Template talk:Infobox drug/Archive 13

Making the template more useful for people
I propose we move all the identifiers to the end of the article into the Template:Authority control. These identifiers would than be removed from the infobox drug template. Doc James (talk · contribs · email) 07:33, 1 August 2015 (UTC)
 * strong support moving the identifiers down, seems very reasonable--Ozzie10aaaa (talk) 09:34, 1 August 2015 (UTC)
 * Strong oppose – I don't think this is even possible. None of the Infobox drug identifiers are support by Authority control template, nor do external authority controls even exist for these identifiers. Furthermore these identifiers provide useful external links that are of interest to a significant number of readers.  Like it or not, drugs are chemicals and it is important to display this information prominently near the top of the article instead of buried at the bottom.  Quite to the contrary, removing useful links from this template makes it less useful for people. Boghog (talk) 11:34, 1 August 2015 (UTC)
 * The infobox is fairly long. It would be fairly easy to support these identifiers in the authority control template. I could see keeping the ATC code but do we really need the DrugBank, ChemSpider, PubChem, CAS Registry Number and IUPHAR/BPS in the lead? The one in 10,000 who want this information would be able to find it at the bottom of the article. Doc James  (talk · contribs · email) 11:45, 1 August 2015 (UTC)
 * Moving the drugbox identifiers to the authority control template is mixing apples with oranges. Authority control provides bibliographic information about the subject of an article. The drugbox identifiers provide links to external chermical/drug databases. They serve two very different purposes.  And yes, we really do need DrugBank, ChemSpider, PubChem, CAS Registry Number and IUPHAR/BPS in the lead because the number of readers interested in this information is considerably larger than 1 in 10,000. Why are you so intent on purging chemical information from drug articles? This proposal is a really bad idea. Boghog (talk) 12:13, 1 August 2015 (UTC)
 * A number of these external databases such as ChEBI, ChEMBL, ChemSpider, DrugBank, and IUPHAR/BPS provide reciprocal links back to Wikipedia drug articles. This drives traffic to Wikipedia and helps to legitimize the Wikipedia drug articles as useful sources of information.  Furthermore this type of traffic is very valuable since it tends to be technically competent readers some of whom might add content to the Wikipedia article. These readers in particular would appreciate the links being prominently displayed. Boghog (talk) 12:39, 1 August 2015 (UTC)
 * I suggested moving some exceedingly detailed content or little general interest to lower in the article. I never proposed "purging chemical information from drug articles"
 * With respect to the reciprocity I see no mention of Wikipedia at pubchem, we are mentioned near the bottom at CHEMBL13, we are not mentioned at KEGG, we are not mentioned at UNII, we are not mentioned at ChemSpider, we are mentioned near the bottom at DrugBank, we are mentioned at the bottom at IUPHAR/BPS, we are not mentioned at ATC code and we are not mentioned at CAS Registry. Thus I do not see the couple that do reference us being able to complain if we move them lower in the article
 * We can create a Template:Drug identifiers if authority control is not good Doc James  (talk · contribs · email) 12:58, 1 August 2015 (UTC)
 * ChemSpider does reference Wikipedia, on a record that has as wikipedia entry there are 3 places you can see a reference, 1. In the names section there is a Flag called Wiki, which appears against the names and acts a link to Wikipedia, in the Data Sources tab (under the More heading) the wikipedia article is referenced, this also creates a tab called Wikipedia that displays the first paragraph of the wikipedia article and links to view or edit the article on Wikipedia. --The chemistds (talk) 14:28, 8 August 2015 (UTC)
 * Even if this information were of little general interest which I dispute, I don't see how this proposal improves anything. Logically, if these links were moved, they should be moved to an external links section, not to the very end of the article.  In addition, the infobox is not in the lead, it is to the right of the lead. In a short article (e.g., Parecoxib), the end of the infobox is placed near the end of the article.  What is the difference between having the identifiers placed at the right side of near the bottom in an infobox, or in the external links section of an the article?  The vertical placement would be about the same. In a medium sized article (e.g., Metoprolol), the only consequence of a long infobox is that the width of text in the first sections is narrowed. In an article with a long table of contents (e.g., Aspirin), the identifier section fills up the space to the right of the table of contents which is normally empty. This is a very efficient use of space. Boghog (talk) 20:37, 1 August 2015 (UTC)
 * If you look at the mobile version on a phone the infobox comes before anything else. More than half of our traffic is to mobile. Doc James  (talk · contribs · email) 14:32, 2 August 2015 (UTC)
 * Only about 1/3 of vertical size of the drugbox is devoted to identifiers. With or without the identifiers, it takes considerable amount of scrolling to get past the infobox before ones sees the lead. A better solution would be to collapse all the subsection of the drugbox in the mobile version so that only the chemical structure and name would be uncollapsed by default. Unfortunately the MediaWiki mobile front end does not appear to support this kind of fine tuning of how the infoboxes are displayed on mobile. Boghog (talk) 15:16, 2 August 2015 (UTC)
 * Was planning on making a similar suggestion for the "infobox medical condition" to move OMIM, MeSH, and DiseasesDB to something like the authority control template for similar reasons. Maybe leave the ICD code. Doc James  (talk · contribs · email) 16:03, 2 August 2015 (UTC)
 * As I already mentioned above, the authority control template would be completely inappropriate for this purpose. One would need to create a new special template. Another reason why splitting up these templates is a really bad idea is that it makes maintenance of this data by bots and synchronization with WikiData more error prone and difficult. Boghog (talk) 16:28, 2 August 2015 (UTC)
 * As this data should now all be in Wikidata the new template can just pull from Wikidata and synchronization will take place their. Added benefit is that it will allow synchronization between languages. Doc James  (talk · contribs · email) 01:44, 3 August 2015 (UTC)
 * As this discussion makes clear, template WikiData is currently a mess with a high number of discrepancies and is no where close to being ready for this purpose. Boghog (talk) 02:08, 3 August 2015 (UTC)
 * It is the best long term solution though IMO. It would make introducing similar content in other languages easier. Doc James  (talk · contribs · email) 11:13, 5 August 2015 (UTC)
 * Oppose – Re: Boghog's statement  Seppi  333  (Insert 2¢) 14:48, 1 August 2015 (UTC)
 * Comment - Just because a chemical has a drugbox does not mean that any given individual searching the article is interested in the pharmaceutical properties or associated information involving that compound. Throwing out chemical indentifiers and information on the assumption from the infobox on those grounds is just an anchoring fallacy.  Seppi  333  (Insert 2¢) 17:01, 2 August 2015 (UTC)
 * They are not thrown out. They are moved to a different place in the article. -DePiep (talk) 15:13, 4 August 2015 (UTC)


