly I don't believe that people on wikipedia will accept any change of that graph, because there is no replacement for it that would 'validly' picture the same content (because that would be too comprehensive and expensive for a study). Thus wikipedians will rather stick to the nonsensical opinion-based graph and revert all your edits. Additionally, from the first impression I don't believe that it does represent that accurately (although I can't tell if it is bullshit, which is a good start). Didn't look into it, no time. But crack cocaine is listed considerably higher than cocaine, although it is the same drug. If you are an addict you don't care if you smoke or inject it at all. A considerably indicator for bias, because the 'experts' probably perceive crack as intensified cocaine or something. Then obviously on cocaine you will be more agressive and prone to violence then on heroin. So I wonder how both can be equal at all. It looks all "meh". 37.201.226.140 00:40, 12 March 2014 (UTC)

Mean Physical harm is misleading, Median Physical harm as alternative

Hello,

I was asking myself how it could be, that in the diagram methylphenidate is rated physically more harmfull as tobacco and only slightly less than slcohol. I didn't find any other source that shows evidence for this rating. But the answer is pretty simple:

The physical harm axis is the mean score of acute, chronic and intravenous physicall harm. But for some substances, the paper hasn't any data for harm caused by intravenous use, and the score is just set to zero (Tobacco, Cannabis, 4-MTA, Ectasy, GHB, ...) or N/A for alcohol.

But the mean score sums also over NA=0 values, which greatly benefits those substances where there is no data for intravenous use. So alcohol gets a score of (1.9+2.4+0)/3 = 1,4333..., Tobacco gets (0,9+2,9+0)/3=1,2666... and Methylphenidate gets (1.2+1.3+1.6) = 1,36.

It is false to say there is "no data for intravenous use". It is known that those drugs are never used intravenously, and it is better to think of NA as "Not Applicable", than "No Data". Since they are not used intravenously, there is no harm from intravenous use, and in computing the mean, a zero was used. Gandulf4 (talk) 05:25, 5 July 2014 (UTC)

There are several ways to fix this problem:

• Using only the Chronic score. Straightforward, easiest to find in the table, and a quite meaningfull measure. But e.g. scores for 'Acute' are the ignored. E.g. methadone is much more harmfull in acute use (lethal dose increases with chronic use due to addiction).

  That does not "fix" any problem, but you are welcome to plot chronic harm, or any other column against any other column. What you can not do though, is make up other calculations, or leave out data in the calculations. That would be original research. Gandulf4 (talk) 05:25, 5 July 2014 (UTC)

• Using the maximum of (Acute, Chronic, Intravenous). Good approach, as it's easy to check, and misses no category. Drawback: Harm of some substances is overestimated. For intravenous use substances get strongly judged by their additives. That's realistic in the case of street-heroine (often intravenously used, bad additives), but highly unrealistic for those drugs which are seldomly used intravenous, as Methylphenidate tablets (but extender in tablets are also bad intravenous).
• median of (classes where data is given) i.e. median (Acute, Chronic, NA) = (Acute + Chronic)/2. Drawback: Needs an additional column in the table, that shows the median score ((Acute + Chronic)/2 not already given).

I think that are the most meaningfull ways to solve the problem. I would suggest to use the last option, because I think it has the fewest drawbacks. I added therefore another column Median Physical harm. Using only the Chronic score would be an alternative, it's the easiest of all versions, and still much better as the mean values (Acute + Chronic + NA)/3 = (Acute + Chronic + 0)/3.

Best, Richard (talk) 22:01, 1 April 2014 (UTC)

There is a similar talk in en.wikipedia. It chooses this option:

• including 'Intravenous' in the mean only if there is given any data (i.e. not N/A or 0). Drawback, two different types of mean values (Acute + Chronic + Intravenous)/3 and (Acute + Chronic)/2. Mean is not as stable as median, and more vulnerable for bias. Nontheless much better as the mean values (Acute + Chronic + NA)/3 = (Acute + Chronic + 0)/3.

Best, Richard (talk) 22:59, 1 April 2014 (UTC)

I created a new svg image, using the mean (Acute + Chronic)/2 for (Acute, Chronic, NA) cases:

It is intended as an replacement for Rational scale to assess the harm of drugs (mean physical harm and mean dependence).svg. I will also create an image using the median method, but the file is named '...harm of drugs (mean physical harm and...', so the median version should be a seperate file, that might be used as an alternative.

Best, Richard (talk) 15:04, 2 April 2014 (UTC)

It is absurd to do new calculations on the data other than what appear in the paper[1]. Of course the data can be manipulated by leaving out NA, or assigning a value of 1.5 to NA, but both severely distort the data. There will always be criticisms of why wasn't this one higher or this one lower, but it is original research to manipulate the data to make drugs be located differently on the chart. The only thing that can be done is publish a new paper so that we can create a chart that is based on that new data. This data is what it is. The new chart File:Development of a rational scale to assess the harm of drugs of potential misuse (physical harm and dependence, NA free means).svg needs to be deleted as original research, and the additional columns in the data table need to be removed. It is though, useful to add a note that "NA=0". The reason for NA being zero is that for example, with alcohol and cannabis there is no harm from the intravenous use of alcohol or cannabis, because neither are ever used intravenously. To leave out those zeroes gives undue weight to the two items that are still included, and erroneously moves the position of the drugs on the chart. Nominated the new chart for deletion at Commons:Deletion requests/File:Development of a rational scale to assess the harm of drugs of potential misuse (physical harm and dependence, NA free means).svg as original research. Gandulf4 (talk) 05:02, 5 July 2014 (UTC)

Commons does not have a "no original research" policy... AnonMoos (talk) 14:39, 8 July 2014 (UTC)

