scratch to reveal album cover
Image via Wikipedia.

When I filed my first Freedom of Information Act (FOIA) request back in November of 2012, I had no idea the volume and content of the materials I might receive. Would it be a small packet of innocuous documents? Just a letter with nothing to follow?

Now, with well over 1,000 pages stacked up and more than a dozen blog posts filed, it seems time to take a look back at what we’ve discovered during 2013 about PubMed Central (PMC), its management, its behavior, and its limitations. This post is an attempt to inventory and summarize what we now know, for the sake of putting it all together a bit better and presenting you with a meaningful retrospective.

Most of the posts discussed below can be found here, along with some related posts.

PMC management plays favorites. This was revealed in a number of ways. Documents show that PMC greased the skids for eLife to promote its launch and allowed eLife to use PMC as a publishing platform while other publishers were ignored or actively rejected. Documents show that PMC allowed Wellcome Trust to appoint its own National Advisory Committee (NAC) member replacement. Emails show that PMC has done favors for BioMed Central and PLOS, while shutting other publishers out entirely or affecting an air of unbiased fairness in order to maintain the appearance of objectivity while at the same time doing favors for their cronies. And documents show that PMC management speaks of “relationships” and “friends” as they play favorites.

PMC allowed US government resources to be used by a UK startup. While open access (OA) is supposedly all about making sure taxpayers get ultimate value from the research their taxes funded, PMC spent US taxpayer money via contractors and technology deployment to support the launch of eLife, a funder-backed journal with its primary offices in the UK.

PMC spends a lot of US taxpayer money uploading and checking author manuscripts. The budget for PMC is about $4.5 million, and nearly all of it is spent dealing with author manuscripts, a dubious expense given the high duplication rate these have with publisher deposits. In addition, PMC’s efforts to increase compliance by NIH researchers could lead to a doubling of expenses in an environment of government spending caps, putting the PMC premise in question.

US taxpayers are not the primary beneficiaries of PMC. Traffic estimates show that US taxpayers provide PMC with only about 40% of its traffic, making others the major beneficiaries of a system US taxpayers are paying for.

PMC management pulled the wool over the eyes of its own advisory committee. The PMC NAC was kept in the dark about how PMC and eLife were collaborating on launching eLife content early on PMC, so much so that the June 2012 meeting — contemporaneous with a high level of activity around the eLife launch on PMC — featured a talk by Mark Patterson from eLife during which nothing of the joint effort was mentioned. The same meeting also featured David Lipman, Executive Secretary of the NAC, speaking as if nothing unusual were going on between the two groups.

The PMC NAC may be too weak to actually oversee PMC. Despite all of the revelations this year about conflicts of interest, insider dealing, improper use of government resources, and direct questions about its own role in such matters, minutes from the June 2013 PMC NAC meeting reflect no discussion or even awareness of these issues among members of the group. This suggests that the PMC NAC is not designed to provide effective oversight, or lacks the ability to carry out this function.

Two journals are using PMC as their primary publications platform. The Journal of the Medical Library Association (JMLA) and the Journal of Biomolecular Techniques (JBT) slipped into PMC during an early phase, and have benefited from free government publishing services ever since. In fact, JBT moved from commercial hosting to PMC-only hosting when its editor was named to the PMC NAC, another sign of insider dealing. PMC’s hosting of these journals not only deprives technology providers from revenues and costs taxpayers money (both lowering tax revenues and driving traffic to PMC, increasing bandwidth costs), but makes these organizations look like they are freeloading.

PMC sought to hide is work with eLife from the community and the US taxpayer. Numerous emails show that staff and management at PMC and the NIH sought to conceal their efforts with eLife to expedite its launch using PMC. They were clearly uncomfortable about what was being done, and worked to sweep it under the rug.

eLife and PMC collaborated on eLife editorial material and publisher messaging. PMC staff were asked to review eLife launch editorial drafts prior to publication, because the two groups were so cozy, leading to changes that obscured how the two entities had worked together. They also coordinated communications to other publishers in order to divert suspicions.

Funders aren’t saints. When Wellcome Trust, HHMI, and Max Planck announced their intention to publish eLife, the entry of funders as publishers struck many as fundamentally wrong. But the reputations of these groups are superlative, so things were allowed to proceed. If the past year has demonstrated anything, it is that funders need as much monitoring as any other publisher, if not more. Their incentives are different, but they are incentives that can lead to shortcuts, collusion, and exploitation. Funders are not saints, nor are the editors they hire or the staff working for them. For example, one of the editors of eLife publicly blamed PMC for eLife’s launch assistance, attempting to deflect blame despite documentation showing the request came from eLife (via a Wellcome Trust email account).

David Lipman was not honest or open about his role in helping eLife publish first on PMC. Starting with assertions that eLife had followed “the usual process” and that he knew nothing of the details, Lipman sought to distance himself from the eLife situation. Documents revealed that he not only was a primary instigator of the effort to publish eLife on PMC, he remained heavily involved throughout. When confronted by a citizen and taxpayer, he equivocated and evaded responsibility or accountability. He has not spoken publicly since, except to say in an interview that he felt no need to respond.

eLife management and editors also ran away from the situation. Through a combination of emails and comments after the first emails were revealed, it became clear that editors and staff at eLife also wanted to distance themselves from the situation and not take personal responsibility for their roles in getting special treatment for eLife via a US government agency.

The relationship between Lipman and Vitek Tracz requires close examination. Favors done by Lipman for Tracz apparently include actively working to get F1000 Research more actively addressed by PMC contractors and staff, getting special treatment for BioMed Central titles, and perhaps even doing something else in the early days of BioMed Central, which current emails only hint at. This relationship requires a very close look, especially given Lipman’s role as an editor on a BioMed Central journal (hired by Tracz for this role).

Leaders at Wellcome Trust, PMC, and eLife apparently have no grasp of what constitutes a conflict of interest. Evidence shows leaders in these organizations exploiting conflicts of interest and speaking in ways that show they have little grasp of what constitutes a conflict of interest, suggesting these people need some remedial training on these issues.

The search results of PubMed are designed specifically to compete with publishers, and keep traffic going to PMC. Emails show that the rationale for the design that places links to PMC versions on the search results list while links to publisher versions are one layer deeper, requiring another click and page load, is primarily driven by the desire to retain traffic and compete with publishers.