 * Support moving the technical details to the bottom, per WP:TECHNICAL. I am very sympathetic to the pharmacology/chemistry people  - it is essential chemistry information that provides a lot of information, compactly, but those fields are not important to your everyday WP reader, in my view.   We should definitely keep them in the article - just not at the top.  Especially not at the top, in view of the reality of WP's mobile display software. Jytdog (talk) 16:25, 2 August 2015 (UTC)
 * The solution is to fix the mobile display software. Boghog (talk) 16:32, 2 August 2015 (UTC)
 * Assuming it is not possible to fix the mobile display software, it would be better to collapse the identifiers section by default. At least that way, all the data is kept in one template. Splitting data between templates is going to create a lot of maintenance headaches. Boghog (talk) 16:35, 2 August 2015 (UTC)
 * I fail to see how including identifier data in the drugbox harms average readers, but as a compromise, would collapsing the identifier section be acceptable? (see sandbox version above). This would be much easier to implement, maintain and much more acceptable to readers that do value this information. Boghog (talk) 22:27, 2 August 2015 (UTC)
 * I like the "some stuff" entry :) I am OK with the collapse thing. Jytdog (talk) 23:04, 2 August 2015 (UTC)
 * Yup happy with collapsing. I think this is a fair compromise. Wondering if we could move the collapsed identifiers to just above the verify bar? Doc James  (talk · contribs · email) 01:48, 3 August 2015 (UTC)
 * collapsing sounds very good--Ozzie10aaaa (talk) 10:44, 3 August 2015 (UTC)
 * Please explain what is 'broken' in mobile view to be 'fixed'. Then about collapsing: (see my comment two bullets below) -DePiep (talk) 08:31, 4 August 2015 (UTC)


 * Strong oppose - Identifiers are not the same as Authority control. Whereas authority control is used to check whether we are talking about the same compound, the identifiers are assigned by organisations working in chemistry, and are a gateway to getting more information (either one follows the link to the organisation webpage on the compound and uses their information - pubhcem to get to publications for the compound, chemspider for other physical data and publications, or ones uses the identifier to find information elsewhere on the internet (copy the CAS, go to your preferred chemical supplier, paste the CAS in the their search engine and press 'search' to find their supply of compound, same goes for SMILES and InChI). I see the problem of the length of the box, and collapsing the section could be an option, I do agree that it is not really info that one needs to see (like one would for a boiling point e.g.).  --Dirk Beetstra T  C 03:25, 3 August 2015 (UTC)
 * By now, at this point, contributors agree that is should not be "authority control". End of subissue. The proposal still has that the externals links could be outside of the infobox. -DePiep (talk) 23:24, 4 August 2015 (UTC)


 * Oppose moving to authority control; that makes no sense in terms of data management and would be huge maintenance hassle. Collapsing works, though I'm not too excited about the basic argument here. Opabinia regalis (talk) 00:11, 4 August 2015 (UTC)
 * Comment about collapsing. Mobile view does not collapse a show/hide box. Mobile view always does "show" with no "hide" button. See for example SMILES and InChI for aspirin mobile view. So collapse them is not an option. Quite the opposite: we should design content pages and tables being uncollapsed. (Tip: in desktop screen, you can always check the mobile view through a link in the very last line of the regular page). -DePiep (talk) 08:31, 4 August 2015 (UTC)
 * After I wrote the above, I also realized this. I therefore created this request to add infobox collapse support to the MediaWiki mobile front end. Again, I think the problem is in the mobile front end and not the drugbox. Furthermore this display problem extends to a large number of other articles containing big infoboxes.  Perhaps the mobile front end should collapse all subsections of large infoboxes by default. In any case, I am still not convinced that the identifiers section of the drugbox harms average readers. Boghog (talk) 08:48, 4 August 2015 (UTC)
 * We'll see where the request leads. I'm on the fence in this. Maybe I can understand the design issues (like ACCESS) that mobile view is based upon, but not oversee.
 * As for the "harm": I support the core question by Doc James: what is that data doing there? IMO it is trespassing some page design & layout points, which I will call 'harm'. I don't know if collapsing would solve the issue(s). A Reader is enormously helped by stable layout & presentation &mdash; without realizing this. But when that is missing, the reader starts stumbling and leaving the article or site. -DePiep (talk) 09:50, 4 August 2015 (UTC)
 * I think you missed the point of of what Doc James was getting at. One could ask the exact same question of some of the links in the "Clinical data" section. However Doc James was not proposing to move all data from the infobox, just the pharmacological/chemical data in the identifiers section. Also how exactly are any of these links trespassing page design & layout points? Many, many infoboxes contain external links to databases. One could argue that not having these links in the infobox would confuse readers. Boghog (talk) 11:20, 4 August 2015 (UTC)
 * This is about collapsing, and I wonder if collapsing would solve the issue(s). And yes, some other data (outsided of subheader 'Identifiers') is eligible for displacement too, the Doc did not exclude that. I'll describe my take later on, in a new bullet I'm chewing on. -DePiep (talk) 13:03, 4 August 2015 (UTC)
 * Wait wait. It looks like editors line up for "collapse solves it". To be clear: Collapse is not a solution, as I explained above. We need a broader look at the topic. -DePiep (talk) 23:30, 4 August 2015 (UTC)


 * Notes on side-issues. These background topics might come in view, but that are not core to the proposal. To keep this talk simple, I think they should not enter as an issue here.
 * 1. For changes in Drugbox, I always keep an eye on the option to give Chembox the same treatment wrt that data. has some 20+ "identifiers". For example, the ATC Code could look & link the same in both; recently the IUPHAR/BPS ligand was unified in this. In this proposal, any outcome we reach here would be stronger if it can be applied to  as well. For simplicity, I propose not to add the  situation here: it is enough to do the mental check: "How would this work in " (answer "Just as well" is good).
 * 2. Most identifier links are actually WP:EXTERNAL LINKS (el). They should show formatted like (for aspirin):
 * DB00945 ✅, not current format DB00945 ❌: you are leaving the wiki site. el's are done this way in infobox medical condition, see cholera.
 * 3. See not MeSH.
 * 4. ATC Code (and possibly other data) is not exactly an identifier, but a classification. So today it is under the wrong subheader. For now we can leave it, but in any new template we should do this right right away.
 * 5. Technical, illegible and possible very long strings like SMILES and InChI might require a different treatment/showing.
 * 6. Here with the same issue was discussed.
 * -DePiep (talk) 09:37, 4 August 2015 (UTC)


