Wikipedia talk:WikiProject Medicine/Archive 130

See Wikipedia:Manual of Style/Medicine-related articles#Product pricing. For current discussion see Wikipedia talk:Manual of Style/Medicine-related articles#Current wording. QuackGuru (talk) 14:07, 21 November 2019 (UTC)

commented(product pricing[1])--Ozzie10aaaa (talk) 20:15, 21 November 2019 (UTC)

The discussion on that page is ongoing but I have discovered specific issues with the widely used source: International Medical Products Price Guide. This is cited as a reference for wholesale prices in the developing world. The problem is we are misusing it, as what our articles claim is just plain incorrect. See Wikipedia talk:Manual of Style/Medicine-related articles#International Medical Products Price Guide. There are similar problems with our use of (retail) prices in the US from Drugs.com. Never mind your views on whether Wikipedia should offer price information, what our articles are currently doing is wrong on so many levels and for so many years that it is rather embarrassing to WP:MED. Please discuss on the guideline talk page. -- Colin°^Talk 18:31, 25 November 2019 (UTC)

Thank you. I've been rather shocked at all this. --Ronz (talk) 18:43, 25 November 2019 (UTC)

This characterization is incorrect. The source in question gives the Defined Daily Dose and prices for LMIC.[2] For sodium valproate the price is about US$0.40 as of 2015 per day. We also have these values published in a book co published by the World Health Organization and MSH.[3]

We have NADAC prices from the US government for many medications. And we have pharmaceutical textbooks that list the appropriate consumer price.[4]

Sure we can do better in the area of health economics. Just because we do not do a perfect job and are not able to also update immediately does not mean we should do nothing in this area. Doc James (talk · contribs · email) 19:26, 25 November 2019 (UTC)

We're only going to do better if we follow policy: content and behavioral.

The characterization that editors may be seeing perfection, or want nothing is incorrect and misleading. --Ronz (talk) 19:48, 25 November 2019 (UTC)

And what is incorrect about "Defined Daily Dose and prices for LMIC"? It is based on a World Health Organization published source.

This is the price from the IDA Foundation that is being reported. And they are a wholesaler that sells to LMIC. Doc James (talk · contribs · email) 20:22, 25 November 2019 (UTC)

Working so hard to make a case looks like desperation and WP:BATTLE. It's pretty clear at this point that there's no simple, general solution. --Ronz (talk) 21:35, 25 November 2019 (UTC)

Meh. Nothing wrong with working hard to write encyclopedic medical content. Doc James (talk · contribs · email) 00:26, 26 November 2019 (UTC)

Agree with Ronz, I also think this is not simple. I read both sides arguments, and I think there are merits for both. I think everyone agrees that a perspective on prices (eg, how price evolved over time, variation across countries, etc) is often/always admissible given reliable sources. But I think the point of contention is whether or not drug prices, without any perspective, should be admissible per se, by considering them as facts. Regardless of the debate about whether industries try to prevent this info out, it's difficult for me to say for sure whether this has encyclopedic value. But there are indeed cases where this is accepted, for example age is considered factual and does not require any perspective to be admissible on biographies, or the chemical composition of compounds. Should price be considered a fact in a similar sense? At this point, I cannot decide. But I think it can merit a new RfC to foster discussion on this. --Signimu (talk) 23:13, 25 November 2019 (UTC)

Signimu, the problem with an RFC right now, is that don't actually have something factual, concrete to ask people's opinion about. Earlier people were asked "Do you want prices" and Wikipedia said no and WP:MED ignored that. Now, I'm demonstrating that "Do you want prices" isn't even a sensible question to ask. James hasn't quite got there yet. -- Colin°^Talk 16:10, 26 November 2019 (UTC)

IMO, Doc James makes a better case--Ozzie10aaaa (talk) 02:15, 26 November 2019 (UTC)

We could use these sorts of sources.[5] But they are not as uptodate. Prices there come from the "Red Book: Pharmacy’s Fundamental Reference" which we could use directly and are up to date. But it requires paying for access.

We can use this source from MSF for discussing the prices of HIV meds.[6]

By the way LMIC represent 85% of the global population.[7] Doc James (talk · contribs · email) 03:01, 26 November 2019 (UTC)

James, "The price of what?". This is your problem with all your sources. They give the price of something with a bar code. They have separate entries for each of those things, but you have randomly picked just one. A particular mg size of tablets in a particular size of bottle made by a particular company. There is no way to relate any individual one of those "what"s to a drug article. The drugs are used differently depending on condition, circumstance, weight, age group, renal function, etc, etc. I think you have forgotten that a drug is just a chemical that we can measure in milligrams. But nobody sells milligrams of drugs from pharmacies, they sell specific tablets or syrups. And there is no one dose. James is misusing Defined Daily Dose in a way that the WHO explicitly disallow. There is a reason why none of our sources do what James did: it is just nonsense. The reason why we have WP:WEIGHT and WP:NOR is to stop this kind of nonsense. -- Colin°^Talk 16:06, 26 November 2019 (UTC)

Lets look at the definition of defined daily dose "The DDD is the assumed average maintenance dose per day for a drug used for its main indication in adults."[8]

The DDD is listed here as 1.5 G for sodium valproate. Is that a reasonable dose for adults? Well lets look at a monograph, yup falls in the range of used in adults.[9]

Source gives the price for 500 mg tablets and a price per tablet of 0.1339 as of 2015. The IDA Foundation ships to more than 130 LMIC at this price.[10]

Is there a better representation of the cost of this medication you wish to see Colin? Or is your opinion that it is impossible to represent pricing information for medications? Doc James (talk · contribs · email) 16:38, 26 November 2019 (UTC)

Per "There is a reason why none of our sources do what James did" that is easy to prove false. Here is a source that gives estimates of prices for medications.[11] The price of a month of valproic acid is roughly US$50 to 100 per month in the United States as of 2019. With a dose of 1.5 gm per day (180 * 250 mg tabs per month) this arrives at the same price. 72 USD per month.[12] This also gives prices in this range for the US.[13] Doc James (talk · contribs · email) 16:42, 26 November 2019 (UTC)

Doc James, you link several sources. The first I can't read all of but seems to have some code with the number of $ signs perhaps indicating cost like TripAdvisor restaurants. The Drugs.com link does not give the price information like you did. It gives individual prices for individual products just like I said: it gives one lot of prices for the 250mg capsule x 100 and another lot for the 250mg/5ml syrup, with six prices for six different bottle sizes. The last link is a lay medical website that doesn't meet MEDRS. It has articles like "Don’t Like Vegetables? It May Be Your Genes" and "Skipping Breakfast Before Your Workout Could Help You Burn Fat", and wrongly claims Depokete is valproic acid (it is divalproex sodium) But hold on a second. The DDD of valproic acid is 1.5g per day. And Healthline claim "following prices are the average cost of a one-month supply for each drug". But they quote prices for "ninety 250-mg tablets of Depekene", "ninety 500-mg tablets of Depakote" and "sixty 500-mg tablets of Depakote ER". These are all different doses. -- Colin°^Talk 17:22, 26 November 2019 (UTC)

Colin — You really seem clueless here. You're accusing Doc James of arbitrarily choosing his statistics — when he is advocating the use of the WHO-supported defined daily dose. To me this looks like a tempest in a tea-pot, brewed by someone who is in far too deep water. Carl Fredrik ^talk 18:01, 26 November 2019 (UTC)

CFCF please look at the Wikipedia talk:Manual of Style/Medicine-related articles and search for "Introduction to Drug Utilization Research 2.6 Drug costs". You will find that the WHO themselves explicitly warn not to use the DDD for this kind of drug pricing. For example, WHO says "For example, the cost per DDD can usually be used to compare the costs of two formulations of the same drug. However, it is usually inappropriate to use this metric to compare the costs of different drugs or drug groups as the relationship between DDD and PDD may vary". So, giving a cost per DDD of Carbamazepine in one article and a cost per DDD of Lamotrigine in the lead of its article, is just the sort of thing we shouldn't do. We can use it to say that the 150mg pack of one drug works out more expensive to meet the DDD than a 500mg pack. Or we can use it for drug usage research, which is what it is for. The problem, CFCF, is that you and James seem keen to do an argument-to-authority by saying "my source is WHO" or "this is a WHO metric" but you don't actually stop to work out what the source said or what the WHO metric is for. Carbamazepine is used for epilepsy, neuropathic pain, schizophrenia and bipolar disorder. Which of those many indications do you think the DDD is for? Do you think the same dose is used for them all? -- Colin°^Talk 18:19, 26 November 2019 (UTC)

Lets bring in the entire quote "Total drug costs; cost per prescription; cost per treatment day, month or year; cost per defined daily dose (DDD); cost per prescribed daily dose (PDD); cost as a proportion of gross national product; cost as a proportion of total health costs; cost as a proportion of average income; net cost per health outcome (cost-effectiveness ratio); net cost per quality adjusted life-year (cost-utility-ratio) Data on drug costs will always be important in managing policy related to drug supply, pricing and use. Numerous cost metrics can be used and some of these are shown in the box above. For example, the cost per DDD can usually be used to compare the costs of two formulations of the same drug. However, it is usually inappropriate to use this metric to compare the costs of different drugs or drug groups as the relationship between DDD and PDD may vary."

