Template talk:Infobox drug/Archive 5

Time for new drugbox?
I've been noticing recently that WP has many articles on combinations and associations—we have unique articles on co-trimoxazole, ampicillin/sulbactam, quinupristin/dalfopristin, co-codamol and Preterax (which should probably be moved to perindopril/indapamide), just to name a few. Drugbox isn't, in my humble opinion, the most appropriate option for these, as there is no real need for a structure, such comprehensive chemical data etc. as these will be in the individual components' articles; indeed, many articles on associations do not transclude Drugbox.

So, I have been considering the possibility of our creating a new template for combination medications. It could be a condensed form of the regular Drugbox, with some modifications. The modifications I've thought would be appropriate so far are:
 * ATC code would be the only required identifier;
 * The "Chemical data" section would be eliminated altogether (as in my current draft) or made optional;
 * Should it be made optional, perhaps a plain "Chemical formula" would no longer place the page in, so as to allow CnHnOn · CnHnOn-style formulae?
 * The  parameter would become , to more intuitively allow changing the drugbox name on articles with no image;
 * Perhaps most importantly, the "IUPAC name" section would become a "Combination of" section to note the components (identified by,   etc. parameters).

I have boldly created a draft in my userspace here, and a working example may be seen here; I've chosen ampicillin/sulbactam. I would greatly welcome any input on this: a clean-up of the code (I've still a lot to learn WRT templates), suggestion of other modifications, and most importantly, whether or not this is a good idea! I think so: Drugbox-mab worked beautifully and is currently in use in over a hundred articles. If anyone would rather post at WT:PHARM, that's fine by me. Fvasconcellos (t·c) 21:16, 24 July 2007 (UTC)
 * I like it the idea. I don't have the ability to clean up the coding, but I agree that this is needed. Remember 21:34, 24 July 2007 (UTC)
 * I am pro, too. All new and old drug boxes should be described and categorized for making their context clear. JKW 21:59, 24 July 2007 (UTC)
 * I like the idea, but rather than creating a new infobox, it probably would be more maintainable if it were handled like monoclonal antibodies, and add a new mode to the existing box. Are we considering Sultamicillin a single substance, or two? --Arcadian 23:37, 24 July 2007 (UTC)
 * Well, something like Drugbox-mab is what I meant; a "sub-Drugbox", separate in practice. Sultamicillin is the single-compound form of ampicillin/sulbactam; it's an ester of both rather than a mixture of both. Fvasconcellos (t·c) 23:39, 24 July 2007 (UTC)
 * I dig it. I think it would be nice to keep it in one template though, as Arcadian has suggested. How would multiple bioavailability or metabolism entries be handled for each substance cleanly, in any case? (Ccroberts( t · c · g ) 23:43, 24 July 2007 (UTC))
 * Judicious use of  tags, as is (IMHO) standard Drugbox practice? :D We'd probably name it Drugbox-something—Drugbox-combination or Drugbox-association, then? Fvasconcellos (t·c) 23:53, 24 July 2007 (UTC)
 * By the way, would you rather see something like the top right or something like the bottom right? I'm afraid of introducing too much into the Drugbox (i.e. drug class) and obviating the need for actual article content :D Fvasconcellos (t·c) 00:09, 25 July 2007 (UTC)


 * Ya, I think top right would be sufficient. Drugbox-combination sounds good, also. (Ccroberts( t · c · g ) 00:12, 25 July 2007 (UTC))