 * Support with changes This idea was discussed also at Template_talk:Infobox_medical_condition in a February 2015 proposal that I made based on ideas and feedback that other people were sharing. See also Wikipedia_talk:WikiProject_Medicine/Archive_67 when I and user:Hexatekin (formerly OR drohowa) presented a model for separating cataloging data from human readable infobox data. The problem is that cataloging data for either drugs in this case or diseases in the other is not useful for most people, and other information in this space would be more useful. However, there is a precedent in about 10,000 drug and medical condition articles of putting this content in this place.
 * I agree with Boghog and other critics - "authority control" is not the right term for this information. I still think it is the right concept and the right precedent, and something like Template:Drug identifiers might be better or even Template:Medical identifiers so that the same box can be used for drugs, devices, therapies, medical conditions, and whatever other concepts in medicine are linked to some catalog identifier.
 * I am not convinced that people who have voiced opposition to this proposal recognize that these links will not be removed. The proposal is to make them less prominent by putting them anywhere else in the article in some other template, not remove them. Catalog identifiers which exist for computers to read should not be the first thing that Wikipedia readers see, and right now, this information which is useless to read (it is only useful as a link) is being presented for readers to examine in preference to anything else.
 * The positioning of the template should not be such a concern. The most obvious choice would be to follow the same positioning precedent which is used for the authority control template, because this is an analogous situation.
 *  Blue Rasberry  (talk)  13:43, 4 August 2015 (UTC)
 * I am not convinced that people who have voiced opposition to this proposal recognize that these links will not be removed. – We clearly do recognize it since what we are specifically objecting to is splitting data between templates which in turn will cause maintenance headaches. No one is suggesting that these links/identifiers be removed entirely.
 * Catalog identifiers which exist for computers to read – That is not at all the purpose of these identifiers. These identifiers are links to external databases that provide additional information about these drugs (see for example DrugBank: Linezolid and IUPHAR: ranitidine). Also these links are not the first thing that readers see on the desktop.  They are the first thing that is seen on mobile, but again, this is a general problem with how the mobile front end presents infoboxes, not a specific problem with the drugbox. Boghog (talk) 15:24, 4 August 2015 (UTC)
 * I spoke poorly - instead of saying that others do not understand the proposal, I should have said that I do not understand the perspectives of others.
 * I am not necessarily in favor of splitting the infoboxes, but that seems like the best way to achieve what I do want. I want the catalog data currently at the top of the article to be moved to the bottom of the article, and this could be done without splitting anything if it were moved in entirety to the bottom. In its place I would like a human readable infobox to be placed at the top. Help me understand - would moving the entire cataloging box to the bottom address your concern? Is it your position that this information needs to be at the top? I have been presuming that this box contains least useful information in the article in the sense that it is used by the least people. Instead of having it in the best article location, why not move it to the least valuable location, which is the bottom?
 * I am conscious of maintenance problems but this was sorted out for Wikipedia's one million biographies, which often have an infobox at the top and an authority control box at the bottom. I am not seeing evidence here that the situation for drugs and medical conditions is not analogous, and I am taking it for granted that the situation here is the same. What different do you see here as compared to biographies?  Blue Rasberry   (talk)  16:39, 4 August 2015 (UTC)
 * Moving all the "catalog data" to the bottom of the article partially addresses my concerns as pharmacological/chemical data are not singled out. However the clinical and identifiers sections also contains a lot more than "catalog data". These sections also contains links to highly relevant and high quality external databases.  In addition, drugs have more than therapeutic significance.  They also have pharmacological and chemical significance. Collapsing the display of information, if the mobile front end supported it, would be a reasonable comprise and preferable to burying the information at the end of an article.  The maintenance problems and the limitation of WikiData should be clear from this  discussion. The difference between bibliographic and drug data is that the former do not normally contain ambiguities whereas the later frequently do (e.g., tautomers, stereoisomers, salts, and close analogs). Boghog (talk) 21:24, 4 August 2015 (UTC)
 * I'm with Bluerasberry in this. I like the concept "catalog data", and I don't see the need to quote that as Boghog does (i.e., writing ""catalog data"" as an unaccepted wording). It is useful, even when we might need refinement later on. And as long as collapsing is not an option (because of mobile view), it is not eh an option. I won't do future solutions promises (and won't discuss them early). What we could use from here is a differentiation in what data rows are eligible for repositioning in the article, and which could/should stay in the infobox. -DePiep (talk) 23:13, 4 August 2015 (UTC)
 * I put "catalog data" in quotations not because it is unacceptable, but rather because it is an incomplete description of what is contained in the identifiers section of the drugbox. If the only thing the identifiers section contained were database registration numbers (e.g., CAS numbers, etc.) without links to external database then this would be "catalog data". But as Dirk Beetstra has pointed out above, the identifier links are gateways to additional information about the subject. There are no deadlines in Wikipedia. Just because collapsing is currently not implemented in the mobile front end does not mean that we must immediately implement the only solution that is currently available to us. It may make more sense to wait and do it right. Boghog (talk) 04:03, 5 August 2015 (UTC)
 * The mobile/nomobile css classes may be a workaround (see for example the Seasonal summary and predictions time line in 2002 Pacific typhoon season). These css classes could be used to display a smaller alternative infobox on mobile devices. Boghog (talk) 05:56, 5 August 2015 (UTC)
 * I am not prepared to discuss collapsing or development of the infobox software because I need more information to understand why you would suggest that this information needs to be in the infobox anyway. I continue to grant this content's importance to the Wikipedia users who need it, but I am assuming that fewer than 1 in 10,000 people would find this information or these links meaningful and that the infobox should contain information which is meaningful to nearly all readers. I think that you would agree with me that few users of the article could use this catalog information or follow these links, but let me know if I am missing something. Please tell me again -
 * Considering that the catalog data and links are meaningful to so few people, why do you believe that they should be positioned such that practically all Wikipedia readers will browse them?