We are not using the DDD to "compare the costs of different drugs" but we are simple providing the "cost per defined daily dose (DDD)" which is recognized by WHO as on2 cost estimate. Should we include other cost estimates? Yes certainly. Doc James (talk · contribs · email) 18:26, 26 November 2019 (UTC)

You mentioned carbamazepine dosing for different indications. Let looks at this for adults:[14]

Epilepsy 400 to 1600 mg per day Trigeminal neuralgia 200 mg to 1200 mg Mania 400 mg to 1600 mg per day Bipolar 400 mg to 1600 mg per day. The DDD is for the most common indication and in adults. The typical doses for all listed indication are very similar. Doc James (talk · contribs · email) 18:32, 26 November 2019 (UTC)

And look the DDD is 1 gram. Seem perfectly reasonable.[15] Doc James (talk · contribs · email) 18:52, 26 November 2019 (UTC)

James you are conducting original research. If you'd read carefully, you'd know that DDD is the based on a maintenance dose in adults, not the initial/maximum doses you quote above. It is this sort of carelessness with figures that has got us into this mess. Please read Problems using the defined daily dose (DDD) as statistical basis for drug pricing and reimbursement" See the bit about "drug pricing" in the title. They explain "The defined daily dose is an artificially and arbitrarily created statistical measurement used for research purposes in comparing the utilization of drugs.", "in most cases, the DDD differs greatly from the typical PDD of the drug in question. In some cases, this gap may be exacerbated by the fact that a drug may be prescribed in two vastly different dosages and the DDD represents the average of those outliers." they also say "For most drugs there is no single dose but rather multiple variations prescribed depending on the severity of the patient’s condition or the level of tolerance. Moreover, prescribing practice and medical customs differ from country to country. As a result, an international compromise must be made that often has little to no relevance to actual prescribing practice from one country to another." The WHO says "Basing reimbursement, therapeutic reference pricing and other decisions on ATC/DDD classifications is a misuse of the system. Defined Daily Doses are not designed necessarily to reflect therapeutically equivalent doses." What you are trying to do in the lead is give a therapeutic reference price. Don't. -- Colin°^Talk 19:38, 26 November 2019 (UTC)

This is not "therapeutic reference price". This is what that term means "Therapeutic reference pricing (TRP) places medicines to treat the same medical condition into groups or 'clusters' with a single common reimbursed price."[16]

We already know that the pharmaceutical industry do not want easily and publicly avaliable pricing of medications. That one of their industry groups that you quote International Federation of Pharmaceutical Manufacturers & Associations also holds that position is not surprising.

That you are pushing the industry position to try to WP:CENSOR Wikipedia is concerning. Doc James (talk · contribs · email) 19:57, 26 November 2019 (UTC)

Sorry, James, you just win the Wiki-Godwins-Law prize for using WP:CENSOR in an argument for including something in an article. You asked earlier for civility: you are just acusing me of being an industry shill. You forget James, I'm having a discussion and have not edited the price any one of those articles. I don't have an agenda, unlike you, who have declared yours openly. Ok, I accept I got TRP wrong. However, the text about misusing DDD comes from WHO, not any pharama body. You can read how it is calculated here and how it should be used here. Please read it carefully. We are talking about price here. It wasn't intended for that purpose. It is a very complex technical measurement and its subtlies will not be apparent to our readers who read about two different drugs at two different prices per "dose" and not realise they can't be compared. You are trying to use this research tool metric in a general-reader encylopaedia, and as I keep saying, there is a reason why nobody else does this. -- Colin°^Talk 20:57, 26 November 2019 (UTC)

From the 1st link given by Colin[17]: "Drug utilization data presented in DDDs only give a rough estimate of consumption and not an exact picture of actual use. DDDs provide a fixed unit of measurement independent of price, currencies, package size and strength enabling the researcher to assess trends in drug consumption and to perform comparisons between population groups." After reading both links, I think ATC/DDD have encyclopedic value IMO, being standard approximate measures of consumption, just like we consider incidence and prevalence as encyclopedic infos. And we'd need articles to clearly describe what these measures are, since they can be misunderstood by the general public, as they are more targeted towards researchers (didn't check, maybe they already exist). But indeed, the WHO clearly defines DDD as a measure independent of price, so if that's the goal, it seems like we would need another measure for that. Signimu (talk) 23:53, 26 November 2019 (UTC)

What about the median price ratio mpr and international reference price? See page 12 of [18] Signimu (talk) 08:43, 27 November 2019 (UTC)

Agree User:Signimu and what we are using is an accepted international reference price :-) Doc James (talk · contribs · email) 19:00, 28 November 2019 (UTC)

Above CFCF claims I'm accusing James of arbitrarily choosing his statistics. And further that James uses the WHO-supported defined daily dose. Let's see if that equates with reality:

• Carbamazepine. "between 0.01 and US$0.07 per dose". The source has a low of $0.00121 per tablet and a high of $0.0683 per tablet. The tablet is 200mg and the DDD is 1000mg. The article seems to be using the 200mg tablet as a "dose". However, if you search for Carbamazepine in 2014 you get three database entries: one 100mg/5ml syrup, one 200mg tablet and one 200mg sustained release tablet.
• Lamotrigine. "about 3.57 USD per month". The source has a median price of $0.1178 per 100mg tablet. This works out at $3.58 per month for one 100mg tablet a day. But the DDD is 300mg. If you search for Lamotrigine 2015 you get three database entries: one 25mg tablet, one 50mg tablet and one 100mg tablet.
• Lorazepam. "typical dose by mouth is between US$0.02 and US$0.16". The source has a low of $0.0242 per 1mg tablet and high of $0.16. This matches the article. But the DDD is 2.5mg. If you search for Lorazepam 2014 you get two entries: one 1mg table and one 2mg tablet.
• Ethosuximide. "about US$27.77 per month" The source has a sole price of 0.1845 per 250mg tablet. The DD is 1250mg, which is 5 tablets. This is $28.05 per month using the DDD which is close to the article figure. There is only one record for Ethosuximide and only one supplier, ASRAMES, which only supplies the Democratic Republic of Congo. The claim this is "The wholesale cost in the developing world" is here based on one supplier to one African country, and we have the nerve to quote it to four significant figures.

So there is no consistency in which definition of "dose" James has used -- sometimes the tablet and sometimes the DDD. And he has each time arbitrarily chosen one of many possible tablet sizes. For example, Lamotrigine is given as a price for a month but based on only 1 tablet a day, at a dose three times less than the DDD. So yes, James is arbitrarily choosing his statistics and doesn't seem to have any consistent day-to-day concept of a "dose". Which isn't surprising since there isn't a one "dose" for any medicine. Just lookup the BNF for a drug and you will find a couple of pages of dosage advise. James has tried to simplify a complex concept into just one number. It doesn't work. -- Colin°^Talk 18:41, 26 November 2019 (UTC)

Yes for carbamazapine I used the cost of the least expensive table, I agree switching to price per day of DDD would be good and have done so.

If we were consistently use price per day of DDD (for long term medications) would you satisfied?

Doc James (talk · contribs · email) 18:52, 26 November 2019 (UTC)

Ah for Ethosuximide I used 30 days for a month rather than 31 days. Okay...

This this data was added gradually over many years. Happy to discuss moving to consistency as much as possible. But I am not sure that is what you are looking for... Doc James (talk · contribs · email) 18:59, 26 November 2019 (UTC)

Firstly I'm looking for you to accept that most of the existing prices are based on the price for one arbitrarily chosen pill size. You had multiple options and seemed to pick one at random. You claim you picked the least expensive tablet. So you could have written that instead of "dose", which is wrong. This is a basic mistake and it would be good for you to admit you made a mistake in the lead section of several hundred articles. No, DDD is not a valid metric for this and WHO says that. I asked you to find sources that say things like "The wholesale cost in the developing world is about US$0.07 to US$0.24 per day as of 2015" in their introductions, or indeed, anywhere. You still haven't found any. Find me a source that consistently uses the cost per DDD in its methodology for presenting drug prices for all the hundreds of drugs that source lists -- and only one price per drug please, just like we do. Find me a source that gives the price of "a drug" rather than a specific size of pill, bottle size and brand. They don't exist and we should follow the sources.