We just have to be careful to note what content is vs isn't directly attributable to sources so that the various *.wp with NOR policies don't accidentally use a graphically high-quality "original" item in places where a higher standard for the content is required. DMacks (talk) 14:47, 8 July 2014 (UTC)

More than that, the chart simply contains erroneous calculations. It is based on the statement that the original chart "greatly benefits those substances where there is no data for intravenous use". Well duh. It also benefits those substances where there is no addiction. And no harm. That is what it is supposed to do. And by leaving out intravenous harm, they are actually assigning a value of the mean of the other two values for intravenous harm. The premise is also false, that there is "no data for intravenous [harm]". There is data for intravenous harm for alcohol - there is no intravenous harm for alcohol, because no one uses it intravenously. NA simply means "Not Applicable", not - use the mean of the other two values. The statement "The calculation (acute harm + chronic harm + 0)/3 is really a joke, and his suggested version is much more realistic" is absurd. Zero is perfectly valid because there is no intravenous harm. What is a joke, is the "suggested version", which assigns absurd values to the intravenous harm for the drugs that are not used intravenously, and therefore have no intravenous harm. Gandulf4 (talk) 21:05, 8 July 2014 (UTC)

The paper gives alcohol a combined physical harm of 1.40. To leave out the NA for intravenous and give it a physical harm of 2.15 by averaging the chronic and acute harm treats alcohol as if it effectively had an intravenous harm of 2.15 which is ludicrous. Gandulf4 (talk) 06:22, 5 July 2014 (UTC)

To be honest, what seems to be a joke is the original authors' assumption that averaging those three measures together actually results in a meaningful quantity. It would be much more reasonable and provide a much more informative measure of actual risk of physical harm to be conservative and consider the maximum of those three values (Acute harm, Chronic harm and Intravenous harm) as a measure of Risk of Physical Harm. Here is an obvious example: if a drug has the highest Acute Harm of all drugs, it must have a high Physical Harm, even if it has zero risk of chronic and intravenous and, likewise, if a drug has the highest Chronic Harm of all drugs, it must also have a high Physical Harm, even if it has no acute or intravenous risks. Another obvious point is that it makes no sense that, if a drug has both high Acute Harm and Chronic Harm, the value of Physical Harm should depend on the Intravenous Harm. As far as I can tell, the only think that makes sense (if one wants to aggregate those 3 values into a meaningful and conservative estimate of risk for physical harm) is to either provide the "maximum" or the average value between the two highest values (i.e. the average between the maximum and the median). Regardless of whether this warrants an extra research paper or not, it seems very clear that simply averaging those three things together results in a meaningless measure. In a nutshell... the problem with your argument is that you are assuming that the whole "mean measure" actually has a meaning, and then back-calculating what the "Intravenous harm" would have to be so that the mean and median measures would be equivalent... and then using this to state that using the median implies somehow that we are overestimating "Intravenous Harm". Notice that people do not care about what the specific "intravenous harm" of a drug is... they only care what the *overall* harm of the drug is. And, as many people already mentioned here, taking the mean value of those three estimates seems like a bad choice. --188.37.176.82 08:17, 14 December 2016 (UTC)

Other than the individual columns of "chronic harm, acute harm, intravenous harm", etc., these are the only data values that can be plotted - the ones calculated in the paper. You are free to choose any of the columns you wish to plot against any of the others, but what you can not do, without original research, is come up with a way of calculating an overall value that is different than was used in the paper. You can also calculate an overall mean score, but must also do that in the same manner as the paper, anything else is also original research. Gandulf4 (talk) 06:39, 5 July 2014 (UTC)

Solvents has a slightly higher overall score than 4-MTA, 1.270 vs. 1.267, and in the paper they are not plotted the same either. Gandulf4 (talk) 22:49, 5 July 2014 (UTC)

Drug	Physical harm	Dependence	Social harms	Overall harm
Heroin	2.78	3.00	2.54	2.77
Cocaine	2.33	2.39	2.17	2.30
Barbiturates	2.23	2.01	2.00	2.08
Street Methadone	1.86	2.08	1.87	1.94
Alcohol	1.40	1.93	2.21	1.85
Ketamine	2.00	1.54	1.69	1.74
Benzodiazepines	1.63	1.83	1.65	1.70
Amphetamine	1.81	1.67	1.50	1.66
Tobacco	1.24	2.21	1.42	1.62
Buprenorphine	1.60	1.64	1.49	1.58
Cannabis	0.99	1.51	1.50	1.33
Solvents	1.28	1.01	1.52	1.27
4-MTA	1.44	1.30	1.06	1.27
LSD	1.13	1.23	1.32	1.23
Methylphenidate	1.32	1.25	0.97	1.18
Anabolic Steroids	1.45	0.88	1.13	1.15
GHB	0.86	1.19	1.30	1.12
Ecstasy	1.05	1.13	1.09	1.09
Alkyl Nitrites	0.93	0.87	0.97	0.92
Khat	0.50	1.04	0.85	0.80

What is rational about this scale?

Cannabis is listed as having a higher "Chronic harm" than both amphetamine and cocaine and more "Social harm" than tobacco. Really? I've seen plenty of ridiculous things on the web, but I don't expect blatant lies on an encyclopedic website. — Preceding unsigned comment added by Chrisknepper (talk • contribs) 00:11, 23 June 2014‎ (UTC)

The graph shows neither "Chronic harm" nor "Social harm" and on both of the scales that it does have, cannabis is lower than amphetamine, cocaine, and tobacco.

As for "lies", the research methodology is well documented. Unless you are claiming that the the researchers falsified the responses to their survey, there can be no "lies" here. You may disagree with the results, but the results are what they are, and, even if incorrect, they are not "lies". .     Jim . . . . (Jameslwoodward) (talk to me) 10:36, 23 June 2014 (UTC)