 * Considering the precedent in Wikipedia biographies in which there is an infobox at the top with human readable information and about 1 external link, along with an authority control box at the bottom with catalog data and external links to things like the VIAF identifier and WorldCat, to what extent can that model apply here? There seems to be no great maintenance problem for having an infobox and a catalog data box for biographies, so is there any reason to anticipate one here? Also, is there any reason to believe that users of chemical would be inconvenienced by finding the catalog data at the bottom, or that biographical catalog data is used differently or is not analogous to this chemical data?
 * Thanks.  Blue Rasberry   (talk)  14:53, 5 August 2015 (UTC)
 * Concerning your first point, where are you getting this fewer than 1 in 10,000 people would find this information or these links meaningful? Granted, most readers would not find this information useful, but technically inclined readers are more likely to access, read in depth, and critically evaluate these articles than the average reader. Even if the "less than" estimate where accurate, technically inclined readers are more likely to add insightful comments on talk pages and add high quality content to articles than the average reader.  These links provide these technically savvy readers fast access to additional information that will help them critically evaluate the article, and if lacking, improve it. Hence the importance of these readers to the project is far in excess to their relative numbers. Boghog (talk) 20:36, 5 August 2015 (UTC)
 * Concerning your second point, this is essentially an OTHERTHINGSEXIST argument. Biography infoboxes are not a good analogy since their very existence is controversial whereas infoboxes for scientific topics are much more widely accepted. Biographies generally have many fewer external links because there are many fewer systematic databases for people compared to drugs. There are plenty of other types of articles that have multiple external links in infoboxes including articles that are much more related to drugs than biographies. The most closely related are chemicals (chembox), and drug targets (GNF Protein box, enzymes). Because of the precedence provided by these closely related articles, readers who are looking for this information would expect this information to be in the infobox. Hence these readers would be inconvenienced because the data they are looking for in not in the place that they expect.  Worse yet, they may completely miss the data. Conversely, how are averaged readers harmed by having the information near the top of the article, especially if it were collapsed?  The maintenance problems are caused by splitting the links between two templates which makes it more difficult for bots and humans to cross check the data.  Also the position of the second template may be changed depending on the arbitrary whims of editors (in the external links section, before the authority control, after the authority control).  It is much cleaner to have all the links in one one template that has a predictable location at the top of the article. Boghog (talk) 03:51, 6 August 2015 (UTC)
 * What percentage of page visitors do you think are able to use this content? For something to be in the infobox, to what percentage of readers should content be useful? I told you that I think infobox information should be useable by nearly 100% of visitors. Already this text is data which has no meaning outside of being followed as a link, so it cannot possible be useful except to people who follow the link. That is a major strike against including this in the article's best location. I get the 1 in 10,000 figure because that is the best traffic which links in citations get, so far as I have seen. Also I think it is a good estimate of ratio of professionals to general public who visit popular drug articles. Suppose the number were 1 in 100 - do you feel that would be enough? For being at the top of the article I think that is contrary to any precedent or analogous situation on Wikipedia.
 * Thanks for sharing discussion about the old infobox discussion. I was and still am unfamiliar with this discussion, but it seems to be an argument over routinely using infoboxes and also about what data should go into them. I think everyone here is in agreement about routinely using them and also I am not arguing about fields presented - I only am questioning the need to put it at the top of the article. I agree with you that it is best to have one infobox, but that has to be balanced with what an infobox can hold. I might as you what length you think an infobox should be, whether they can growth indefinitely long, or whether eventually there should be an editorial decision to only include a limited number of things then put more elsewhere. I have two ideas in moving this content - first I want it deprioritized so that not everyone sees it, and secondly I want it replaced with consumer information more like the Mayo Clinic information that Google's Knowledge Graph presents. I feel justified because having Google's marketing researchers suggest what fields are most important seems good to consider. Fields which they include are the follow: prescription or OTC, use, route of administration, link to Library of Medicine, brand names, molar mass, CAS ID, pregnancy risk, diseases which this may treat, and drug class. The Wikipedia way is to include some more things that those, but at least I would more along the lines of "what it does, what does it treat".
 * I continue to acknowledge maintenance problems with a second template. I also would prefer for there to be uniform templates between drugs and proteins, but I have more sympathy for the layman public using Wikipedia as their primary source of health information as compared to sympathy for protein researchers who do important work but on the whole are more equipped to learn a change in layout due to their higher than average education, income, etc. An initial compromise could be to load more consumer information into the infobox along with the other information, but I expect that by greatly increasing the size of the box that will draw more people into the conversation and I am not sure what would happen in that case. I would like to find compromise that everyone likes.  Blue Rasberry   (talk)  15:55, 7 August 2015 (UTC)
 * A compromise that most everyone likes is to collapse the identifiers section. We can also include more consumer information while collapsing other parts of the infobox (e.g., the 3D structure) so that length of the infobox is kept within reason. The only problem is that the mobile version uncollapses collapsed portions of infoboxes. (A closely related problem is that collapsed navboxes are not displayed at all on mobile.) The solution is to fix the mobile front end. This problem extends far beyond infobox drug. The display of any large infobox on mobile is problematic. A potential temporary workaround is to use mobile/nomobile css classes to display a smaller alternative drugbox on mobile. Boghog (talk) 09:40, 8 August 2015 (UTC)
 * re '... everyone likes': Not to redo earlier arguments, I just note that collapsing is bigger than just box size. It is also about having to interact (click) to read content (w3c accessibility, page design for mobile). And about: does that info belong in the infobox? -DePiep (talk) 20:57, 8 August 2015 (UTC)
 * Resolved for me! I am comfortable leaving these links in the infobox regardless of collapse functionality, just so long as there is no objection to also adding more human readable information to the boxes.  Blue Rasberry   (talk)  11:34, 10 August 2015 (UTC)