Some OA advocates failed to see the problem here. While not revealed directly by the documents obtained by the FOIA requests, the fact that some strident OA advocates accused me of being anti-OA throughout this, or anti-science, beggars reason. OA publishers were the ones being mistreated. This was not “anti-OA” but “anti-cronyism” and “anti-deception.” It was about being open, about accountability, and about understanding how a US government agency became a tool used by a UK-based charity cum publisher. Science is about gathering and evaluating evidence in light of a hypothesis to see if the hypothesis proves out. I merely gathered evidence based on the hypothesis that PMC leadership was unwilling to tell the entire story, and I believe the evidence supports the hypothesis.

If someone had told me that the FOIA requests I filed would generate evidence enough to justify even one of the above statements, I would have been surprised. However, over the past year, the scope of mismanagement and its apparent duration have been astonishing to me. Clearly, there has been something rotten in Bethesda, as I first sensed. However, I never could have guessed that so much would be wrong, biased, improper, or unfair. It’s a shocking list.

And it’s not over with yet. Some requests remain ongoing, and are producing materials still. Stay tuned.

For me, 2013 will go down as the year where we saw that PMC has significant management shortcomings and flaws in its governance. It cannot lecture anyone about standards, fairness, openness, or propriety — at least not until it becomes openly accountable and sets things right.

Kent Anderson

Kent Anderson

Kent Anderson is the CEO of RedLink and RedLink Network, a past-President of SSP, and the founder of the Scholarly Kitchen. He has worked as Publisher at AAAS/Science, CEO/Publisher of JBJS, Inc., a publishing executive at the Massachusetts Medical Society, Publishing Director of the New England Journal of Medicine, and Director of Medical Journals at the American Academy of Pediatrics. Opinions on social media or blogs are his own.

Discussion

56 Thoughts on "PubMed Central Revealed — Reviewing and Interpreting the Findings of a Surprising 2013"

So, in the face of all this evidence, one wonders, why is no action being taken in the government to correct all these faults and mete out appropriate punishment for misbehavior?

These posts have gotten some attention from serious people in the government, so 2014 might be a year where the chickens come home to roost. We will see.

You have done nice work exposing what seems to be a serious problem with a government agency. Lots of smoke, maybe some fire. What I have not seen is you taking the next logical step: presenting these findings to the HHS Inspector General and PubMed Central’s congressional funding committees. Have you followed up on this?

Thanks.

I feel I’ve done my part. Do you? The more people who point to all this, the more attention it gets. As a taxpayer, citizen, and scientist, I’m sure your voice will add a great deal.

Kent – You have done way more than your part and I’m hopeful the community has expressed its gratitude to you either personally or publicly. I have been in contact with my congressman’s office as a result of your investigation and hopefully will work toward an understanding of the seemingly disparate treatment of smaller publishers by PMC and MEDLINE (our first MEDLINE report card was so perfunctory it didn’t even have simple math averaged properly and PMC essentially rejected us because we don’t publish review articles even though PMC has over 11,000 roundtable discussions). Consequently, until we do get indexed, our content has been locked behind a paywall rather than open access.

While access to research and other issues are important topics, the reality is “indexing” is critical to survival of any biomedical publication – without it a publication will eventually cease to exist. Assuming the publication has value, the National Library of Medicine knows it has the power to “make or break” publications – particularly since start ups most likely cannot last years waiting for the NLM to index, NLM can simply let these publications wither and die.

For instance, the NLM has been written into the definition of “medical journal” in federal code 21 CFR 99.3:

j) Scientific or medical journal means a scientific or medical publication:

(4) That is indexed in the Index Medicus of the National Library of Medicine of the National Institutes of Health;

This is part of the required elements for pharma companies to distribute off-label reprints – a valuable source of information and of revenue for publishers. Without it, compliance will not sign off on distribution.

Indexing has the ability to allow a business to succeed – and giving eLife and others a VIP pass to the front of the line is certainly worth assessing carefully.

In addition, this past June during the NAC meeting, Dr. Lippman suggested that PMC would start instituting “objective criteria” in an application – like MEDLINE. I think this is an acknowledgment that PMC understand’s it needs to change.

What you see as collusion looks like reasonable collaboration toward making results of publicly funded research freely available to public. If NIH provided only the publishing platform on the initial stage, then the US government contribution probably is fraction of overall cost of eLife operations. Be the way eLife is incorporated at Delaware as well in UK. So your point that it is “UK startup” is not true. But since it is not profit and internet based (i.e. employees are spread cross the world), place of registration doesn’t matter. Probably most or significant proportion of eLife published papers come from US just because where most biomedical research is done. But even they publish only foreign papers , US taxpayers would benefit from it because access to scientific research results is crucial for technological progress . Since most of world pharmaceutical industry is located in US, one can argue that open access to life science publication, will benefit US more then other countries

I don’t think the OA publishers who were rejected during the same time period that eLife was getting assistance would agree that the PMC-eLife relationship looked like “reasonable collaboration.”

I don’t think many would agree that when PMC states specifically that it “is not a publisher” and then selectively provides publishing services to self-described “friends” that this is either fair or a valid “US government contribution.”

eLife’s Delaware incorporation is a distraction. The real action occurs in the UK, and the emails involved in these affairs started at Wellcome Trust and then moved to the eLife offices Wellcome set up down the road. It was a Wellcome start-up in the UK. Splitting hairs does not change the facts of the matter.

Your claim that because “it is not profit and internet based” that place of registration doesn’t matter is baseless. The UK has a different brief for charities and non-profits, and different reporting and qualification requirements. Nice video overview here: http://www.youtube.com/watch?v=Wqf4YVUiRxQ Go beyond the US/UK differences, and being registered in Egypt, India, or other countries introduces more differences still. “Virtual” does not mean “unreal” and often the legal constructs needed to create virtual organizations are more complex than location-based firms.

Saying that US taxpayers benefit from paying for others’ access is a stretch. Most of the traffic coming to PMC originates in China, a country that has been eating our economic lunch for the better part of two decades. US manufacturing has been slipping away, partly because of foolish “virtual reality” thinking and partly because our IP is everywhere. At some point, harsh reality will likely intervene into pipe dreams like these. Ask the long-term unemployed in the US how they feel about taxpayers funding China’s access to US-funded scientific papers.

These kinds of comments miss the point of this series, which was ultimately about fairness.