 * Good job Fvasconcellos. I would go further in the trimming as all the pharmacology section details are meaningless for combinations, i.e. Bioavailability, Metabolism, Half-life, Extretion, as currently has set up (yes each component has specific values for these parameters, but I don't think it makes any sense for the combination). Pregnancy & legal categories of the whole combination do make sense (Paracetamol/Codeine 500/8 is over-the counter, but 500/30 is prescription only in UK). Do we keep the pictures, if so, then each should have a caption parameter - and what about combinations of more than two products? I suggest dropping teh pictures entirely (there are the links to each component's article)
 * My preference, but I think unworkable, would be one drugbox but with a controlling parameter, perhaps "infobox_type" (set to "mab", "combo" or left blank), which hides unnecessary parameters. However this would make each parameter conditional (on top of current conditions of whether parameter defined or not) and all pipes (|) must be converted to templated | . This would make the template very large in its coding and difficult for all but the most experienced template editors to maintain. I'm happy to give it a go (big gulp) but please after my holiday (away after tomorrow for 11 days)...... but I would suggest far easier is to set up a restricted daughter template as discussed above and currently in place for Drugbox-mab.
 * I think "combo" in template title and "comboname" as a parameter is better than "compound" which is open to misinterpretation (amlodipine beselate is a compound, but not in sense we mean here of two separate pharmacologically active agents)
 * I've trimmed it further, see User:Davidruben/Templates/Drugbox-combo, but I like the idea of component and class of agent - clearly 2 items are obligatory for a combinatin item, and I've added ability for optional 3rd.David Ruben Talk 02:38, 25 July 2007 (UTC)
 * Excellent work, David! I would have trimmed the pharmacology section myself but thought it might be too much. As for removing the images, also a good decision although there are some images of mixture/combination products in use (see dimenhydrinate for an example).
 * Thank you so much for adding room for an optional 3rd component—I wouldn't know how to do so, and where would that leave Cortisporin? :D
 * is fine—pretty intuitive :)
 * I'm still undecided WRT a simple "component 1 and component 2" v. "component/class, component/class..." I would welcome further input on that.
 * Now, about making a single drugbox with a  or   parameter—I agree that would probably be best in the long run, but would certainly make the template even more difficult to maintain. I would certainly be no help in the process, as my knowledge of template syntax is very basic, but if you'd like (be willing ?) to take it up I would not object in the slightest. I do think that creating Drugbox-combo would be a nice, readily implementable short-term solution while we, say, prepare for and discuss future implementation of a single drugbox. Fvasconcellos (t·c) 16:20, 25 July 2007 (UTC)
 * Agree set up Drugbox-combo, can always later be changed to simply call a reved-up do-everything Drugbox passing relevant parameter values.
 * Still would need convincing re the images; the example of dimenhydrinate is a case in point, which molecule is which ?
 * Final thoughts re "component 1 and component 2" v. "component/class, component/class...".
 * 1st option inserts the word "and" between the two items, but would need some coding if to allow for 3 or more items (have to use commas except for final conjunction) - yuk. Whereas "component/class, component/class..." simply has one row per item and is much easier to code a simple conditional display/hide (indeed probably should allow for upto 4 items just in case - I'll do this now).
 * In the "component/class, component/class...", which is more important (ie which is the heading)? It could be argued that it is the class effect that is the most important and gets shown in left-column, with the actual class-member component in the right-column (eg Seretide is important for being a long-acting bronchodilator and a steroid, and really not to us doctors for having fluticasone as the steroid rather than beclometasone). Likewise "FuciBet" is Fusidic acid antibiotic and Beclometasone, but if Mometasone steroid and Neomycin antibiotic had been selected for a combo product (lets say "NeoMom"), then I doubt if of much difference (its a potent steroid with antibiotic that is being sought) . That did not work - many class-of-drug articles are very long in their wording compared to the drugs themselves (eg see Seretide example now used in my subpage's instructions), will need be Drug on left and Class on right.
 * Anyway what do others think ? I'll happily code up any finally decided decisions on my return, else feel free to move my subpage User:Davidruben/Templates/Drugbox-combo to Drugbox-combo to go live :-) David Ruben Talk 19:08, 25 July 2007 (UTC)
 * Hmm... OK then. Left would be drug and right would be class—fine by me. I'll wait a couple of days before going live, see if there is no further commentary. Fvasconcellos (t·c) 20:12, 25 July 2007 (UTC)

Single solution for consideration
Ok so I have this version for a drug-combo template, but I also have had a go at a merged single infobox to combine the current main Drugbox with options for Drug-combinations. The advantage of this single merged approach is twofold:
 * 1) having in concept just the single template (general editors need only lookup the one template) and
 * 2) the fact that current Drugbox, Drugbox-combo & Drugbox-mab templates would all have a large number of commonly shared parameters (licensing, administration routes etc) that would be awkward to maintain across the multiple templates.