Arbitrary break #1

 * Support development. Even OP Doc James agrees it should not be through Authority control, but for now we have working title,  and let me add  to cover even more.
 * IMO, the topic boils down to article-specific links to external databases (not just identifiers; could be classifiers like ATC Code too). Strictly speaking, that is not information needed in the article body, because the body should be self-contained wiki. Also, it is not about the authority defining an identifier to the substance. In that case, we wouldn't have to link to that database/site but just write, unlinked: . To cut it short: the link is an external link, and so should be place in section External links. A template could easily provide default links (as the template now has). -DePiep (talk) 15:10, 4 August 2015 (UTC)
 * Per WP:ELPOINTS, external links in infoboxes are allowed. There are many, many infoboxes that contain external links, not just the drugbox. Boghog (talk) 15:31, 4 August 2015 (UTC)
 * It says: EL can be "in the appropriate location within an infobox, if applicable". Good example I can think of is the website of a company in the company's article. But sixteen as in Drugbox is 'appropriate' too? Twenty in Chembox? Anyway, "other infoboxes have them too" may be an otherstuffexists reason. But I'm looking for considerations. -DePiep (talk) 15:40, 4 August 2015 (UTC)
 * The otherthings exits counter argument only applies to cherry picking evidence. "Many, many examples" evidence in contrast demonstrates that the practice is long standing and widely accepted. Boghog (talk) 20:54, 4 August 2015 (UTC)
 * I don't understand what you say. If you have good examples of EL in infoboxes, please expand (e.g., demos). Stating "many other infoboxes have EL" is Otherstuff and is not an argument. That is what WP:Otherstuff is about. I repeat: I'd like to read which other situations make an argument for inclusion. -DePiep (talk) 22:16, 4 August 2015 (UTC)
 * The argument for inclusion is simply that these links provide valuable additional information about the subject. Boghog (talk) 04:31, 5 August 2015 (UTC)
 * Yes a new box of identifiers for the external links section I think would be good. Maybe one or two in the infobox but not so many.
 * This would be simpler to the situation with brandnames. We put a couple in the infobox but not dozens. Also with external links for infobox medical condition. We put a couple in the infobox and rest in the EL section. Doc James  (talk · contribs · email) 11:06, 5 August 2015 (UTC)
 * All the external links should be kept either in the drugbox or in a new template placed in the external links section. Splitting the links between two templates located at the beginning and end of an article respectively would confuse readers and would be a headache to maintain. A cleaner long term solution is to keep all the links in the infobox and collapse (both on the desktop and on mobile) the less important ones. Boghog (talk) 11:31, 5 August 2015 (UTC)
 * re Boghog: you noted that lots of other boxes have ELs. I was asking for the argument in those situations (as in: how is that EL addition applicable). Of course, everything added to an infobox (that is not nonsense) adds information. However, infoboxes have restrictions (otherwise, we'd be adding the full article prose in the infobox). I am searching for more sound reasons to add ELs. -DePiep (talk) 11:15, 5 August 2015 (UTC)
 * What I think distinguishes an external link in an infobox from a link in the external links section is (1) the degree of relevance and (2) the fraction of articles where this link would be appropriate. According to WP:EL, all external links should be relevant in some way to the article. The bar for inclusion of an external link into an infobox should be higher.  The link should be highly relevant to a major topic in the article.  Links that are especially relevant to an infobox are to specialist databases that systematically cover the subject matter of an infobox. In the case of the drugbox, links to individual drugs in drug databases would be both highly relevant and applicable to a large number of articles in which the drugbox is transcluded. Boghog (talk) 12:42, 5 August 2015 (UTC)
 * re your "specialist databases". I do not dispute the current links are relevant. Of course they are, the rest is deletable nonsense. My quest is: how are they important enough to be in the infobox? They are an EL. Most other ELs in an infobox are defined by, and limited to, the topic (e.g., Coca Cola inc topic has one).


 * My second repetition note is, that say the DrugBank link has nothing to do with the wikipedia article. It is a new road to information. Interesting, relevant, but not part of the article. Maybe we shold treat is as a source (we could say: "The DrugBank id = ABC[1]", and that's the ref. -DePiep (talk) 20:13, 9 August 2015 (UTC)
 * Deletable nonsense?? What I wrote is common sense. Also where is it written that infoboxes can at most contain one external link? There are many, many infoboxes that contain little more than external links. Furthermore the external links guideline specifically allows external links in footnotes and in infoboxes. The guideline does not say only one external link allowed per infobox. Finally why would you move the external links to footnotes?  This is necessary when referring to an external link the in the text of an article, but again, infoboxes are exempted from this requirement. Doing this would make it harder to readers to follow the links (requires two links instead of one) and unnecessarily clutters up the footnotes. Infoboxes are really no more part of the article than the footnotes.  Both are supplements. Boghog (talk) 20:57, 9 August 2015 (UTC)


 * re Deletable nonsense: I meant to say, everything that is not related to the article is nonsense (think: vandalism). Not our discussion.
 * Next, all current present ELs are under discussion here. I don't think all these should be in the infobox. Do we recognise that not each and every drugbox EL data row is OK? Some could stay, but then all 20+ in ? -DePiep (talk) 21:18, 9 August 2015 (UTC)
 * re Also where is it written that infoboxes can at most contain one external link? . Dear : no one claimed so. If that were a rule, I'd claim it and would not spend time on you.-DePiep (talk) 21:24, 9 August 2015 (UTC)
 * OK, sorry for the misunderstanding. Yes, I would agree that many of these infoboxes have gotten out of hand and should be trimmed. Boghog (talk) 21:38, 9 August 2015 (UTC)
 * I met an ec :-) when I tried to 'revert'/nulledit my prev harsh ec es. Glad to stay on topic. -DePiep (talk) 21:42, 9 August 2015 (UTC)

Cost information
This is of a great deal of importance to both healthcare providers and the general population. This website provides global data per dose and provides the ration between high and low costs.