Is it fair for PMC to have “friends” for whom it bends policies, courts conflicts of interest, and deceives the public? Is it fair for PMC to spend taxpayer funds in wasteful ways? Is it fair for PMC to have US taxpayers overpay for the benefit they receive from PMC? Is it fair for US taxpayers to help UK companies? Is it fair for some journals to get free publishing services from the US government?

After reviewing all the evidence, I think PMC is unfair, biased, conflicted, and dismissive of its obligations to the public it supposedly serves.

“Saying that US taxpayers benefit from paying for others’ access is a stretch. Most of the traffic coming to PMC originates in China, a country that has been eating our economic lunch for the better part of two decades. US manufacturing has been slipping away, partly because of foolish “virtual reality” thinking and partly because our IP is everywhere. At some point, harsh reality will likely intervene into pipe dreams like these. Ask the long-term unemployed in the US how they feel about taxpayers funding China’s access to US-funded scientific papers.”

Firstly, it is grossly wrong to talk just about funding access to US funded research. Whilst it is true that NIH funding requirements generates a substantial amount of content (particularly from closed access publishers who wouldn’t otherwise participate), PMC is taking all of the [eligible] content from the publishers, regardless of where it was funded.

Secondly, even leaving aside any other potential benefits of collecting this information in a single place, it is still of benefit to people in the US.

If you take the case of Jack Andraka by itself, then the health care and economic benefits of what he has done would more than justify the cost of running PMC for many years. Now, does other countries accessing PMC actively harm the US economy to an equal or greater degree? Or, does it improve global health care, fuel global economic growth – does increased wealth in China marginalize it being the world’s sweat shop, increase imports (e.g. from US companies)?

I’m not saying it does or doesn’t – but it’s a vastly, vastly more complicated equation than is painted above.

PMC is not charged with taking content from beyond the US, and correspondence from within NIH and PMC shows a real discomfort with the amount of content coming from outside the US. Funding of government initiatives is politically charged, and the politics of US taxpayers paying for things non-taxpayers use and benefit from is risky.

The “benefit” to people in the US is tenuous at best, as I point out. PMC asserts that 52,000 non-US papers reside in PMC in one year. For each taxpayer, they’d have to read 140+ papers per day to “benefit” from this. Considering that these papers are not only written at a level that most Americans aren’t literate enough to read, and that they are highly specialized reports of limited incremental value, this assertion of “benefit” is very weak.

The case of Jack Andraka is worth revisiting for a variety of reasons. It’s often tossed up as an unalloyed example of the benefits of OA and lay research, but both are in fact wrong. First, Andraka comes from an elite family in both educational and financial terms. His parents are both well-paid science professionals, and his household income is probably $200K to $400K per year. He can afford to purchase any content he wants. His parents both are science enthusiasts, as well, and his older brother is probably a more accomplished scientist than Jack, having earned many financial benefits from this. But now to Jack’s science, which was accepted by the media with little scrutiny but which has since been refuted in larger studies by real scientists. So, it’s really not the breakthrough you think. The story I get out of this is that a rich kid with science parents made a splash with findings that have since turned out to be wrong. He could afford to pay for content, so OA made no economic difference in the outcome (except, perhaps, he should have paid for more content perhaps rather than relying on OA publications, Wikipedia, and YouTube). And his findings are not holding up. So, what’s the lesson now? That free content is inferior? That the cost of him using free science to get a bogus finding justifies PMC? Huh?

Also perhaps worth noting that Andraka’s results have never been published in a peer-reviewed journal and that he has patented those results, locking up their use behind a paywall.

The first NLM goal https://www.nlm.nih.gov/about/functstatement.html:
“(1) assists the advancement of medical and related sciences through the collection, dissemination, and exchange of information important to the progress of medicine and health; “
Internet is most obvious and cost efficient medium to achieve this goal. Apparently “side effect” of this activity is that foreigners have access to NLM hosted scientific information. I personally firmly believe that scientific and technological progress is global by nature and spreads benefits globally. Simply put, you would not mind if you or your child is cured by a drug invented and produced in China. As I told before, currently US is the center of biomedical research and drug development. US based pharmaceutical companies pay corporate taxes in US, their employees – income and local taxes. My point about non-profit status of eLife was about that they don’t pay corporate taxes in UK. What alternatives do you suggest? Block access to nlm.gov from China or disseminate scientific information by paper?
eLife clearly stands up among open-access crowd by caliber and financial commitments of their bakers. As far I see it is only open access publisher who doesn’t charge publication fees (thanks to its mostly non-US bakes). So even any rules were bended, it is probably justified in this case. Though I might agrees that PMC have should be more open about their backing eLife.

This list bears some examination. For instance, most of these are in markets where print is still strong, and many charge subscription fees for print. So, they may be OA online but subscription in print. These are also not mainstream journals by any stretch of the imagination. This is just a way to get to an impressive percentage, but another interpretation is that this is bad news for OA in the long run — that is, the vast majority cannot charge, so may be doomed to oblivion. Is that good news?

Fair enough, but I think the point still stands–eLife isn’t unique in this model, and more importantly, shouldn’t be given special status because they’re rich enough to cover everyone’s author fees. Inclusion in PMC shouldn’t be based on how much money you have.

One question this begs is whether a mission from the 19th century should be updated when something like the Internet becomes mainstream. Or do you just keep doing 19th century things (like collecting papers in an online repository)?

All OA research publications in the biomed field should be in PMC/EU-PMC, in my view. “OA publishers were the ones being mistreated.” No names, though, or quotes from those mistreated OA publishers. Why not? Have I missed them being mentioned in the post?

BTW, taxpayers’ money should of course go to publishers, preferably to for-profit ones, rather than to an infrastructure-like open scientific communication service like PMC. No?

I believe there was at least one angry tweet from the journal Biology Open, mentioning that they’d been forced to wait for months after they’d begun publishing on their own to get what eLife was given in advance.

Really? It’s a bloody shame. All OA journals worthy of being covered in PubMed (actually, all OA journals that are not obviously and demonstrably below par) should be immediately in PMC. Subscription journals with embargoes before they open up their articles can wait — after all, that’s what they like prospective readers to do.

Again, why the infatuation with PMC? It’s expensive and old-fashioned. Why make copies when you can link? Isn’t that what the Internet solved? It’s like a group of misinformed old men got together and thought up PMC based on pre-Internet computer systems. Hmmmm. Maybe that is what happened, now that I think about it . . .