You are still claiming that IMPPG is a reference for the min and max prices for drugs in the developing world. It isn't, as I explained at length on the other page. And you are still juxtaposing wholesale prices in the developing world with retail prices in the US (but without saying the US price is retail). And Ethosuximide still falsely claims the price in the developing world is "about US$27.77 per month". James, I am asking for basic maths and basic source->article honesty. Your source gives the price in the DRC and that is all. That is not "the price in the developing world". And Ethosuximide is far from the only article that stretches "one or two suppliers" or "the contract price in one or two nations" to become "the price in the developing world". -- Colin°^Talk 19:16, 26 November 2019 (UTC)

No not arbitrarily chosen...

This source lists price of HIV regimens per person per year.[19] And goes on to say "Prices are rounded up to the third decimal for unit price and to the nearest whole number for price per person per year (PPPY). The annual cost of treatment PPPY has been calculated according to the WHO dosing schedules, multiplying the unit price (one tablet, capsule or millilitre) by the number of units required for the daily dose, and by 365." Doc James (talk · contribs · email) 20:06, 26 November 2019 (UTC)

James, this still isn't what I asked for. The article does not say they use WHO Defined Daily Dose. They have doses for adults and children separately. Of the retroviral drugs I have looked at, they all have trivial dosage regimes. Mostly 1 tablet a day for any adult. This isn't something where you start off on 5mg and we see if that isn't enough to prevent you dying. It is pretty much a fixed dose and cross your fingers and hope not to die. The article you link has a handful of drug sizes x a small number of suppliers giving up to a dozen prices per drug. The linked article doesn't claim this handful of suppliers represents the most and least expensive suppliers in the entire developing world, as you have done hundreds of times in our articles. There is a reason why our reliable sources give pricing in such a complex matrix of tables and footnotes and *exceptions. It is more complicated than you want it to be. -- Colin°^Talk 20:57, 26 November 2019 (UTC)

Doc James I think the approach where I list everything that is wrong with what is in our article text and then you adjust/argue just one point alone isn't working. So let's take this one step at a time

1. The source is not a reference for all prices in the developing world. Therefore a statement that "The wholesale cost in the developing world is between US$0.02 and US$0.16" cannot be derived from it. This is because what you wrote claims a minimum and maximum price "in the developing world". We sometimes have a single price, a few prices or a handful of prices. So do you want to give a price range or just a single price? If a price range, we're going to have to describe the source of values better. Something like "The International Medical Products Price Guide gives prices in the developing world for a 1mg tablet that range from XX to YY". If you want just a single price, then how would you calculate it? Btw, if you are thinking that this is getting a bit wordy, let's consider that we really should be writing article body text first, so let's worry about that for now. -- Colin°^Talk 08:55, 27 November 2019 (UTC)

Lets look at the text in question "The wholesale cost in the developing world of a typical dose by mouth is between US$0.02 and US$0.16 as of 2014."

So yes it provides a range. But a range does not necessarily mean an absolute maximum and an absolute minimum. Frequently ranges in health care are 1) one standard deviation ie the confidence intervals 2) the highest and lowest value in a set. What we have here is the second one. Doc James (talk · contribs · email) 14:55, 27 November 2019 (UTC)

I think Colin has a valid point about attribution here, James. If I understand the argument, then the suggestion would be that we should be writing a section of the article that contains reliable sourced pricing information, which is attributed to the source, rather than asserted as a simple fact.

I can see that something like "The <ABC price guide, published by the WXY,> gave prices for <this drug> in <whatever year> for <a particular region> of <this many dollars/cents (or range)> per <tablet/dose/etc.> ref:whatever" would be an obviously neutral formulation. By utilising the formulations imposed by the source, we avoid any hint of analysis on our part. In the article text, there is less need to be concise, and we could include more than one source, allowing the reader to form their own conclusions. Unless, of course, we also have reliable sources specifically commenting on particular pricing for particular drugs (perhaps as we might find for insulin?), when we could expand the section with a summary of that commentary. What do you think? --RexxS (talk) 15:22, 27 November 2019 (UTC)

If we parse the text "The wholesale cost in the developing world of a typical dose by mouth is US$0.02 and US$0.16 as of 2014."

• 1 mg is a typical oral dose. This is common knowledge but sure we could state 1 mg.
• "wholesale cost in the developing world" does not mean absolutely every single possible country simple that these represent some of the wholesale prices in the developing world, which they do

Per adding "ABC price guide, published by the WXY". Imagine if we were required to start every sentence on Wikipedia by stating the journal that published it. The year of publication. The authors who published it. The institution the authors come from. All the caveats for the data in question including the patient population that the data comes from such as their country or origin, ethnicity, sex breakdown, age breakdown etc.

We do not do this but rather we attempt to provide a reasonable summary and people can look at our sources for more details. Pricing data is actually no more complicated than any other statement we make on Wikipedia.

I would be satisfied with the rest of the suggestion "The prices in <whatever year> for <a particular region> of <this many dollars/cents (or range)> per <tablet/dose/DDD etc.> ref:whatever" Doc James (talk · contribs · email) 15:58, 27 November 2019 (UTC)

For example we summarize "Male circumcision reduces the risk of HIV infection among heterosexual men in sub-Saharan Africa." based on this Cochrane review[20]

We do not say "A 2009 Cochrane review by Siegfried et al of three RCTs of 2,274, 4,996, and 2,784 men from the general population in South Africa, Uganda, and Kenya carried out between 2002 and 2006 at 12 months and 21 or 24 months resulted in an incidence risk ration of...."

The second of course is also entirely correct. And one could argue that each bit is critically important, our goal is to provide a reasonable summary and give our readers the ability to dive deeper.

Doc James (talk · contribs · email) 16:14, 27 November 2019 (UTC)

At present, we've got a content dispute about whether it's accurate to state without qualification a fact about pricing. We both know that in some Wikipedia articles, we have to start almost every sentence by attributing the statement made because of controversy surrounding the subject. In some sections, we actually do start almost every sentence "According to <this review> ...". Of course that's not the same as "stating the journal that published it. The year of publication. The authors who published it. The institution the authors come from" and so on. My suggestion to attribute was because that might represent a compromise that moved the discussion forward. I do agree, though, that if the "ABC price guide" was well known, there wouldn't be any need to qualify it with the publisher. --RexxS (talk) 18:02, 27 November 2019 (UTC)

Well-known, or a standard, just like WHO ICD, we don't repeat in infoboxes that "this disease is classified as xxx by the WHO in ICD11", we just properly wikilink and ref it. I could see something similar with drug prices, if there is a standard measure. Signimu (talk) 19:15, 27 November 2019 (UTC)

James, et al, let's not get side-tracked. (The claim that "1mg a typical dose is common knowledge" is, well, bollocks). If you keep raising other wording issues then we won't resolve the one in this step alone. James, I agree the range is a set: and the set you claim it is is the set of prices in the developing world. You and I both know that if you look at the source most of the drugs, for one size of pill, have a very very limited set of data. Many times there are no supplier prices at all, which are the gold standard prices in the guide because it is a real price on offer -- so it only has buyer prices, which it places big warnings around because those prices aren't available to anyone else, and may reflect peculiar negotiations, bribery and corruption, who knows.

Wrt attribution, consider the polls being conducted prior to the UK election in three weeks. We don't say "43 percent of the UK population intend to vote Conservative", because the polls don't actually ask the whole UK population. Instead we say "The YouGov poll published on 26 November put the Conservatives on 43%..." So people know it is a poll (typically of a few thousand people) by a certain organisation on a certain date. We do not have facts about prices in the whole developing world, so we can't make that claim to our readers.

James, if you took the blood pressure of the next three patients that walked into your hospital, or counted the number of pages in 5 random books in your bookshelf, you can't claim "the blood pressure in humans is between X and Y" or "novels in English literature are between 100 and 450 pages long". The source itself makes no claim to be comprehensive. Its purpose is to list a handful of indicative prices to help purchases make good deals. They have no need to list 100 prices if 10 will do, and no need to seek out outliers. Even your claim of "developing nations" isn't actually mentioned in the guide at all. It just so happens that the buyers and suppliers seem to be located there.