Am looking at adding the info to Wikidata but Wikidata does not yet take dosing information. I guess the question is do we add it to En Wikipedia to be moved latter to Wikidata? Doc James (talk · contribs · email) 07:36, 1 August 2015 (UTC)
 * I believe this is a very good idea (it could help many readers- both patients and providers)...therefore I support--Ozzie10aaaa (talk) 09:30, 1 August 2015 (UTC)
 * It would also interest policy makers, investors, economists, and anyone else interested in money or financial matters. I support including this information.  The only real question is, how?
 * I would be happy to have the details written up in sentences and paragraphs in a ==Cost== section. I'd be happy to have a table there that lists the median price (or both the high and low) for as many years as we can get that information.  This could presumably be assembled a bot.
 * If we are going to put it in the infobox, then perhaps just the median price, with an external link for more information:
 * Global wholesale price     US $0.20 per pill (2014)
 * (I used Truvada as an example; for things with multiple forms and doses, like penicillin, it'd be much harder to get one number—but that's easily handled: you just skip those cases.)  I can't really see adding multiple lines of cost information to such a huge infobox.  Doing all of the above might be best, of course.  What would you choose to start?  WhatamIdoing (talk) 22:00, 1 August 2015 (UTC)


 * The source is International Drug Price Indicator Guide. They have a big disclaimer: "What this Guide cannot tell you: Exact information on prices or any other descriptive piece of information. No guarantees are implied in the list as to price changes, quality of products, or specific sales conditions".  Here are their sources - most of it comes from agencies servicing the developing world.  I am not sure how much of that information is relevant to the US market, or the EU market.  They also make it clear that those prices are nothing like what somebody pays at the pharmacy - the info is there for purchasing agencies, I believe.   We would need to explain what the price listed there actually means.  Also I did a bunch of searching in it - lots of information but it all seems to be about generic drugs.  No celecoxib, no gleevec, no gardasil.  Their coverage of essential generic meds seems good though -they have Oseltamivir and Alprazolam and atorvastatin for example.  So hm.  Jytdog (talk) 00:59, 2 August 2015 (UTC)
 * I see the price information of essential medicines as more important as you can 1) find these in most areas of the world 2) they are essential medicines.
 * One can than supplement this with information from drugs.com for what the stuff costs in the USA (which has the most expensive meds in the world). Doc James  (talk · contribs · email) 11:26, 2 August 2015 (UTC)
 * How do we explain to readers simply that these are not retail prices nor even what any agency or PBM might actually pay? Maybe more importantly, how do we keep this field maintained?  Once it is available people are going to add all kinds of crazy stuff there - I see it is as significant future pain in the butt to keep high-quality and contextualize.  For example, as discussed here Alirocumab, people are up in arms about the price of the new PCSK9 inhibitor which is about $15K/year, which looks terribly high compared with generic statins.  What the NY Times pointed out, is that this drug doesn't replace statins; it is for people whose cholesterol isn't controlled with statins and diet, and who undergo apheresis (like dialysis) to clear cholesterol from their blood - a procedure that costs about $96K/year.  PBMs are really upset, of course, because costs have shifted to them, but the system overall will save money. (If the drug remains restricted to those people; if the label expands into the general population, that will become a big problem, if the price doesn't come down).   A raw number in a field doesn't provide context like that. Jytdog (talk) 15:44, 2 August 2015 (UTC)
 * The cost difference between stuff like clarithromycin and amoxicillin however is significant. And we should make this more clearly apparent. Same for sumatriptan and rizatriptan though their is likely little clinical difference between the two.
 * The Global Wholesale Price does partly match the retail price. Stuff with a higher wholesale price has a higher retail price. Doc James  (talk · contribs · email) 16:06, 2 August 2015 (UTC)
 * Partly yes, but the field is going to be a point, not a dataset. People can look at one article and then the other, yes... but humans being human, readers are going to see a dollar price and think that is the price they should pay, and when they find a mismatch when they go to actually acquire the drug, they will think WP is messed up or they are getting ripped off, and neither will be true.  Maybe if we call it something like "agency purchasing price" instead of just "price"? Jytdog (talk) 16:19, 2 August 2015 (UTC)
 * Agree we will need to find the correct term as what we are talk about is not retail price. Doc James  (talk · contribs · email) 13:32, 3 August 2015 (UTC)


 * Oppose. This is market information. I wonder if and why WP should add this: how is it encyclopedic? I know of no other topics/articles that do market information. Maybe currency exchange rates has, and stock exchange listed companies. Also, drug markets are not free and open (as the currency and stock markets are). Drug companies apply secret pricing, and there is government interference (aren't HIV drug for Africa payed for by WHO?). Then there is the practical side: markets are played regional and time-based. We'd need an updating habit for say 200 countries (or 10-50? regions), and a timestamp with each. This is much more than the existing examples, which can be stated to be world-wide (at any moment, there is just ~one price). To compare, already we have to look at the list of jurisdictions for drugs (legal_AU). But mainly, I'm hoping to find a good argument on whether this is/isn't encyclopedic. We're not Bloomberg. -DePiep (talk) 10:13, 4 August 2015 (UTC)
 * Cost details are here for the iPhone.  Doc James  (talk · contribs · email) 12:28, 4 August 2015 (UTC)
 * Yes. Not in the infobox, not per telephone type (as is proposed for drugs), and I could easily agree this iPhone pricing is an interesting development. Will does the proposal for Drugbox include production costs? It occurs to me that iPhone cost development is relevant in historical view and for multiple angles (production site issues, marketing policy). Could be interesting too for HIV drugs. IMO still not encyclopedic for 6000 drugs. -DePiep (talk) 12:52, 4 August 2015 (UTC)
 * A discussion in the text is probably a first place to start with inclusion. The cost of the 300 or so essential medicines is well discussed in the literature as this plays an important role in deciding if they should be included in the list. Doc James  (talk · contribs · email) 10:59, 5 August 2015 (UTC)
 * Doc, your "... to start with inclusion" means to say: 'in article text is best place to start price mentioning'? -DePiep (talk) 18:00, 5 August 2015 (UTC)
 * Yes I am not sure how to distill pricing information into an infobox format yet. Add text on cost and putting that price into context in useful and likely a good way to start. Doc James  (talk · contribs · email) 06:28, 9 August 2015 (UTC)