It’s a tough call – on one hand you can argue that PubMed / PMC should not be giving credibility to bogus publications. On the other, even bogus publications may have attracted some decent content, and they shouldn’t be making life harder for new entrants to the market.

This is a question that becomes more pertinent as we see so-called predatory publishers try to enter the market. Where do you draw the line?

It’s easier to trust new journals from existing publishers / known entities than from completely new organisations. But it does seem that PMC needs to be fairer and less conservative in it’s acceptance of new journals.

But this isn’t the problem that was addressed here, in fact, quite the opposite. We saw favoritism and a faster than usual approval process given to an entity that had never published a single article, while at the same time you had established, reputable publishers like the Company of Biologists being forced to wait for months and months. Oxford University Press has been publishing for over 500 years, yet we must wait for entrance into PMC until the journal actually exists and has published something like 15 articles on its own platform.

I don’t think that’s necessarily an unreasonable policy. What’s unreasonable though, is that PMC and the NLM in general seem to have a lot of unclearly defined policies that are enforced in a somewhat capricious manner. I think what most publishers (and likely most readers) would like to see is a clear set of policies enforced consistently. Clearly state what the requirements are for inclusion in PMC and make everyone who wants in meet those requirements. Do things as quickly as possible, but do them on a first-come, first-served basis rather than giving special favors to cronies.

Similarly, the NLM has a stated set of rules for indexing in MedLine (so your articles appear in PubMed search results), but again, these are not consistently enforced and every publisher I’ve spoken to has had experiences that are all over the map as far as acceptance. I’ve seen established journals that have been around for 50 years and are ranked highly in their fields denied acceptance, while their direct competitors are accepted. No reasoning is generally given in a rejection letter, just the status of the decision. It’s also become increasingly confusing to see PMC articles from journals that haven’t met the high standards of MedLine acceptance listed equally alongside MedLine articles in search results. This has provided a back door for quickly getting your journals into PubMed search results while circumventing the review process.

I agree with the need to have clearly defined policies, that are fairly enforced, for the most part. On the other hand, I could also make the case for allowing some journals to be treated separately, as a pilot / trial for modifying the policies, and / or acknowledging that some new types of journals or publishers weren’t really envisioned when the policies were established, and that this is the development of a set of new policies to cover certain cases. That should all be done with a greater degree of transparency, mind.

Why would you put everything into something like PMC? It’s completely strange to create copies on the Internet. The point of the Internet was to create hyperlinks so you could point to source material and not have to make copies. Why make wasteful and redundant copies? The PMC architecture is from the 1950s. It’s pitiful. CHORUS is a modern and far cheaper alternative that accomplishes the same goals while saving the taxpayer money.

To your next point, OA publishers were the ones being mistreated. Some of the names are in the posts, if you’d bother to read them. Two of them are PeerJ and Frontiers. Other OA publications from larger publishers were also complaining, and some non-profit OA publishers as well.

The “taxpayer funded” saw has essentially backfired at this point. Taxpayers pay a lot of salaries at universities, but do you see tuitions dropping? Taxpayers pay for a lot of airport parking garages, but do you see the parking fees falling? What you’re missing is that taxpayers pay LESS in aggregate when there is a private source of funding to complement taxpayer funding — so they pay less for the parking garage than they would have because of bond initiatives, long-term parking fees for improvements, etc. In science, taxpayers pay some for research (and oddly, when they pay more to NIH, they seem to get less as far as publishable results, hmmmm), but tuition, fees, personal subscriptions, advertising, and licensing revenues provide publishers with the bulk of their revenues either through institutional subscriptions, personal subscriptions, advertising, or b2b deals. Without this multitude of inputs, taxpayers will pay more than they do now for the same product or an inferior one. Your solution to “taxpayer funded” is boiling down to taxpayers paying more — for PMC, for RCUK, for EU-PMC. You guys accuse everyone else of burdening the taxpayer, then do it yourselves very egregiously and without any offsets to blunt the costs on their behalf (CC-BY eliminates offsets through licensing, and PMC sucks traffic away from publishers, traffic they could turn into advertising revenues to further blunt the cost to taxpayers). It’s astounding that you do not see this.

“Why would you put everything into something like PMC? It’s completely strange to create copies on the Internet. The point of the Internet was to create hyperlinks so you could point to source material and not have to make copies.”

Does a research paper contain the full content of 30 other research papers? Or does it have 30 citations of identifying information that can be used to obtain the full text? Which of course takes time to look up…

The purpose of a hyperlink is to aid navigation, not to obviate the need to create copies of information. In fact, internet architecture is designed to incorporate copies (caches) of information, having multiple servers and even multiple servers in multiple locations to provide capacity and reduce latency. (Even peer-to-peer downloading has legitimate uses in providing resiliency and capacity through multiple copies.)

But it’s true that this is not necessarily obviously to the user / at the expense of the information’s identity.

However PMC does not really come at the cost of a paper’s identity – which should still either be cited via is bibliographic information, or electronically via the DOI assigned by the publisher.

“Why make wasteful and redundant copies?”

There are a number of reasons why they aren’t:

1) It is a resource that can survive, even if the journal / publisher doesn’t

2) It has a role in allowing new entrants into the market, to aid competition / provide innovation – largely by obviating fears about the first point.

3) It’s an additional resource, when the original is temporarily unavailable (net routing issues, server downtime, DDoS, etc.)

4) It has a role in standardizing markup of information, which can be beneficial for activities such as text-mining

5) It provides subject collection(s) of material across a number of sources, which can be beneficial for general users, as well as text-mining

“CHORUS is a modern and far cheaper alternative that accomplishes the same goals while saving the taxpayer money.”

Does CHORUS allow for full-text searching of research material, combined with Medline/PubMed indexing restrictions? You could search PubMed abstracts, and link to the CHORUS provided full-text, or possibly search the full-text across a number of resources without the additional PubMed indexing, but not to include both in the original search.

Having been involved with online search provision of the Medline database before PubMed was created, I have first hand experience of that effort not only aiding researchers, but front line health care too. The potential of PMC providing value in excess of CHORUS alone, not just to research but for actual provision of health care, is pretty clear.

And you state: “The budget for PMC is about $4.5 million”.

The annual, global, STM publising revenue is about $10 billion – half of which comes from the US. Without getting into the debate about what portion of that comes from the taxpayer (and it’s certainly not all of it), the budget for PMC is still a tiny proportion of even what is spent on publishing, let alone the $60+ billion the US spends on non-defence research.