If you want to give a range, you need to define the set. And in this case it is the set of prices for one mg size of pill in the International Medical Products Price Guide. If however, you just wish to give a single indicative price, then we might be able to come up with a wording that avoids having to mention the guide. If we are happy to accept this guide provides indicative prices for the developing world, then perhaps we can use that sort of language. -- Colin°^Talk 19:35, 27 November 2019 (UTC)

"We do not have facts about prices in the whole developing world, so we can't make that claim to our readers." That would be true if we were generating the content ourselves. But we're not. We're summarising what has been published in reliable sources. If a reliable source states that the price of a given drug is between $X and $Y in the developing world, we can most certainly report that as the price range for that drug in the developing world. We rely on the sources for our claims, not our own analysis. --RexxS (talk) 20:24, 27 November 2019 (UTC)

"If a reliable source states that the price of a given drug is between $X and $Y in the developing world, we can". RexxS, that's exactly the problem. No reliable source states that the price of a given drug is between $X and $Y in the developing world. We not only have a source->article disconnect, but a big WP:WEIGHT problem, but that can wait for later discussion point. Please have a look at the source we are discussing, it is very much raw data. To use my polling analogy, it is a bit like seeing the raw results from the YouGov poll with "John1234: Green, Susan0458: Conservative, Edmond0405: Labour...." and trying to make a point about who is going to win the election ourselves. Ok, I exaggerate a little with the polling, but there's a lot of WP:OR and hand waving going on. The Ethosuximide claim is a bit like looking at the data for the Cambridge South constituency, finding they only polled Greg3055 who is keen on the Brexit party, and declaring that Cambridge South is pro-Brexit. We often only have one or a few datapoints. -- Colin°^Talk 16:16, 28 November 2019 (UTC)

The WHO supports the use of this source as a international reference price.[21] Doc James (talk · contribs · email) 18:50, 28 November 2019 (UTC)

agree w/ Doc James, and the WHO support speaks for itself--Ozzie10aaaa (talk) 19:12, 28 November 2019 (UTC)

James / Ozzie10aaaa "supports the use of this source" doesn't really solve the problem. It is a good source, for certain things. But you have to use it correctly and you have not been doing that. If you care to read the "Measuring Medicine Prices, Availability, Affordability and Price Components" book, page 41:

"Median prices listed in MSH’s International Drug Price Indicator Guide have been selected as the most useful standard since they are updated frequently, are always available and are relatively stable. These prices are recent procurement prices offered by both not-for-profit and for-profit suppliers to developing countries for multi-source products. When no supplier prices are available, buyer prices are used, but a single supplier price is still preferable to multiple buyer prices".

Nowhere have you found a source that says the price of a given drug is between $X and $Y in the developing world. So, can we draw a line under that and all agree that using the MSH to claim min/max prices for the developing world is not supported or recommended by sources. Anyone with statistics knowledge will tell you that min/max prices are of limited value as they always tend to give outlier values rather than representative values. So the prices manual recommend you pick the median of the supplier prices, whereas you have been sampling min/max from both sets of prices, which, while not exactly apples and oranges, are certainly oranges and satsumas.

Can we then follow their advice for a single reference MSH price: the median of the suppliers or the median of the buyers if no suppliers. If we agree on that, we can come to the next step. -- Colin°^Talk 14:38, 29 November 2019 (UTC)

Yes we all agree that these are not min/max prices. No one has every claimed that they are as far as I have seen. Just like when one states confidence intervals one is not stating min and max of possible values. We hae already been over this. Not all ranges are min max. Doc James (talk · contribs · email) 16:29, 29 November 2019 (UTC)

James, your statements are very unclear. It seems we don't agree on min/max. Every drug article you have edited with a range from/to is a claim of a min/max in the developing world. If you won't accept this, then we'll need to go get a third opinion from someone with authority in this area. I could do a post on the VP if you want. There is a reason nobody is presenting price information like you do. It would be simpler if you would just agree to stop using "from X to Y" style of presenting the prices, and agree with the text in bold as the method to pick. -- Colin°^Talk 18:30, 29 November 2019 (UTC)

2. The source only gives price values for a particular dose strength, pack size and formulation per database entry. James only links to one entry, which appears to be chosen quite arbitrarily. Again look at this manual. Assuming we use their algorithm for price it comes back, as always, to "the price of what?" So far in our articles, we have not always said, or have said "dose" without defining that. Mostly the existing articles use "dose" to mean "one pill" of a completely arbitrary and unspecified strength. So the existing practice isn't telling our readers what the price is for. The manual lists several attributes a price is for:

• Generic INN name (the drug)
• Dosage form -- cap/tab, millilitre, gram, dose (e.g. inhaler puff), modified release tablet, pessary/suppository.
• Strength -- usually milligrams per dosage form, but sometimes e.g. per 5ml where the dose is per 1 ml.
• Pack size e.g. 100 tablets or 600ml bottle.
• Brand or low-price generic (I'm guessing we'd concentrate on the generic if available).

The people using these reference prices and price catalogues are doing so because they want to buy a bulk order of bottles of 100 x 2mg tablets of generic lorazipam. They will do a separate price for 1mg tablets. And another for vials for injection, etc. So how do we pick which database entry or entries to cite and how do we express the "price of what" to our readers. All our sources use language like "price for 100 tablets of 100mg wonderpam" or "28 tablets 100mg wonderpam". Example: BNF has over two dozen prices of lorazepam. Drugs.com has three injections doses and three tablet doses, each with different prices depending on how much you order. If you think pack size doesn't matter much, the 1mg tablet costs $9.83 each for one, but $0.65 each for 30, $0.34 each for 1000. Remember, I'm thinking of article body here, so we have room to explain and list more than one if required. -- Colin°^Talk 14:53, 29 November 2019 (UTC)

I have already stated that these are not chosen arbitrarily. The lowest price formulation to make the DDD was generally chosen.

Sure I am happy to go through and clarify the exact dose or make it clear that the price is per DDD when this has not been done.

It is tough for me to see the BNF website itself as I am outside the UK and thus generally use a specific print version. Looking at BNF 76 I see two prices for oral lorazepam. One for 1 mg tabs, one for 2.5 mg tabs. Doc James (talk · contribs · email) 16:33, 29 November 2019 (UTC)

James, you didn't say "lowest price formulation to make the DDD" before, you mentioned you chose the cheapest tablet for carbamazepine but you didn't even consider DDD until recently. The cheapest to make the DDD isn't necessarily a reasonable choice. Take Diazepam for example. You quote the price of 10mg, and the DDD is 10mg, but the source has no suppliers for 10mg, only two buyers. Compare instead the source for 5mg has 8 suppliers and 5 buyers. It is clear that 5mg is more popular and indeed the practice for most uses is to take as divided doses; divided doses are common. So really, if one wants to pick just one size, you need to pick the one that is commonly chosen for whatever condition the DDD is for. James, when you pick a dosage by some hidden choice that is not actually an official one, that is "arbitrary". It is also original research.

We need to include the same information our sources do. No you can't use DDD. You need to do what our sources do, James. They don't pick just one pill size and ignore the others. They don't conduct OR and present the prices in DDD. They don't pick the cheapest supplier price and the dearest buyer price. You can read the Drugs.com one. How many prices on that page James? There is a reason nobody is presenting price information over-simplified like you do. -- Colin°^Talk 18:30, 29 November 2019 (UTC)

Part of the core of this issue is one of knowledge equity. It is by far easier to write about content that is of "importance" to people in the United States and Europe. The fact that the list price of onasemnogene abeparvovec is more than $US2.1 million is easy to support using the popular press.[22][23] This medication is of course not avaliable in LMIC. I am not seeing anyone (without a connection to industry) seriously arguing that we should remove pricing information from this article.