 * Support Inclusion of prices has been a regular topic of discussion, most recently at Wikipedia_talk:WikiProject_Medicine/Archive_67 when this proposal was made without having a comprehensive price database ready to share.
 * I think the inclusion of price data for drugs is worthwhile to share because a history of precedent in Consumer Reports publishing has found that people in the United States demand and use this data. I work for Consumer Reports, a nonprofit organization which has for decades published the price of drugs in the United States. Through Consumers International partner organizations publish drug prices in as many places worldwide as possible. I have always wanted to contribute some of Consumer Reports commentary on pricing but I know that this is difficult to do because of difficulty in getting source data. If we have good source data to share then I would like it shared in Wikipedia, because Wikipedia has little coverage on how the price of medical treatment is a barrier to people accessing care. Physicians routinely include patients in decisions about what drugs they will take but physicians almost never provide pricing comparisons. If that information were in Wikipedia as it is in Consumer Reports then I think the same proven interest in pricing information that Consumer Reports has identified would also be present in Wikipedia.
 * The reason why it is important to share this is because for most people, normal typical people, getting a drug is a consumer marketplace decision. This is true in the United States where uniquely among developed countries the healthcare is not nationalized, and it is just as true in countries with less developed economies where by necessity drugs are distributed through the marketplace instead of only through dispensaries with government oversight.
 * Please reconsider the past discussion on pricing to which I just linked in the June 2015 talk in WikiProject Medicine archives. There are some barriers to overcome here but I like the concept a lot.  Blue Rasberry   (talk)  14:04, 4 August 2015 (UTC)
 * We can share it, but how is it encyclopedic? -DePiep (talk) 22:19, 4 August 2015 (UTC)
 * I discuss pricing information fairly frequently with patients. The price is an important aspect of a pharmaceutical so yes encyclopedic. Whether or not it is generic also plays into it. Accessibility is also encyclopedic. The fact that a drug may be available in the US but not Canada or Europe is important. Doc James  (talk · contribs · email) 11:02, 5 August 2015 (UTC)
 * It is encyclopedic because reliable sources like Consumer Reports have established that the price of a drug is a necessary part of discussing the drug. Encyclopedias cover precedent so the precedent is the first reason. A philosophical reason for why it is encyclopedic is because it is not possible to understand a drug's usefulness without understanding its price. A Wikipedia article which explains the uses of a drug is missing crucial information if it neglects to say that the drug is out of reach for a patient or even entire population because of its price. Price also is helpful in communicating unnecessary health care, as in places of excess, consumers tend to think that higher priced drugs might be better for them even if their doctor and the available evidence suggest that they should choose less expensive alternatives. In practically all major medical lawsuits, and certainly in all cases listed at List of largest pharmaceutical settlements, the issue is that an expensive drug is used in haste, and never because there was some error with an older, cheaper drug. There is some relationship between drug prices and the largest pharmaceutical industry media circus lawsuits.
 * A likely argument for saying that pricing data is not necessary to understand a drug is to assume that price is not an influence on how a drug is used or its place in society. This argument probably could not withstand scrutiny. If there are other arguments for not presenting price, then please share, or please reconsider those arguments from the archives which have already been given. I agree that many people intuitively object to providing price information.  Blue Rasberry   (talk)  14:04, 5 August 2015 (UTC)
 * re you both. First, let us assume we can find a form to write price information in infobox style. It can't be prose, but somehow it says "Costs: $ x,- to y,- per dose treatment in the US under Obamacare insurance rules". I think I'm addressing some basic cost parameters this way. This illustrates that the "cost of a medicine" depends on many, many parameters. Already the example is narrowed down to a specific situation (US, Obamacare insurance rules), I still can not see how this would work for the numerous regions/countries and insurance policies. Also, for the range of treatments that are usually possible for a single drug (number of doses, amount per dose) and any, any pricing policy by the drug industry and patent issues. With this, let's use this example.
 * It still does not say anything for the accessibility of a drug (as in: which patients use the drug or is the price prohibitive?). I understand this is what Blueraspberry means when saying "understand a drug's usefulness without understanding its price": of course not the curing, clinical usefulness, but the possibility to reduce a specific disease in a population. The price, even when detailed as in the example, does not answer this question. So yes, the set pricing parameters may be encyclopedic in certain situations, but even then it is not fit for an infobox. In other words: if the pricing is that relevant to the general public being cured, it better has a paragraph in article text. HIV treatment in Africa comes to mind. (Just as the iPhone pricing history has, the earlier example in this thread).
 * An other solution could be if a relative-cost-indicator (say, y Consumer Reports) were available. Note that this would require a per-country (or market, region, ...) list.
 * A third option would be to add an external link to the International Drug Price Indicator Guide. In the opening page is has a "This is/This is not" list, that says it does price comparision (consumer information), but that might be the "non-encyclopedic" trigger. Does (not a) WP:LINKFARM apply? (to be looked at). -DePiep (talk) 19:02, 8 August 2015 (UTC)
 * TL;DR: the price infirmation relating to the usefullness (how accessible is the drug for the public) has many parameters. If it is (encyclopedically) relevant to describe the popular effect of the drug (through access to medicine), it needs textual description not an infobox shorthand code. And there are many regions or countries to cover. -DePiep (talk) 19:33, 8 August 2015 (UTC)
 * Sure I think that is a good place to begin. Doc James  (talk · contribs · email) 07:09, 9 August 2015 (UTC)
 * Doc, I hope you mean like this: "Drugs where price and marketing and public availability circumstances are relevant for its effect, should have a section in the body text". (I add: the infobox can not solve this). -DePiep (talk) 18:46, 9 August 2015 (UTC)
 * What you describe is also what I imagine. What I read into what you are saying is that the dataset will not only be able to give prices, but also have other information. If a price of a pill is given, then that price also has to be matched to the dose of the pill, date of the price, region of the price, and cultural context like is the pill subsidized by most insurance or by government subsidy. Right now we hardly manage any databases in Wikidata and drug pricing will not be an easy place to begin. An issue you did not raise was price of generic versus branded, or price of a family of equivalent drugs when patients and doctors agree that any of a set of drugs may be tried. No one can promise that Wikipedia / Wikidata can manage these complications immediately. Right now, I want to explore options and add what is possible without introducing misleading information. The rule which is being tested here is WP:PRIMARY - we are saying that Wikipedia readers want a new kind of data which we have not routinely given before. I want to be cautious but at the same time, I support anyone's attempt to try a few new things and see what is possible for a trial. I agree - so many of the things you listed are problems which no one is ready to address, and I am not sure what the answer will be. In some cases data can be only in the infobox, but in other cases, the infobox data would make no sense without an explanation in the article text.  Blue Rasberry   (talk)  19:22, 11 August 2015 (UTC)