“then do it yourselves very egregiously and without any offsets to blunt the costs on their behalf (CC-BY eliminates offsets through licensing, and PMC sucks traffic away from publishers, traffic they could turn into advertising revenues to further blunt the cost to taxpayers). It’s astounding that you do not see this.”

Advertising revenue is only a very small portion, particularly for the online component (and besides, what about a debate about whether certain taxpayer funded sites could contain some advertising to contribute to their budget?).

OA / CC-BY publishing – based on current publisher quoted prices – is already 25% – 75% cheaper than the estimates for average income for closed access articles (put at around $4000-$5000 in some cases). There is a lot of guess work and speculation, granted, but – and with increased competition around services instead of knowledge – there is every reason to believe that OA / CC-BY publishing will be cheaper for the taxpayer than purchasing access, even taking into account licensing!

But even if it isn’t, that’s not necessarily a problem. Take for example – http://www.jisc.ac.uk/reports/value-and-benefits-of-text-mining :

“For example, MGI predict that effective and creative use of these large data sets in the US health care sector could generate more than $300bn in value per annum and reduce national health care expenditures by around 8%.”

Now, I take it that that also means processing the raw data, not just text mining the research papers. However, there is also this (just considering the UK?):

“If text mining enabled just a 2% increase in productivity corresponding to only 45 minutes per academic per working week, this would imply over 4.7 million working hours and additional productivity worth between £123.5m and £156.8m in working time per year.”

The potential benefit of CC-BY publishing can easily outweigh the “offset through licensing”, and any potential increase in cost (if there was any – which there doesn’t appear to be).

Yes, it’s speculation, and it may not hold true to the wider effect on the economy (e.g. if others can ‘freeload’ from the CC-BY) material – which is far too complex a question to even try to address here – but there is a much bigger picture here, which needs to be looked at.

Reference linking is common on all good publishing platforms, and is used probably about 100 million times per year across the literature. It works. You don’t need to store papers in a central repository on the Internet. That’s why hyperlinking was designed.

You claim PMC doesn’t come at the cost of a paper’s identity. Read this: http://scholarlykitchen.sspnet.org/2013/10/08/how-the-nlm-justifies-linking-to-pubmed-central-versions-directly-from-pubmed-search-results-lists/ It shows that PMC designed its search interface and branding approach precisely to obscure journal branding and compete with publishers for traffic.

Apparently, you are ignorant about all the archiving good publishers (aka, most publishers) do now, archiving that makes PMC even less necessary. These services (LOCKSS, CLOCKSS, Portico) make PMC’s archiving function pale in comparison.

MEDLINE searching has been largely superseded by Google, and PubMed is no longer a major search engine referral for most biomedical journals. It’s been this way for years now. I think you’re being nostalgic, and I’m urging you to get with the times.

The rest of your supposed analysis is void of facts or logic. For instance, comparing PMC’s budget to the overall publishing economy makes no sense. Not only are your numbers estimating the size of the STM publishing economy way off (you’re undershooting), but the two are not related. The question about PMC’s spending is, “Is it necessary or simply redundant?” It’s been mostly shown to be redundant (migrating author manuscripts into PMC). Online advertising revenues provide millions to many publishers already, and are growing steadily. So, you’re wrong/out of date there as well.

CC-BY licenses are the least popular licenses among authors.

OA publishing is “cheaper” because it offers less, is subsidized by the much, much larger subscription businesses, or a combination of the two. And it’s growing quickly, but on a very small base. It’s growth in 2013 in absolute dollars is a fraction of the lower-rate growth of the core subscription businesses. It’s really a speck on the economic landscape for most publishers.

So, you’re defending an approach that reduces the ability of the market to offset taxpayer burdens, a licensing approach authors detest, a storage approach that is expensive and unnecessary, and a search model that’s outdated and already ignored by most practical users, and you’re doing this all with poor data, outdated data, and a lack of understanding of the facts.

Wow.

The fact that a paper appears in PMC does not change it’s bibliographic record – so any citations in papers will use the journal name, along with article title, etc. Nor does it change the DOI, so any electronic citation / link should be to the DOI resolver, not directly to either a publisher OR PMC.

I’ve done some searches in PubMed, and I’m simply not seeing what you are claiming. Yes, the text “Free PMC Article” appears prominently in the search results, but the link (for both the title and the free article text) goes to the PubMed abstract page.

On the PubMed abstract page, the top right has equally prominent links to both the publisher site, and PMC – providing the publisher link out information has been supplied. The link at the very end of the abstract (which is often buried in the page) is arguable, but the presence of images from PMC article does provide a more prominent link to PMC. Clearly, there is no way they could provide that without linking to PMC – but are they right to even include that?

That there are links to PMC at all will take traffic away from the publisher site, and how it is presented may have changed since you published that article, but I find it hard to make the same degree of argument against what I have seen as you have done.

I’m well aware of LOCKSS, etc. but these are only a partial solution to providing access during temporary outages. PMC as a known (and discoverable) resource that people can go to when they find they are unable to access still has a role to play here. It’s not an alternative to LOCKSS, or made redundant by LOCKSS, but they serve slightly different, complementary purposes.

And I’m not advocating CC-BY licenses, merely arguing that their cost is not prohibitive, and actually provide benefit over not having liberal enough licenses. I’m open to debating the possibility of -NC licenses, providing they offer the balance of rights for research / scholarly use vs the capability to offset costs via commercial arrangements. As they currently stand, the Creative Commons -NC licenses do not necessarily do that, and it appears the benefits of using CC-BY outweigh using -NC licenses.

With regards to PMC, you are asking a very particular question, which is designed specifically give greater prominence to a particular answer. Much like the licensing, I prefer to ask what is the right thing to do, not what is necessary or not. If it has value to researchers (and health care practitioners, lets not forget – what price to you place on a life?) that exceeds it’s costs, then it shouldn’t be killed just for being technically redundant.

Of course, that doesn’t mean that it is above scrutiny of how it operates, or that the funding shouldn’t be looked at, so that the entire cost does not fall on the US taxpayer.