Writing articles about medications that are approved in LMIC but not approved in the US or EU is much more difficult but in my opinion is equally important. Using sources from the World Health Organization is perfectly appropriate to discuss issues in LMIC. And no it is not surprising that the Western popular press frequently does not cover issues of importance in LMIC. Does not mean we shouldn't either, of course not. Doc James (talk · contribs · email) 16:44, 27 November 2019 (UTC)

• I think adding the price of essential medicines is responsible because it's an important information for readers and with regard to the developing countries, I think adding the range of prices is helpful and we can use the formal resources in that countries. For example the website for Jordanian food and drug administration provides details about the prices of medicines. Regards---Avicenno (talk) 20:09, 27 November 2019 (UTC)

• I can't easily read the site you link, but it follows the examples I see elsewhere. Prices are quoted for things with barcodes:

• Ampidar Forte Caps 500mg / 500 capsules / 500mg / Manufacturer (or Supplier?)-name / 28.33
• Ampidar Forte Caps 500mg / 16 capsules / 500mg / Manufacturer (or Supplier?)-name / 1.4
• Ampidar Forte Caps 500mg / 20 / 500mg / Manufacturer (or Supplier?)-name / 1.75
• Ampidar Forte G.F.Susp / 100ml/ 250mg/5ml / Manufacturer (or Supplier?)-name / 11.37

None of our sources then takes this raw data and says "The price in developing countries for ampicillin is between US$0.13 and 1.20 for a vial of the intravenous solution". None of them forget to mention the vial is 1g of the drug. All of them list many options. -- Colin°^Talk 18:46, 29 November 2019 (UTC)

Agree we shouldn't make a ratio over the ddd unless we've got sources who did the same, otherwise we are making a new ratio that although it may be very pertinent, we don't know what it measures exactly and hiw robust it is. I think we should stick wit what we have: ERP for a reference price for a specific formulation (the choice of which to show depending on the article's context: if eg injection is more pertinent then show pricing for that, otherwise may be sensible to choose the formulation where more data is available), atc/ddd for average dosage consumption and to reference the atc code of a family of drugs, and mpr for comparisons. All of these are well defined and widely used for their respective purposes, with clear methodologies so we don't have to infer anything ourselves. Btw for ranges, the hai provides 25% aed 75% interquartile prices for some medications in some countries. Signimu (talk) 22:35, 29 November 2019 (UTC)

There is also the affordability measure used by who/hai, which is defined as the numbers of wage days someone needs to pay to get one month of a medication, so it'svvery accessible to the lay public as this is the purpose, but i need to check if and where the database is. Signimu (talk) 23:10, 29 November 2019 (UTC)

Correction: the cost per DDD is a real metric and is used by some European policy makers, although this is a misuse as stated by the who themselves. But comparing drugs of similar atc class (level 4 or even 3) is ok and named internal reference pricing, but this generally can't be used to compare prices internationally, only nationally. Finally, cost is most often expressed either as standard unit (tablet, pill, etc any one dose) or per gram, and there's a great review about their differences [24]. Both can be used internationally as does the review. Signimu (talk) 22:22, 30 November 2019 (UTC)

• Comment: It seems to me that prices of medications are a important fact about medications. Flyer22 Reborn (talk) 21:49, 29 November 2019 (UTC)
  • Flyer22 Reborn, see Wikipedia talk:WikiProject Medicine/Archive 84#Price of medications. There was an RFC, started by James, to get approval for his practice of routinely adding prices to medicine articles. The community said no. Further, nobody has found a source we can use to present the prices in the way James wants to. And the source he is pushing, would only permit the sort of very detailed per-barcoded-item pricing that every other side (Drugs.com, BNF, JFDA, etc) does. Which, fails WP:NOT. The price of lots of things seems important to some, but it isn't necessarily Wikipedia's job. There are just too many "prices of medications" per drug, and no reasonable way for us to summarise that if our sources do not. The world has moved on from when Wikipedia was a go-to resource for medical information. It very much isn't any more, and Wikipedia should concentrate on what it delivers best, and leave others to maintain huge databases on the prices of drugs. -- Colin°^Talk 13:26, 30 November 2019 (UTC)

Colin, knowing how popular Wikipedia is in the search engines and that so many use Wikipedia as a medical resource, I don't agree that "the world has moved on from when Wikipedia was a go-to resource for medical information." What resource do you think is now the go-to resource? No need to ping me, by the way, since this page is on my watchlist. Flyer22 Reborn (talk) 23:49, 30 November 2019 (UTC)

http://mshpriceguide.org/en/single-drug-information/?DMFId=1390&searchYear=2015

Please add your versions below. Once we have suggestions by people we can begin to discuss the merits of each. If you are happy with the version suggested by someone else feel free to add your name. This section is not for adding reasons why you think versions suggested by someone else are "wrong" that can come later or go above.

Doc James

"The wholesale cost in the developing world is about US$0.40 per day as of 2015." Doc James (talk · contribs · email) 16:51, 29 November 2019 (UTC)

James, you haven't chose a good example to aid discussion. Nor have you started from the right direction. You've picked one database record from several options. Your record has one supplier price. So it doesn't answer many of the questions we have. If your drug article is valproate and you want to make use of the MSH guide, can you do that without breaking WP:NOR, WP:WEIGHT and WP:NOT? This one article attempts to cover three drugs. In the US, the "valproic acid" and "valproate semisodium" forms are more common, whereas in Europe and elsewhere, "sodium valproate" is more common. See this for details -- the various forms are not directly interchangeable. Anyway, MSH have six entries:

• Sodium Valproate 200mg 100 enteric-coated tablets 5 suppliers. = $0.52 per day
• Sodium Valproate 250mg/5ml in a 120ml bottle No suppliers. 1 buyer. = $1.20 per day
• Sodium Valproate 500mg 100 enteric-coated tablets 1 supplier. = $0.40 per day (price in article)
• Valproic Acid 200mg/5ml in a 300ml bottle No suppliers. 1 buyer. = $1.10 per day
• Valproic Acid 200mg 100 enteric-coated tablets No suppliers. 1 buyer. = $1.30 per day
• Valproic Acid 150mg 100 tablets 1 supplier. = $0.82 per day
• There is no data for valproate semisodium in MSH.

It is quite a mess if you attempt to give a price for "valproate" because there is no such thing for sale. I see you quote a wholesale price in the US as $1.30 per day, which you are juxtaposing with the developing world price of $0.40 per day. But the developing world price is based on a bottle of 100 500mg enteric coated tablets of sodium valproate, whereas your US price is based on (unknown pack size) 250mg tablet (unknown if enteric coated) of valproic acid. The US source also give the price of 250mg/5ml solution. Maybe we should list the wholesale price per kilogramme of the pharmaceutical intermediate in 200kg drums at Alibaba :-).

• NADAC (filter results for "valproic acid"):

• VALPROIC ACID 250 MG/5 ML SOLN = $0.59 per day
• VALPROIC ACID 250 MG CAPSULE = $1.30 per day (price in article)

It is interesting that the NADAC gives price per day that is much much cheaper for the syrup vs pill. It is also odd that it only lists one pill size for valproic acid, and no prices for sodium valproate or valproate semisodium. This suggests some problems with usability of their database. The syrup should be more expensive than the pill. And the price for the pill seems high. I suspect they are giving the price of a brand name rather than a generic for the pill. They also don't say what kind of capsule it is -- see below because the price for different kinds varies greatly.

James says he can't access the BNF in Canada, only a printed edition which omits detailed prices. So I've reproduced the BNF information below. FYI you can also find prices in the NHS Drug Tariff December 2019 which lists prices in pence corresponding to the "Drug Tariff" price below. There are two prices listed below: an NHS Indicative Price and a Drug Tariff price. Pricing is complex. My understanding is that if a doctor prescribes a named brand drug, the pharmacist should generally supply that brand and should be compensated for the dearer price vs generic. If the doctor prescribes a generic drug name, the pharmacist generally will supply whatever they can get cheaply. How much the pharmacist is then compensated by the government depends on the prescription, the tarff code below, whether the medicine is currently on a list of price consessions or known shortages. There is also a problem with the assumption that off-patent generic drugs are cheaper: if there is no competition among suppliers, the generic price can rise. See Drug Tariff and PSNC Dispensing Supply. That's just for England & Wales. Scotland has its own drug tariff.

• BNF Sodium Valproate

• BNF Valproic Acid

Above are just the range of prices for England & Wales for one very commonly prescribed medicine. Most European countries will have a similar range of prices and options. The developing nations might have fewer options, but likely more than appears in the MSH guide. I have no idea why the US price list James used only has two prices: one syrup and one tablet, and with a very dodgy-looking price for the tablet. It seems clear that US source isn't adequate for our purpose.

It is interesting to see that the Drug Tariff price is proportionate wrt the dose per tablet for one given formulation: in most of the above tables, the Drug Tariff price, if multiplied to give 1500mg DDD, is equal for that table. The NHS Indicative Price is also generally proportionate wrt the number of tablets (30 tablets is 30% of the 100 tablet price), but does vary from one manufacturer to another for the same pill (see Gastro-resistant tablet and Oral solution). The Modified-release granules from Sanofi cost the same per sachet regardless of dose, and Desitin Pharma is considerably cheaper than Sanofi. This is likely a result of the UK's regulated socialist healthcare drug price policy, as the results from MHS and US show variation per pill size or quantity purchased. For example, diazepam wholesale price is about 2 cents for the 2mg, 5mg and 10mg pills, making the "cost per day" exquisitely sensitive on the totally unknown factor of how many divided doses are taken per day.