 * Comment - IDK how pharmaceutical pricing works outside the United States, but the actual price that a consumer pays for a given drug at the point-of-sale in the US can vary significantly between different health insurance plans. E.g., I changed insurance plans about 2 years ago and the price of an Rx that I regularly fill nearly tripled (this higher price is about 10% of the drug's retail price). In contrast, with my new plan, the price I pay for generics of the most commonly Rx'd pharmaceuticals (e.g., drugs like fluoxetine and sertraline) is about an order of magnitude less than the price I'd pay for the same drug under my old plan. Given the high rate of variability in pharmaceutical pricing between insurance plans, I don't see how wholesale price info would be useful unless it's provided together with price info for related pharmaceuticals - even then, some health insurance plans include tiered pricing for the insurer's "preferred" drugs (e.g., patent-protected drugs may be priced in a more expensive tier if substitutable generic drugs are available) for a particular indication, which reduces the relevance of retail/wholesale price comparisons.  Frankly, I think it's more informative to indicate the availability of generics instead of provide actual price data.  Seppi  333  (Insert 2¢) 00:44, 9 August 2015 (UTC)
 * re Seppi333: yes. The topics better be covered by articles about: "How is drug price made by: insurance -- patent -- markets -- ... -- policies". That is the encyclopedic approach. (WP describing the workings, and any thinking reader then can detect systemic failures :-) ). -DePiep (talk) 18:54, 9 August 2015 (UTC)


 * Someone is paying a set price if not you. As pharmaceutical pricing vary from less than a cent a day to more than 250000 cents per day a rough ballpark is useful IMO. Doc James  (talk · contribs · email) 06:26, 9 August 2015 (UTC)
 * "a rough ballpark is useful IMO". Is that a "rough", 1 to 250.000? To me, that is non-info. You say, any number would be OK in this? Really? Even then. I understand that is a price per day (better be per treatmnent, but alas). Now please refine: what is the encyclopedic meaning if WP states: "this drug costs US$ x per dose/day/treatment"? Or, in your view: "this drug costs US$ x to x00000 per dose/day/treatment". And I omit the obvious Q: how about non-US? -DePiep (talk) 18:39, 9 August 2015 (UTC)
 * Sure so currently we say nothing which means every page states "1 to 250.000" which is useless. What is much more useful is "Amoxicillin is available as a generic medication. It has a wholesale costs between 0.02 and 0.05 USD per pill. In the United States ten days of treatment costs about 16 USD. "

Doc James (talk · contribs · email) 04:35, 10 August 2015 (UTC)
 * ... this in an infobox? US only? If it is $160, is it accessible? DePiep -00:10, 12 August 2015 (UTC)
 * Doc, I'm not that familiar with English (let alone US colloquial). Why not introduce three examplary drugs with a rich price relevance. -DePiep (talk) 19:42, 9 August 2015 (UTC)
 * Not sure what you mean by "three examplary drugs with a rich price relevance"? Do you mean you want three expensive medications to see how this would compare to three inexpensive medications? Doc James  (talk · contribs · email) 04:51, 10 August 2015 (UTC)
 * It think we see what it would look like if we have develop some examples. -DePiep (talk) 00:08, 12 August 2015 (UTC)


 * Comment I assume if the price is added to any drug article, it should be always clarified that it's within the United States and as such the currency is USD, with the relevant year, unless the reference says otherwise. However, that may be US-centric. Brandmeistertalk  11:07, 11 August 2015 (UTC)
 * Much of the pricing info for essential medicines is from the developing world. However it is in USD as that is still very much a global currency. Doc James  (talk · contribs · email) 13:17, 11 August 2015 (UTC)
 * That's already a simplification. There is not just a price in USD, it's also about price in the US. Also, there are marketing regions, insurance rules. And the treatment is not simple for one drug (eg period, drug dose per day, ...). There is no such thing as "The COst Of A Drug". -DePiep (talk) 00:08, 12 August 2015 (UTC)
 * Yes agree there is not one cost but there are "costs for a medication" or a "range of prices". The cost vary by manufacturer / supplier / region. But the same applies to all commodities. We have an article here Gasoline_and_diesel_usage_and_pricing for example. Doc James  (talk · contribs · email) 06:11, 12 August 2015 (UTC)
 * By the way MSF has agreed to release their complete medication / medical equipment price list for us to use. Doc James  (talk · contribs · email) 06:14, 12 August 2015 (UTC)


 * Yes, a full article on gas pricing! That's how complicated it is (as in: many many parameters are in play). Just as the pricing policy history of the iPhone has a full section. We can not squeeze that into an infobox. It's not a "melting point at STP", or a "company's stock value per x-y-zzzz (dae)", or a "currency value against USD per x-y-zzzz". Just to get a feeling: what would be the lefthand text in the i'box? Or another idea: should we create a template like   for relevant drugs? Sounds like 'typical' is the escape route from huge lists of information. It's won't be consumer info though, nor useful in Doc's consult discussion with patients. -DePiep (talk) 08:29, 12 August 2015 (UTC)
 * One could have an upper and lower cost limit ( X to Y dollars ). Agree it is best to work on it as text first. Doc James  (talk · contribs · email) 10:44, 12 August 2015 (UTC)

Semi-arbitrary break #1

 * Altogether, few new arguments lately IMO. We could wrap this up. Maybe into a different proposal. -DePiep (talk) 19:28, 13 August 2015 (UTC)

CAS RN should link to CAS
At the moment the CAS_number input links to the external MeSH site. I think it should link to the CAS site. There could be an extra link to the MeSH site. And those external links should be id-specific, not just search help.

For example aspirin, 50-78-2. now links to:
 * ❌

That is using the CAS RN to the MeSH-site, and not even reaching the aspirin special page!

Were the Chembox used for the CAS RN link, it would link:
 * ✅

And for example the MeSH link for carbon dioxide:
 * ✅

Questions, proposals:
 * The CAS RN data should actually link to the CAS site, specific target page.
 * We can add a MeSH option and add that link to the specific MESH page directly.

Thoughts? -DePiep (talk) 22:04, 25 July 2015 (UTC)
 * I reopen this point. It is not correct then the CAS RN links to MeSH (and a bad page at that). -DePiep (talk) 20:12, 6 October 2015 (UTC)
 * Can't say I know too much about this but it doesn't seem like anyone wants to input. If the commonchemistry.org link is more official than the MeSH link then that is the link we should use IMO. Sizeofint (talk) 17:40, 12 October 2015 (UTC)


 * If a MeSH external link were needed, it should be to the right MeSH page and it would need its own parameter (name-like). At Chembox: see MeSHName in Chembox Identifiers. -DePiep (talk) 14:18, 24 October 2015 (UTC)
 * If a MeSH external link were needed, it should be to the right MeSH page and it would need its own parameter (name-like). At Chembox: see MeSHName in Chembox Identifiers. -DePiep (talk) 14:18, 24 October 2015 (UTC)