I think I’m beginning to understand why the state of so much of the US infrastructure is so deplorable. One of the biggest frustrations in science is the fragmented nature of the literature. That’s what PubMed is addressing, with a single source for all biomedical abstracts, and what PMC subsequently started to address with a single source for all open access full-text (EU-PMC even with semantic enrichments that are most useful to scientists – in fact, I always advise people to use that instead of PMC). I’d have to agree that had the publishers collectively come up with something like CHORUS some 15 years earlier, the need for PMC might not have arisen. Private enterprise (and government initiatives) can go both ways: BioMed Central was the first open access publisher, private and for-profit, without any tax advantages or government subvention, and CHORUS a Johnny-come-lately, and until proven otherwise expected by many to be too little, too late, probably also as a result of it being a private enterprise initiative, the model that you seem to elevate to a panacaea.

PS. I did read the post; no mention of PeerJ or Frontiers.
PPS. If you’d like to make better use of your tax money, consider having an aircraft carrier or a couple of stealth bombers less and you’d be able to finance all the world’s scientific literature for quite a while. What a contribution to the world that would be!
PPPS. If reducing weaponry is not your thing, you, as US taxpayer, may consider cutting the annual $12billion+ fossil fuel subsidies and spend some of that on the advancement of science and medicine via open knowledge sharing. I’m not holding my breath.

You didn’t read the posts covering this apparently. Were they too fragmented for you? All I had to do was use a search engine. Those are things on the Internet that help you find information located in disparate locations.

Frontiers comes forward to thank me for blowing the lid off PMC’s erratic handling of applications here: http://scholarlykitchen.sspnet.org/2013/02/12/dont-shoot-the-messenger-keeping-our-eye-on-the-real-meaning-of-the-elife-pubmed-central-scandal/

PeerJ’s inquiries about why eLife got special treatment (and showing eLife and PMC trying to mislead Peter Binfield) are here: http://scholarlykitchen.sspnet.org/2013/02/07/why-were-pubmed-central-and-elife-discussing-peerj/

I agree that now that CHORUS is here, PMC is no longer necessary.

I’d like to add a comment as a small open access publisher who have been included by PMC. It took quite a long time before PMC would accept us, and although this meant a lot of work on our side, the process was completely fair and forced us to make improvements to our journal which have benefitted both our authors and readers enormously. After acceptance into PMC, our submission rate rose by over 500%, and the (free) service they have provided to us continues to be highly professional.

The benefits for us have been:

Significant rise in submissions due to the perception among authors that only high quality journals are indexed by PubMed, as well as their (correct) assumption that far more people will read their work than if it’s only published on the journal website.

Improvement in production quality of articles.

PMC provide us with stats on how many times each of our articles have been read on their site, which we incorporate with our own stats.

As our journal is small and still relatively unknown, the far greater visibility that our articles receive through being included on PMC has been a great bonus.

Significant rise in submissions due to the perception among authors that only high quality journals are indexed by PubMed

According to the NLM website, your journal has not been accepted by MedLine for indexing, so the impression given to authors that it has is deceptive. Acceptance into PMC offers a backdoor into PubMed search results without having to pass the rigorous standards of MedLine. Clearly readers and authors do not differentiate between these two levels of quality standards.

You are correct, we are not currently in Medline and will be applying this year – however, PMC do have rigorous standards for accepting journals for indexing, they rejected us a couple of times because we did not meet these standards and gave us a lot of helpful advice on how to bring the articles up to the level they expect. I see this as a very useful service which is of particular help to newer open access journals and which can only lead to the improvement of journal publication standards as a whole.

I’m not in agreement that PMC has “rigorous standards” – in fact, PMC doesn’t publish any objective criteria beyond quoting the vague guidelines of the NLM Collections Manual. This past summer, after 10+ years of operation, Dr. Lipman suggested PMC may start developing criteria for inclusion. The Medline online application is far different than submitting a URL and account information to PMC contractors for review. I am in complete agreement on the “bonus” to being included in these indices.

I’m not suggesting they’ve done anything improper.

The problem lies with the NLM. They have two systems (Medline and PubMed Central) that have tremendously different standards for inclusion. Yet the place where nearly all interact with those systems (PubMed) merges both with no indication of which standard a given journal has passed.

There is in general an enormous amount of confusion among researchers and most don’t know the subtle differences–it’s all just PubMed.

So most readers who see a journal’s papers in PubMed search results think it has passed Medline screening, and as noted in the first comment in this thread, assume the journal is of high quality. This backdoor effectively subverts the entire Medline screening process.

I do, however know of journals that have deliberately used this process, depositing articles in PMC, to give the impression of Medline approval, as a way of getting around both the requirements and the often long wait involved.

Thanks for clarifying. I agree there is a fair amount of confusion among authors and deception possible – primarily because of the way NLM has the search results and databases set up.

Yes, I agree that the difference between PMC, PubMed and Medline is confusing, and, as T Anthony Howell states, nowhere on our journal website do we indicate that we are indexed by Medline.

However, the fact remains that PMC do have publication standards that have to be met (which as you say are not as rigorous as Medline) before a journal is accepted for inclusion, so authors and readers are partially correct when they assume inclusion in PMC reflects a journal’s quality standards. Otherwise why has there been such a fuss about eLife allegedly getting into PMC “via the back door”, if its so easy for anyone to get in?

There has been a fuss because even with the reduced standards of PMC (as compared to Medline), some journals were allowed to bypass them entirely and were accepted before they even existed. As you state in your first comment, you had do a lot of work and spend “quite a long time” before PMC would accept you. Is it fair that others, who are better connected to PMC, don’t have to make those same efforts? Why should other OA journals get to jump the line ahead of you?

PMC were right to initially reject us – our articles needed improvement. As a new OA publisher using a unusual business model at the time (7 years ago), we were greatful for all the help and advice that PMC gave us, as well as other associations such as OASPA. I have no doubt that the production standards of eLife articles were already up to scratch, although I think the process could have been more open.

Also, I’d just like to clarify that we do not charge APCs – we aim to provide a free service to authors who may not be able to afford to publish in other OA journals. Our top viewed article in 2013 was read over 4000 times on PMC – that’s a brilliant outcome for the authors of that article, and as we publish in the field of cancer, it’s obviously beneficial to the medical community to be able to find and read articles of importance, a service that PMC provides to them for free.

Katy – eLife wasn’t even close to being “up to scratch” when it was accepted to PMC, it didn’t even have a website – and therefore was ushered to the front of the line by PMC leadership. eLife used US Government platforms to launch its publishing business, even without a single final paper for PMC to review. If you read some of Kent Anderson’s posts on eLife and PMC, the tale unfolds….Tony

I’m sure that’s true, but if you look at eLife articles in PMC they are of a high standard, both scientifically and in their production standards. These articles being on PMC benefits both their authors and readers greatly, which is what PMC is for.