Not only is there no such thing as "a" price for "valproate", but Wikipedia talk:WikiProject Medicine/Archive 84#Price of medications RFC does not allow us to include the price of valproate unless "sources note the significance of the pricing" (such as for example, the extortionate US price is a result of low competition, or the high UK price meant NICE did not recommend its use). I know you wished the RFC went a different way, but it didn't. -- Colin°^Talk 20:20, 29 November 2019 (UTC)

RFC does not allow us to include the price of valproate unless "sources not the significance of the pricing", that's not true, the RFC's result was no consensus, which means that the prices may very well be included, or not, depending on local or future consensus (a possibility that is explicitly noted at the end of the closing message). It seems to me there are two issues here: whether prices, without a perspective, are admissible, and whether there is a price measure that is stable accepted enough to be of encyclopedic value. Although several editors voiced a positive opinion for the former issue, I would suggest to stick to studying the latter for now, as anyway without a good measure we can agree on, we can't add prices, and at worst even if we don't get any consensus, the info we gather can be used to extend articles on the topic. Signimu (talk) 19:08, 30 November 2019 (UTC)

Signimu, James asked the community to support his inclusion of pricing information in drug articles and the community did not give it. He does not have consensus to add drug information to articles. He failed to get permission to deviate from WP:NOT, which is policy. Specifically:

An article should not include product pricing or availability information unless there is an independent source and a justified reason for the mention. Encyclopedic significance may be indicated if mainstream media sources (not just product reviews) provide commentary on these details instead of just passing mention. Prices and product availability can vary widely from place to place and over time. Wikipedia is not a price comparison service to compare the prices of competing products, or the prices and availability of a single product from different vendors or retailers.

So the closing remarks on the RFC just echo WP:NOT. We need a source providing commentary on e.g., the high price of valproate in the US, in order for that to be significant enough to pass [[WP:NOT]. It isn't good enough, as James has claimed, that people are interested in drug pricing in general. It isn't sufficient for pricing to be listed at Drugs.com or BNF or any other catalogue because that is raw data like bus timetables, TV guides and restaurant menus. James feels strongly about this, has an open agenda about fighting Big Pharma secrecy, declared above that anyone disagreeing with him is a Pharma shill wanting to censor Wikipedia, and has past form for edit warring against community consensus. So I have no doubt that anyone removing prices will simply be edit-warred into submission. I can't read the paper you linked above but do note the comment "a major conclusion of this analysis is that international drug price comparisons are extremely sensitive to choices made about certain key methodological issues". We see this above, where James has picked a weirdly expensive form of valproate for the US price, which is a different dose strength to the developing world price, and juxtaposes them in the lead. He also has generally juxtaposed the developing world wholesale price with a US retail price, without declaring the US price is retail, which, you know, might make folk suspect we've got a POV-pushing agenda here. His "methodological" practice has been shown to be quite arbitrary, inconsistent, and not recognised by any authority, or used by any of our sources. If this was someone providing price information for homeopathic remedies, WP:MED would have flattened them like a pancake by now. There is no such thing as "the price of valproate". -- Colin°^Talk 13:28, 1 December 2019 (UTC)

Yes there are prices for valproate. And lots of sources to support this... Doc James (talk · contribs · email) 00:56, 2 December 2019 (UTC)

Once again, there's a huge amount of OR here, while there's clearly no one price that we can use in general. I'm afraid we'll be heading for ArbCom. --Ronz (talk) 16:29, 2 December 2019 (UTC)

James you are being wilfully obtuse wrt price/prices. You make a claim that "The [singular] wholesale cost in the developing world is about $0.40 per day [original-research] as of 2015 [out-of-date source]" and cherry-pick one database record from a source, which happens to have only one supplier. You then conduct OR to present this as a daily price of $0.40 per day. You ignore that the same source would give prices of $0.52, $0.82, $1.10, $1.20, $1.30 per day, using your method, on other database records. And you are not acknowledging the statistical nonsense you performed elsewhere with your min/max pricing. For the US price, you cherry-pick one database record from a source to get $1.30 per day, for not only a different dose pill but a completely different chemical form (valproic acid, vs sodium valproate). You ignore that the same source would also give a price of $0.59 per day if you chose a different record. Hmm, one might almost think you had an agenda, picking prices that way. I agree, I think we need to take this to a different court to get people competent in policy to examine. The practice of inserting incorrect and misleading prices in to leads and body text, without meeting WP:NOT requirements, seems to be done solely by Doc James. -- Colin°^Talk 18:03, 2 December 2019 (UTC)

Colin stop pinging me as I have previously requested. "About" is not a singular price. Doc James (talk · contribs · email) 19:04, 2 December 2019 (UTC)

James As long as you won't drop this issue, you'll get pinged whenever I mention your name. As an admin, you hold a position of power and responsibility. That means you must be accountable to the community and are required to respond when challenged on your editing. If you wish a quieter life, you can resign as admin, and accept a topic ban on drug pricing. "About: adjective, "in the vicinity of", "approximately", "nearly", "close to". Are you really claiming $0.40 is approximately $1.30? There is no statistical method where the range of prices for different pill sizes or syrups from different suppliers can be combined to produce an "about" value that conveniently picks the cheapest price in the developing world, yet conveniently also picks the dearest price in the US.

At Ethosuximide you claim '"The wholesale cost in the developing world is about US$27.77 per month. In the United States the wholesale cost as of 2016 is about US$41.55 per month for a typical dose."' You say "typical dose" even though WHO says "DDD is sometimes a dose that is rarely or never prescribed because it is an average of two or more commonly used doses" and "The DDD is a unit of measurement and does not necessarily correspond to the recommended or Prescribed Daily Dose (PDD). Therapeutic doses for individual patients and patient groups will often differ from the DDD as they will be based on individual characteristics such as age, weight, ethnic differences, type and severity of disease, and pharmacokinetic considerations." It isn't a "typical dose" at all, and quite dangerous at misleading to suggest to our readers that it might be. Your "developing world" price is only for one supplier who only sells to government or NGOs in the Democratic Republic of Congo. So you've claimed your figure is "about", expressed to four significant figures of precision, as though there was a range of prices nearby, when in fact, you have absolutely no idea what the price of ethosuximide in other developing nations is. And the US price from this source. You'll have to help me here, because I fear I have made a mistake with the calculator. I filter the results for "ethosuximide". I get two prices.

• $0.27656 per ML of "ETHOSUXIMIDE 250 MG/5 ML SOLN". Using your "Defined Daily Dose: 1.25 G" OR-calculation, that means I need to multiply the ML price by 5 to get 250MG and by 5 to get 1.25G. Which is $6.914 per day or $207.42 for 30 days.
• $0.99956 per "ETHOSUXIMIDE 250 MG CAPSULE". Using the above we multiply by 5 to get 1.25G which is $4.9978 or $149.93 for 30 days.

Perhaps I have made a mistake? Or do you think $41.55 is approximately $207.42 or about $149.93? -- Colin°^Talk 22:37, 2 December 2019 (UTC)

Ah yes I see I interpreted the NADAC to be for 250 mg when it was only for 50 mg. Have fixed and updated to 2019.

Sure ""DDD is sometimes a dose that is rarely or never prescribed because it is an average of two or more commonly used doses" But it is not in this or most cases.

With respect to valproate I used the lowest cost formulation. Doc James (talk · contribs · email) 00:19, 3 December 2019 (UTC)

James That error of yours was present for three years. For three years, Wikipedia reported the price as $41.55 rather than $149.93. Not only are these prices calculated and cherry-picked by an arbitrary and original-research method, but many of them are just plain wrong. Wrt valproate, no, you used the lowest cost for the developing world, and the highest cost for the US. And you compared enteric-coated 500mg sodium valproate tablets with unspecified 250mg valproic acid capsules. As we can see above, the price for crushable, enteric-coated, gastro-resistant, capsules and tablets are all different, yet the US weirdly only lists one and does not specify what kind of capsule. And we can see from e.g. diazepam, that the US with its dysfunctional drug pricing system, has a wholesale cost-per-pill that does not necessarily take dose into account at all.