At top of this thread are some points for and against such merged single infobox, so what do people think ? Below are some examples of how single template can be used for both single and combination products:

First off for Salbutamol:

And then for Fluticasone/salmeterol (Seretide)


 * If people are happy with this single approach then, question is whether to also merge in Drugbox-mab?
 * Of course before going live with this, I'd like a few others to help check the template coding is robust and "fit for purpose" as the current Drugbox coding. The Drugbox template is of course protected to prevent any disruption, and I'd hate to get this wrong :-) David Ruben Talk 00:42, 9 August 2007 (UTC)
 * Wow. That's...excellent :) Sure makes a case for WP:BOLD! I'd like to see the "Chemical data" section optional so as not to make it empty, though; is that possible? If it really works than I'd go for merging drugbox-mab as well and making the elusive unified Drugbox :) Fvasconcellos (t·c) 01:18, 9 August 2007 (UTC)
 * OK, I think your asking that the Section header of "Chemical data" becomes optional and gets hidden too (given that all its parameters are hidden), yes I'd missed out decission control on that and now so done. David Ruben Talk 15:46, 9 August 2007 (UTC)
 * P.S. I've still got some thinking to do re the merge of all chemboxes. Fvasconcellos (t·c) 01:18, 9 August 2007 (UTC)

Ok, I was feeling brave and I've now tried to include Template:Drugbox-mab, just set type to "mab" rather than combo.

Coding for this trial at User:Davidruben/Templates/Drugbox-combo. If those with the knowledge could preview the coding for bugs and give it a testing, I'd be grateful. Then if the consensus we can go live with this, and depreciate Drugbox-mab (making it redirect to the new Drugbox initially) David Ruben Talk 21:36, 10 August 2007 (UTC)
 * No-one up for reviewing the code? Fvasconcellos (t·c) 01:03, 17 August 2007 (UTC)


 * Ok, lets, make this easier - could a few non-techies (i.e. must be everyone given no one responded so far) try out a few tests on existing drug articles with User:Davidruben/Templates/Drugbox-combo (ie use the described template but use  {{Davidruben/Templates/Drugbox-combo  as the template calling name and use "Show preview" rather than "Save page"):
 * Are the instructions on use of the 3-flavours ("type" not defined, or set as either "combo" or "mab") understandable ?
 * Does this all work in the way one expects the template to work ?
 * Any glaring errors occuring ? David Ruben Talk 19:47, 17 August 2007 (UTC)
 * Now running on ampicillin/sulbactam, perindopril/indapamide and co-codamol. No problems so far. Fvasconcellos (t·c) 20:34, 17 August 2007 (UTC)
 * Hmmm, have I over coded in restricting pictures for combos? Given the example of Ampicillin/sulbactam - either the pictures are not needed in that article at all (as they duplicate those in sub-component articles), or should be allowed in the infobox (making it a long infobox and perhaps look untidy) or remain as pictures within the text & outside of this infobox ? David Ruben Talk 23:40, 17 August 2007 (UTC)
 * I quite like the images in table format in this specific case, as I put them there :) I do think the parameter could be left in, just made optional, in case an article comes up where an image might be useful, e.g. dimenhydrinate: we have no article on 8-chlorotheophyllinate (and I frankly don't see a reason for one to be created), so if one wanted to find out its structure there'd be nowhere to go on-wiki. Fvasconcellos (t·c) 23:52, 17 August 2007 (UTC)

Merge
Ok I think, if no one objects, that it is time gets upgraded and incorporates in the features of  - see this proposal. However I note that several languages link to this, so only seems polite to add some merge tag warnings and wait a day or two.