Agreed, but that quality was unknown before those articles existed, which was when they were approved. And are those articles so much more important than the ones in your journal and other high quality OA journals that they deserved to jump the line ahead of you?

Some of the OA publishers who were not given the same treatment included presses that have been around for decades and decades, and who have proven and high editorial standards.

Also, if this was all so noble, why did PMC seek to hide it from not only the public but from their own advisory committee? Don’t forget the charade of June 2012, when Lipman and Patterson were deep in discussions about eLife getting launch assistance from PMC, yet the two of them basically conducted a ruse to tell the PMC advisory committee a story that hid these facts. And don’t forget the classic line from an NIH employee to eLife — “Now that we know that you love us, we’d like you to pretend that we’re complete strangers in public.” Or how PMC and eLife colluded to tell a story to PeerJ that hid the facts of what they’d done. Or how Lipman himself, when asked, either lied or stonewalled about his involvement or what was going on or both. Or how Randy Schekman threw PMC under the bus publicly when these revelations came out, saying eLife had nothing to do with it, that it was all PMC.

If this is all so noble, why have all these illustrious scientists been running away from what the record shows they did?

I understand there is a certain amount of “credibility” that can be presumed with publishers of proven quality – particularly with new ventures. However, this is not a private organization handing out VIP passes – PMC is the US Government handling a taxpayer funded platform with wide discretion and little oversight or objective criteria. Everyone knows – including PMC leadership – that acceptance to PubMed Central and / or MEDLINE are business necessities (noting the common example of the increase in your journal’s readership as a result). The actions with eLife, F1000, and subsidizing at least 2 online society journals (JMLA is one) and that can well afford to pay their own way, I see this action as unreasonable favoritism.

Here as some of the “primer” discussions:

3 Main Points discussion:

http://scholarlykitchen.sspnet.org/2012/10/29/pubmed-central-or-oa-central-more-strange-behaviors-at-pmc-and-nlm-paint-a-portrait-of-biases-and-poor-process/?relatedposts_exclude=34624

eLife, PLoS, BioMed Central discussion:

http://scholarlykitchen.sspnet.org/2013/10/15/pubmed-central-and-elife-new-documents-reveal-more-evidence-of-impropriety-and-bias/

JMLA discussion:

http://scholarlykitchen.sspnet.org/2013/10/30/should-the-journal-of-the-medical-library-association-jmla-stop-using-pmc-as-its-publishing-platform/

I’d rather not get involved with the criticism of eLife, as I said before, I just wanted to make the point that we, as a small OA publisher, don’t feel that PMC is “compet(ing) with publishers”. Our dealings with them have always been of benefit to us as a company. Our authors and readers are the most important factors for us, and many of them have told us that our articles being indexed in PMC is of great importance to them.

“as a small OA publisher, don’t feel that PMC is “compet(ing) with publishers”.”

Here is some reading to educate you about how PMC is competing with publishers for traffic and branding prominence:

http://scholarlykitchen.sspnet.org/2012/09/20/is-pubmed-central-complementing-or-competing-with-journal-publishers/

http://scholarlykitchen.sspnet.org/2013/04/04/pubmed-central-reduces-publisher-traffic-study-shows/

http://scholarlykitchen.sspnet.org/2012/11/12/what-pubmed-centrals-drag-on-publisher-traffic-could-mean-financially/

http://scholarlykitchen.sspnet.org/2013/10/08/how-the-nlm-justifies-linking-to-pubmed-central-versions-directly-from-pubmed-search-results-lists/

Long story short, PMC competes for traffic, which costs publishers money. For subscription publishers, it costs them because it reduces the traffic measured through COUNTER, and if the journal carries advertising, it reduces page views available for sale. OA publishers are also affected — in fact, PLOS’ traffic is decreased more than subscription publisher traffic, and costs them probably around $200,000 in lost ad revenue per year.

PMC also designed its search interface explicitly to steal traffic from publishers, as emails obtained through a Freedom of Information Act request showed. They also have suppressed journal brands through some of their initiatives.

The reason they get away with stealing traffic, dominating their supposedly neutral search results, and pushing publishers around generally is exactly the reason you state — publishers need and authors expect the validation that PubMed or MEDLINE inclusion brings. (They’ve also purposely created confusion between the two, and created a back-door into PubMed via PMC and adoption of OA.)

So, PMC is competing with publishers. And they get away with it because everybody’s afraid that pointing out these flaws and competitive aspects will threaten their indexing status.

Again, as a small OA publisher who does not charge APCs (and does not depend on advertising revenue), we do not feel that PMC is competing with us, but rather complementing our publishing service. We include the statistics that they give us along with our own statistics which we provide authors and readers with for article views. At no point in our four year relationship with them have we felt “pushed around”.

I don’t doubt that many other publishers feel aggrieved by PMC, as you say, but I’m sure there are other small OA publishers who have experienced similar benefits to us by being indexed in PMC. I suppose it comes down to this: who are the most important participants in scholarly publishing – the scientists who write and read the articles, or the publishers? The former are extremely well-served by PMC.

The most important participants in scholarly publishing are not the authors, but the readers. Authorship is just a mode of existence, not a person. Most authors are readers 99.9% of the time. And most readers want relevant, filtered content from sources they can trust. They want to read less, and more of things that matter. In that regard, PMC doesn’t do much for them. It just throws a whole bunch of journals together, increasingly driven by business model.

Your company is a UK company (http://www.ecancer.org), with some roots in Switzerland. As a US taxpayer, I’m so happy that my tax dollars are helping to validate your OA journal, and that my tax dollars are also helping to pay for about 25% of your bandwidth and infrastructure costs. So, you’re basically saying that you like PMC because it validates your journal and provides you with free bandwidth (aka, “complementing our publishing service.”).

So glad that US taxpayers can help another UK-based publisher keep the lights on. And who said that Revolutionary War thing was forever? Talk about taxation without representation. You get to tax us again now by utilizing US infrastructure without paying for it, and we have no voice in the RCUK debates at all. Lovely.

So what percentage of the world’s biomedical – or even just cancer – research is conducted by the US? More than 25%, right? Which means the US probably accounts for more than 25% of the readership of ecancer. What about the percentage of authors are from the US?