James, do you realise that at Ethosuximide you are now comparing a wholesale price in the Democratic Republic of Congo in 2014 (falsely claiming this is the current "developing world" price) with a suspiciously incomplete database record for US wholesale prices in December 2019. Five years later. So you are showing a price now of $95.97 (extraordinary precision) when the date-comparable price in Dec 2015 was $166.17 (if we ignore for now, that you are comparing apples and oranges, and the price in DRC of all places, a country in civil war in 2015). The fact that these prices randomly bounce around enormously when you use different database records, different years, different suppliers and the fact that nobody has cared about the price enough to even check it is in the right ballpark, just adds to the argument that this is not Wikipedia's job. --Colin°^Talk 16:39, 3 December 2019 (UTC)

Frankly I'm puzzled by all this discussion. The cost of the daily dose for the healthcare system is an obviously important piece of information for the readers interested in healthcare (i.e. everybody). Pretty much all the world coordinates on this, except one big outlier, and except some rare cases of compulsory licenses. The supposed "errors" could have been fixed simply by reducing the precision of the claim, for instance in my opinion "less than 1" is better than "0.57" and "less than 30" is better than "27.77", unless the body of the article is debating the differences in the decimals. Nemo 07:25, 3 December 2019 (UTC)

There is no such thing, for many drugs, as "the cost of the daily dose". There appears to be a political agenda by one user attempting right great wrongs and we all know that in politics, the truth no longer seems to matter, nor does it seem to matter if a claim makes any sense. --Colin°^Talk 16:39, 3 December 2019 (UTC)

User:Nemo_bis I have a medical textbook that has price ranges for medications in the United States.[25] Less accessible than this[26] so I have generally gone with the later. Doc James (talk · contribs · email) 02:55, 4 December 2019 (UTC)

Colin — It is abundantly clear to me that this isn't about one user trying to right great wrongs by including prices, but rather the opposite: One user with a radical agenda to invalidate WHO and government set standards for measuring price — Prices that have been included on Wikipedia based on high quality, if not perfect quality sources. Your position is clearly the one that is radical, and it should be abundantly clear to anyone who has spent time writing on Wikipedia that what we don't include is as much an ideological choice, as what we do — and for the indisputable ideology that Wikipedia stands on (WP:PILLARs), we have guidelines and policies such as WP:CENSOR. We can nearly always source things better, but your argument seems to rest on the fact that sources say different things, but we can and will, as per our standards, simply use the best available sources and summarize what is WP:DUE: detailing who said what and when. This is standard practice and not up for discussion, and you are quite clearly WP:BLUDGEONing the process here. We've heard what you have to say and it doesn't hold up to scrutiny. Carl Fredrik ^talk 09:53, 4 December 2019 (UTC)

CFCF hope you don't mind me not rebutting the many mistakes in your comment. Life is too short, and I'd rather discuss mistakes in all our drugs articles, and an egregious breach of WP:NOT by one editor with a fanbase. -- Colin°^Talk 13:38, 4 December 2019 (UTC)

Colin — I do mind, quite a lot in fact. I find your refusal to engage in the points as indicative of your attitude towards Wikipedia collaboration overall. You are clearly versed in the intricacies of Wikipedia's policies and guidelines, but so are others and you do not hold any unique capacity to interpret them. You are regardless of your opinion objectively engaging in WP:BLUDGEONING. Carl Fredrik ^talk 16:40, 5 December 2019 (UTC)

CFCF, why should I respond to an editor who's opening remarks earlier here "Colin, You really seem clueless here", as I took apart James's editing and sourcing claims bit by bit. CFCF, you played the WP:CENSOR card. Bzzzt. End of discussion. -- Colin°^Talk 17:02, 5 December 2019 (UTC)

@CFCF: I suggest you look more closely, while avoiding escallating a BATTLE situation in the meantime. --Ronz (talk) 16:49, 4 December 2019 (UTC)

• Wikipedia:Administrators'_noticeboard/Incidents#User:Colin I have brought these concerns to ANI.--Ozzie10aaaa (talk) 01:31, 5 December 2019 (UTC)

• Comment This seems to me to be in essence a philosophical dispute about where Wikipedia should sit on the information⟷knowledge spectrum. My view is we should, per WP:NOT, learn towards the "knowledge" end of this and only include drug prices where there is discussion in secondary literature to establish due weight. Alexbrn (talk) 08:34, 5 December 2019 (UTC)
  • Alexbrn thanks for that. I agree, though the "philosophical dispute" was long settled, and this is more a case of editing against known community consensus. And with so many source->content errors, it raises serious questions of competence to edit in this field. It seems however, WP:MED intends to silence any criticism of its approach. -- Colin°^Talk 10:21, 5 December 2019 (UTC)

Above James wrote "I have a medical textbook that has price ranges for medications in the United States.[27] Less accessible than this[28] so I have generally gone with the later.". I don't have that textbook but Amazon.com let me peek at a little of it. I found two pages of interest. The first explains their drug pricing methodology. I'll quote it: (the underline italics is theirs)

The price in the Tarascon Pocket Pharmacopoeia 2019 for Diclofenac is $$$, i.e. $50 to 99. And indeed James used that price in an earlier version of the article, which said "As of 2015 the cost for a typical month of medication in the United States is 50 to 100 USD." So the price in 2015 must be similar to what I can read in the 2019 version. Neither the body text nor footnote gives any of the above explanatory text for how this "cost" is derived, how it should be treated with caution, and that in fact it is an wholesale list price that may not even represent the cost to a hospital, or pharmacy, let alone the retail cost to a patient or reader.

James then updated the price using the National Average Drug Acquisition Cost (NADAC) weekly reference data as of 2018-12-19. The article now says "In the United states the wholesale cost per dose is less than 0.15 USD as of 2018" When I examine that source, I find lots of prices for "diclofenac": 0.1% eye drops, 1.5% topical solution, 1% gel, tablets in 25, 50, 75 and 100mg, tablets containing the sodium salt, the potassium salt, tablets in combination with misoprostol, enteric-coated tablets, delayed-release tablets, extended-releases tablets. Most of the diclofenac sodium tablets are indeed less than 15 cents a tablet which is I guess what James used to mean "dose" when he wrote this (vs "defined daily dose" when he wrote "dose" on some other articles). The 100mg sodium tablet is oddly expensive at $1.86 and the diclofenac potassium tablets are 37 cents. The tablets combined with misoprostol (to protect the stomach) are about $1.70 each.

Let's try to compare the two figures. Above James argues for the use of defined daily dose to perform our cost-per-day (or month) calculation. The DDD of diclofenac is 100mg. So lets pick one of the cheap ("less than 0.15 USD per dose") tablets 50mg at $0.126 each. Twice a day comes to $0.252 per day or $7.56 for 30 days supply.

So we've taken James's original-research calculation methodology and produced a figure of $7.56 a month vs Tarascon Pocket Pharmacopoeia 2019 which claims the typical wholesale cost is between $50 and $99 per month. That is a factor of 7x to 13x difference in "cost" depending on how you work it out. The retail price of 50mg delayed-release tablet is about $30 per month if bought as a pack of 60 tablets.

We can't reproduce Tarascon's methodology. We don't have a source for "the most common indication and route". For all the reader knows, voltarol gel for arthritis could vastly outsell the tablets, and we have no idea how popular the various enteric-coated/extended-release forms are, or whether the misoprostol combination is routine or exception. We have no idea if the potassium or sodium salt variants are chosen most commonly. We don't know if it is used as maintenance long-term, or for short term occasionally. And any of this could vary from country to country depending on prescribing practice, brand availability and licencing.

Is the answer to go back to using the Pocket Pharmacopoeia? Well not really. The $$ symbols are clearly there as "a rough guide only" to aid the prescribing US physician as to whether this was generally a cheap or expensive drug in the US alone. But they know the limitations and know that price range is for one kind and size of tablet only, for one typical indication, in adults, and that the eye drops, topical solutions, gels, patches and other variants will likely not be in the $50 to $100 range but considerably more, or that the price for children's doses of medicines have to take weight and age into account, but also other formulas such as syrups. And they know this is a "wholesale list price" whereas James claimed it was the "cost for a typical month of medication in the United States" which most readers would assume was the retail price a patient pays. This is all too complex to explain in every article. Furthermore, the Tarascon $$$ price guide is an expert opinion, not a fact, and reproducing that across multiple articles on Wikipedia, could lead us to foul copyright (just as we couldn't open every article with verbatim definition from the Oxford English Dictionary, or reproduce the DSM-IV definitions).