From: {{Drugbox-mab ... To: {{Drugbox | type = mab ...
 * Initially I would edit to call, setting "type=mab" in the process.
 * Then articles could have the infobox call edited:
 * Finally if we ever get to no articles calling Drugbox-mab (and there are a lot of MAB articles out there), then Drugbox-mab would be fully depreciated and converted into a redirect (which would preserve edit history unlike a full TfD) David Ruben {{sup| Talk }} 23:50, 23 August 2007 (UTC)

Been bold (well not really, there has been enough time for people to comment or review coding). Few points as I did this: Please report quickly any major problems - I'm away on holiday in couple days, and if need be we'll just revert the template back until after my return. :-) David Ruben {{sup| Talk }} 00:26, 25 August 2007 (UTC)
 * 1) I followed Fvasconcellos suggestion to leave in the pictures for combination articles, but unlikely to be commonly needed, so parameters not listed in the short version of the "type=combo" example.
 * 2) Whilst I had in my user-subpage proposed a "combo-name" parameter, the template already had an undocumented "imagename" to be shown as an optional alternative to the automatically generated  {{Pagename}}  of the calling article. "imagename" was fine if one wanted to add a manually specified image caption, but if there was no image then "imagename" was meaningless as a choice of parameter name if one sought to specify the name at the top of this infobox. "combo-name" would have been fine for "type=combo" combination articles, but not for general single drug or MAB articles. Therefore as an alternative to "imagename", I've allowed for a more generically termed "drug-name" parameter to be defined to replace the article's default  {{Pagename}} .
 * 3) Tidied up some of the internal documentation as to what is being done.


 * MAB articles still seem working, here is switch for direct use of Drugbox (vs Drugbox_mab) by Infliximab. I've switched the test examples of ampicillin/sulbactam, perindopril/indapamide and co-codamol to using the real Drugbox and added to Fluticasone/salmeterol. David Ruben {{sup| Talk }} 01:16, 25 August 2007 (UTC)
 * Excellent :) Not to be a pain, but is there any way to place articles with no images in {{cl|Drug pages needing a structure drawing}} only if they are not  or  ? Fvasconcellos (t·c) 14:52, 25 August 2007 (UTC)
 * Ok, categorised only on that proviso (works for combo, need check for mab...) David Ruben {{sup| Talk }} 15:39, 25 August 2007 (UTC)
 * Thanks, that was quick! Works for mab as well. Fvasconcellos (t·c) 17:40, 25 August 2007 (UTC)

I find these mab names still very confusing and whilst there is a link for Therapeutic monoclonal antibody, would add ing a link at the top of the infobox to Nomenclature of monoclonal antibodies be helpful? Rather than linking the article name itself which would seem strange, better (and easier to code) would be to add a linked superscripted "?", eg Xxxxxmab{{sup|?}} David Ruben {{sup| Talk }} 15:55, 25 August 2007 (UTC)
 * Hmm. Sounds good. Fvasconcellos (t·c) 17:40, 25 August 2007 (UTC)

Drugbox & Chembox
Hate to be a wet blanket but I'd like to once more suggest that drugbox, explosivebox be combined into chembox new because they all have the same basic identifiers: formula, mw, appearance, density, bp, mp, cas, pubchem, smiles, etc. I think that it is very difficult to arbitrarily draw a line and say that such and such compounds are drugs, such and such are explosives, and such and such are just industrial chemicals. Case in point is nitroglycerine which is both a drug and an explosive; other examples include simple building blocks which are listed as drugs, such as dimercaprol, hydroxyurea, and nitrous oxide.