As a journal that doesn’t charge authors or readers, then if the US is only paying about 25% of the bandwidth / infrastructure (not even the total) costs, then maybe there is a case there to say you really aren’t pulling your weight, considering what you get out of it.

But maybe you can withdraw that support, and help ecancer turn the lights off. But then where are the US researchers going to access the already published research from, if they need it? Oh, yes…

As for using US infrastructure, I think you’ll find RCUK talks of using institutional repositories, and Europe PubMed Central.

Is it really worth getting territorial over the tiniest percentage of public research funds? The damage far exceeds any possible perceived gain.

The NLM has powerful and established brands – MEDLINE, PUBMED, PUBMED CENTRAL etc – they all are well established, predominantly because they were first and did a relatively good job of selecting, curating, indexing and providing a centralized searchable database. (Albeit Google Scholar is on the verge of eating their proverbial lunches.)

The concern I’m mostly interested in that was unearthed by Kent Anderson’s FOIA requests is the lack of objectivity to curate material for PMC, coupled with its position as a taxpayer funded entity – and then playing favorites by doling out “skip the line” cards to industry friends, ignoring conflicts of interest and established procedures and hiding the reality of its actions (“Friends in private, strangers in public”).

Access to these brands can make or break new businesses due to the Catch-22 nature.

That is simply unjustifiable and they are not being held accountable for it. Playing favorites with access to its brands by a taxpayer entity through circumventing objectivity is inappropriate.

So, if you’re suggesting that it’s better if people pay for the content they use, then I’m with you. I think it does cultivate better incentives overall.

If you’re unaware of the growing territorial aspects of taxpayer-funded research publication, then I hope this starts your awareness.

As for ecancer.org, it looks a bit slicker than you might think. While it is supported by some foundation money, it is a private entity (“The leading oncology channel”). It is a private limited company in the UK http://corporatedir.com/uk/company/04595666-cancer-intelligence-limited. So, why does a private UK company get to use US taxpayer-funded infrastructure to help it make its ends meet? It is not non-profit, and its OA journal looks like simply a way to drive traffic to these skeptical eyes.

“So, PMC is competing with publishers. And they get away with it because everybody’s afraid that pointing out these flaws and competitive aspects will threaten their indexing status.”

Kent – do you have any evidence for this? I see that your own journal is indexed in PMC, despite your continual public criticism of them:
http://www.ncbi.nlm.nih.gov/pmc/journals/?term=The%20Journal%20of%20Bone%20and%20Joint%20Surgery.%20American%20volume&titles=all&search=journals

“Which means the US probably accounts for more than 25% of the readership of ecancer.”
That is correct, currently 26.5% of our readership is from the US.

“What about the percentage of authors (that) are from the US?”
14% of our published articles are by US authors – if you average out a publication cost of $2000 per article, that’s $84,000 worth of publishing fees we have saved US authors so far, all provided by UK, European and Latin American funders.

We also offer a free translation service to Spanish-speaking authors (there are 38 million Americans who speak Spanish as their first language at home). Again, none of the funding for that has come from the US.

However, we do not see that as an issue at all, as Graham Triggs stated “Is it really worth getting territorial over the tiniest percentage of public research funds? The damage far exceeds any possible perceived gain.” We have published articles from 33 countries over the past 6 and half years, we don’t only publish articles from countries that have provided us with funding.

Lastly, to reply to your comments specifically regarding ecancer:

Cancer Intelligence is a private company, but ecancer itself is non-profit, it is supported by the ECMS foundation, a charity registered in Switzerland. We have the largest collection of self-generated oncology videos in the world, all completely open access, and have been publishing articles for free since 2007, again all those articles have always been open access. We also provide open access educational modules, our latest modules provide free training for African healthcare professionals in palliative care.

I have worked in publishing for 16 years (for large and small players) and truly believe in what we are trying to do at ecancer. We have an exemplary editorial board including many past presidents of all the major oncology societies and are committed to disseminating cancer information that ultimately benefits patients. And I think I’ll end on that note!

Lots of assumptions here. First, you’re assuming every paper you published would have been published in an author-pays OA journal. That’s not clear or certain. If the articles had been published in a traditional subscription-based journal, authors typically are charged nothing. Second, if you’re assuming that US taxpayers are in fact paying for ecancer traffic and that because percentages match up that any of that makes sense, you’re assuming that saving a UK-based company money is a goal of US taxpayers. Not the case. If US taxpayers were getting these free articles from ecancer’s own servers, then great. If not, US taxpayers are saving this UK-based company money. That’s a UK-based company exploiting US infrastructure. End of story.

The evidence of PMC competing is in the links I’ve put elsewhere and the posts I’ve published. So some homework, please. I can’t be your tutor. This is taking far too much of my time to educate you when a good search engine and your own elbow grease can be applied.

So, again, a private company with a non-profit branch benefiting from US taxpayer-funded infrastructure. We all have good missions we believe in. That’s not a differentiator, and it doesn’t excuse freeloading behavior, especially when your private company should be footing the bill, full stop.

Not so sure ecancer.org is just a branch of the Cancer Intelligence publisher.

The structure I could piece together is something like the swiss noprofit ECMS Foundation and some other non-US charities fund into ecancer.org – which is published by the for-profit CancerIntelligence that lists its business purpose as web portals.

Considering CI is operating with large staff and leadership overlap with ecancer.org, seems like the lines between for and not-for profit are blurred – although financial information is not available about non-profits in privacy protected Switzerland like here in the US so confirmation is impossible.

(Additional indicia include CI isn’t listed on LinkedIn while ecancer.org has its own company profile as a writing company with 11-50 employees, there is practically no vapor trail for a company founded over 10 years ago, and on CI’s website the only project it runs is ecancer.org – its other silos of expertise are blank without example).

What is clear is that the ecancer OA journal in PMC drives traffic back to the other ecancer.org offerings, even if PMC siphons off some for its own purposes – its all very charitable (including the GSK sponsored advertising on its home page).

Cancer Intelligence ltd was founded by Prof Gordon McVie and is the publisher of ecancer.org. The ECMS foundation supports ecancer to help disseminate cancer information for free (all of our offerings are completely open access). If you want further information you can always contact me directly at katie@ecancer.org.

My final thought is that as publishers we should all be working together and supporting each other. We all have different goals and aims but they don’t have to be exclusive.

Comments are closed.