A further problem with the Pocket Pharmacopoeia's $$ price scale is that information has been lost when the "average monthly cost" was inserted into one of 5 symbolic forms for brevity (the $$ appear next to a whole load of other symbols). You can't then say "the cost for a typical month of medication in the United States is 50 to 100 USD". Their data might have the cost of a typical month of medication ranging from $20 to $300 depending on supplier, and when they calculate their average (they don't say if this is mean or median) they get a single figure such as $55, not a range at all. The only reason for the range is to arbitrarily assign how many $ symbols. So in fact the "cost for a typical month of medication in the United states is 20 to 300 USD" might be the truthful statement according to Tarascon's PP raw data, but we will never know. The information was lost and the numbers 50 and 100 have no special meaning for our readers whatsoever. They are artefacts.

So is it $7 a month, $30 a month, $55 a month, or $50-100 a month? This is why we don't allow original research. And why claiming a "drug" has "a cost" is not appropriate for Wikipedia. -- Colin°^Talk 16:28, 4 December 2019 (UTC)

Yes I used that price back in 2016. Yes I updated it to NADAC and clarified that this is wholesale.

Sure the prices in the Pocket Pharmacopoeia is a rough guide. Nothing wrong with that.

We provide expert opinions all the time. We are not including $$$. We are presenting this data in a different form so no it is not copyright infringement.

We include DSM criteria but we paraphrase. Those are also expert opinion by the way. No different than here.

The underlying data actually comes from "Micromedex RED BOOK" from what I understand [29] But these experts summarize it. Doc James (talk · contribs · email) 22:09, 4 December 2019 (UTC)

that is not original research, concur w/ Doc James--Ozzie10aaaa (talk) 00:38, 5 December 2019 (UTC)

By the way NADAC and average wholesale price (AWP) are not the same. There are a few types of wholesale prices and I will link to the specific one I mean going forwards. The AWP is greater than the NADAC and thus the pricing discrepancy. AWP is often closer to actual retail price. Doc James (talk · contribs · email) 22:16, 4 December 2019 (UTC)

James, you can't just tweak one mistake and leave in all the others. For example, the meaning of "dose" varies depending on which article you edited, with this one being a small dose tablet that needs to be taken several times a day, but in other articles you pick a large dose tablet that could only be taken once a day. I'm afraid, James, you don't understand database copyright, but there are enough mistakes here without getting bogged down in that one. You haven't appreciated the central point is that your Pocket Pharmacopoeia gave "wholesale" prices 7x to 13x higher than the price you are now calculated by Original Research, and higher than the retail price from Drugs.com. This isn't an NADAC vs avg wholesale difference. This is a whole different ballpark difference. Are they picking an expensive formulation of tablet or a gel because they know something we don't about what's commonly prescribed? Your last comment, about the complexities of what we even mean by "wholesale price" is the very first time you have accepted that pricing might not be simple to calculate, to present, or for our readers to understand. There's a whole terminology wormhole our readers could fall down trying to understand different forms of wholesale pricing, discounting and rebates, and that's just the US. And that's before we even consider if some hard-to-explain variant of wholesale price is evening meaningful to a reader paying a retail price or a prescription charge. The problem James, is that, as you say "these experts summarize [the underlying data]" for your Pocket Pharmacopoeia, which is aimed at a medical professional audience, but you User:Doc James, are not allowed to do that, and are failing to present anything coherent to a lay readership of an encyclopaedia. -- Colin°^Talk 23:16, 4 December 2019 (UTC)

We simplify a whole lot of stuff. And we base everything we do on the expert opinion within published sources. There is a massive debate in academia about the appropriateness of p<0.05 for example. Saying something is complicated does not mean we should give up.

Per your claim "NADAC vs avg wholesale" is not the reason. Did you check the Red Book? I do not have access so have not. Doc James (talk · contribs · email) 23:26, 4 December 2019 (UTC)

So lets look at what different sources say about the cost of a month of diclofenac in different parts of the world.

Lets start with the DDD which is 100 mg.[30] Is that reasonable? Listed as a reasonable dose in the product monograph so yes.[31] So 60 X 50 mg tabs for a month.

• MSH gives a price of 0.0024 to 0.0296 per 50 mg tab.[32] so 0.14 to 1.78 per month in LMIC. Summarized this as less than $US2 per month wholesale as of 2015.

• NADAC gives a price of 0.13 per 50 mg tab.[33] So a NADAC wholesale cost of 7.8 for a month at DDD in the US.

• Lowest retail prices are listed at 0.23 each when buying 1000 tabs (13.80).[34] But are around 30 USD for a month worth of the 50s.[35] This site lists prices as low as 22.30 for this amount for a month.[36]

• The average wholesale cost, is the wholesaler or manufacturers offered price, and is listed as 50 to 100 USD as of 2019 by Tarascon. I have generally moved away from these for generic as they are inflated as we see in this case. This review lists the AWC as 79 USD so I would not be surprised if this is what the manufacturer still lists it for.^[1]

We thus see a US retail price for a month of DDD between the NADAC and AWC. And we see a cost in LMIC at less than in the US. Nothing really surprising here. Useful, definitely. Doc James (talk · contribs · email) 23:55, 4 December 2019 (UTC)

• James, while it has been interesting to pick apart the prices in, well, basically any article I've looked at, it doesn't move us forward to keep performing original research on numbers that frankly neither you or I or anyone else here really understands. The cost per DDD is not an accepted method for presenting price data to a lay audience -- it has very limited use for research purposes or very limited comparison purposes. You continue to use the high/low prices from MSH despite MSH explaining that those could be outliers and recommend using the median supplier price. You quote "0.0296" as an upper price, but this is a "buyer" price which MSH say "should not be used as international reference prices". And you link to External reference pricing as an easter-egg link for "wholesale price". Even looking at just the supplier prices, it is clear AMSTELFAR is an outlier, which further emphasises why MSH recommend the median. While MSH can be used for external reference pricing by expert researchers who know what they are doing, that doesn't begin to justify citing raw database record results, choosing the 50mg tablet vs some other size or syrup, and conducting original research. After doing all your maths above, to get $7.8 you then write in the article "typically less than US$9". James, where did $9 come from? This is a clue about why original research is a bad thing: a different editor using the same raw data will come up with different numbers. That is an interesting admission you make about the Tarascon: "I have generally moved away from these for generic as they are inflated as we see in this case". So these experts, with their expert methodology that I quote above, are dismissed by you as being "inflated", based on what? Your own expert opinion? In the earlier RFC you were quite happy to claim this source and earlier you only rejected it as being hard for others to access online, but now it seems you reject it because you think your methodology is better than theirs? Hmm. -- Colin°^Talk 17:13, 6 December 2019 (UTC)

No, inflated per sources.[37][38] Average wholesale price is still a reasonable estimate and I have no problem with people using it. Estimating prices is not black and white. This is similar to, well everything in medicine. Doc James (talk · contribs · email) 20:18, 6 December 2019 (UTC)

Ah, now you are getting into politics. What do we mean by "cost" -- the answer is not simple you and a few others have repeatedly claimed cost is a simple thing readers want to know. If "cost" is Medicare's reimbursement then the AWP "list price" is closer. If "cost" is what the pharmacy paid after discount then perhaps the NADAC is closer. Your first source says the AWP was typically inflated 20%. But your pocket book gave a cost for diclofenac of $50-100 whereas your original research gave a cost of $7.80, and then weirdly, you wrote $9 in the article. So the pocket book was no where near 20% inflated. That's 540% to 1180% higher. The UK BNF prices have a similar issue: which cost do our readers want and can we even begin to explain to them what it is. The "cost" in the MSH is also complicated, and you still haven't accepted that you really mustn't use the Buyer price for your reference price. They explain about cost: "Direct comparison of prices between suppliers for a given product may be unrealistic because of the varying terms and services offered by different suppliers. When calculating and projecting real costs, it is important to consider factors such as financing, delivery times, modes of transportation (air/sea/inland), and handling charges. An analysis of these factors is beyond the scope and intent of this Guide. However, it is reasonable, for estimation purposes, to add 10% for shipping costs to the listed price for suppliers in the guide.". So are we going to do some more OR and add 10% to all the MSH prices? And if "list price" isn't what you want to use in the US, why are you accepting "list price" from suppliers in MSH. There could well be further discounts achievable -- after all, helping negotiating a price is part of what the MSH guide is for. It really comes down to "which cost". And that is far from simple. -- Colin°^Talk 14:42, 7 December 2019 (UTC)