The fact is that chembox new already contains drug and explosive parameters. So, why don't we use them and contain all data as a whole? If you wish to incorporate the combo parameter, incorporate it into chembox new instead? --Rifleman 82 03:51, 26 July 2007 (UTC)
 * Don't hate to be a wet blanket, especially since you're not being one :) I'm aware that this has been suggested before, and I do believe that eventually these templates probably should be collated into one. My idea was for drugbox-combo to be a solution for an immediate problem—the generally sorry state of infoboxes in combination drug articles, simply because they're not appropriate for them. As an aside, I honestly prefer the overall look of Drugbox to that of chembox new—of course, it wouldn't be fair of me to oppose something in the project's best interest on purely stylistic grounds. Bottom line: I do think that these articles, although not as numerous as those on monoclonal antibodies, would benefit—at least in the short run—from an infobox more suited to them. We should certainly discuss the future of these templates—perhaps somewhere more central?—but I don't really see the harm in a quick fix. Fvasconcellos (t·c) 02:25, 27 July 2007 (UTC)
 * Tend agree with Fvasconcellos, Chembox really is not appropriate in majority of drug articles, as most of its parameters fail to meet notability threshold for this a general encyclopaedia (who but an industrial pharmacist cares about most of Chembox's parameters if say applied to Ibuprofen or Trimethoprim). Whereas, most non-medicinal chemical articles are about the physical properties of the chemical and so such parameters are of course notable.
 * So compare how Ibuprofen's infobox and indeed that whole article is really mostly about clinical usage (rather than the manufacturing and biochemistry of its pharmacology - I assure you no doctors nor patients are ever interested in Ibuprofen's Acidity, Basicity, Dipole moment, Spectral data etc etc), against say Hydrochloric acid article which is very much about describing the chemical itself, its most important manufacture, associated industry activity and chemistry. Sure I agree there are some chemicals which have both medical and more widespread uses (eg Potassium permanganate or Glyceryl trinitrate), but these are generally rare exceptions (virtually all analgesics, opiates, antibiotics, drugs for hypertension etc etc have no industrial chemical uses).
 * So either, for those very few articles where clinical infobox details need to be added to a more generalised chembox, have a bloated Chembox including very rarely used drug parameters, or I would suggest preferably remove these parameters entirely from Chembox and accept that a few articles will use both Chembox and Drugbox infoboxes. David Ruben Talk 00:03, 9 August 2007 (UTC)


 * I disagree. In Chembox new and Drugbox, all parameters are optional. What should remain constant (and probably universal) is formula, mw, smiles, inchi, cas #, pubchem #. All drugs should have this data.


 * I think you misunderstand that the acidity, basicity, dipole moment, etc. are routinely filled in in chemboxes. Even when I fill in chemboxes of some chemical compounds, I use the "simple" option because such detailed information is not (easily) available from MSDS etc. What I do fill in is the parameters above, b.p., m.p. Everything you can find from an MSDS. --Rifleman 82 02:26, 9 August 2007 (UTC)


 * Ok, fair point :-) I think though that if Chembox used, then temptation would be for all the chemical data to be entered when used on pharmaceutical topics, and as stated, for most drugs such information is not notable for the wider readership. Image what arguments would ensue if parameter information were to be deleted once entered - WP is not a textbook (WP:NOT), and "merely being true or informative does not automatically make something suitable for inclusion in an encyclopedia" (WP:NOT). In contrast any unfilled parameters for industrial/commercially important substances such as sulphuric acid, iron oxide etc does seem to be of importance and notable as those articles are about the chemicals themselves, rather than drug articles which tend be more about how the chemical is used. Would be good to have additional editors views :-) David Ruben Talk 22:23, 18 August 2007 (UTC)


 * Well if there were too much info it'd be moved to a (data page) so as to reduce table creep. This mechanism already exists in chembox new. In any case, I can safely say that you won't be able to find the dipole moments, etc. for most drugs, or even most chemicals. --Rifleman 82 02:40, 19 August 2007 (UTC)


 * I'm opposed to a merger for now, but I could support one in the future. The ChemboxNew template has only been usable for a few months, and some of the functionality is still stabilizing. And we're still figuring out what to do with Drugbox -- in many cases a "drug" is a molecule, but it can also be an antibody, a mixture, a polymer, or an unrefined plant product. It's going to take some creativity and time to figure out the best ways to handle all the possibilities (and I commend those who are working on that issue as we speak). Once the boxes have stabilized, we would be able to make a more informed decision on how they could coexist. --Arcadian 18:49, 20 August 2007 (UTC)


 * I clearly vote for a merge. This is required for avoiding a lot of work in the future by editing hundreds of articles. Furthermore is it easier to maintain one centralized solution then having several drug, chem, explosive, and whatever boxes. If people can not agree, I vote for a general chembox intersecting all boxes out there. If there is then still a need for additional information I would prefer an add-on or a plug-in schema for the boxes. JKW 22:16, 12 September 2007 (UTC)