As I’ve discussed previously, much of the strangeness of the PubMed Health-NLM-Google affair arises because NLM doesn’t seem to have an appreciation of search engine optimization (SEO), and the value of being ranked in Google.

Another aspect of this misunderstanding that came out during the NLM presentation at the recent MLA annual meeting, is that NLM is frustrated (!) that PubMed Health pages are getting a top ranking in Google because they consider the new resource to be in an uncompleted, “pre-alpha” state. Bafflingly, they apparently didn’t anticipate that Google would find the PMH pages until they were “ready.”

To anyone in the dotcom world, getting a high ranking in Google is invaluable — The Ultimate Goal, The End of the Rainbow. To them, the idea that NLM would not be celebrating (!) a high Google ranking would be hard to fathom.

I realize that government websites like NLM don’t have the nimbleness of dotcom sites, so they have trouble adapting to unexpected happenings. But, still, it’s interesting, I think, to ask how a dotcom site would handle the current PubMed-Google situation …

What would a dotcom do if they had a new site that was under development, not ready to be used by the public, and it suddenly and unexpectedly started getting high rankings in Google? I think if this happened a dotcom would drop everything else and get the site in a finished state as quickly as possible. And while they were working on this, they would inform users about their progress in getting it finished.

NLM’s response to getting a top Google ranking has been very different — From all appearances, they have done nothing different at all because of the high ranking. They are working at a slow pace to implement the new site, on some pages, but they have done nothing to inform users about their progress, and when implementation will be completed.

Beyond NLM – Building Library Discoverability with Google & SEO

My point here is not to be hypercritical of NLM. It’s rather to use NLM as an example of a more general problem in libraries. As I’ve discussed before, I think libraries should be more aware of the effect of SEO and Google on how our users find our sites.

What’s unfortunate about NLM’s reaction to Google’s ranking of PMH is that it almost appears as if they really don’t care whether users find their PMH pages or not — The pages are certainly going to be found and used more if they get a high ranking in Google, so NLM should rejoice, instead of grousing about Google finding them.

So I say the same thing to libraries in general that I say to NLM — We have good stuff! Let’s help our users find it! Taking advantage of Google and the principles of SEO to help us do this doesn’t mean we’re “in it for the money” — It just means we want to make our resources more discoverable for our users!

Related articles:

Eric Rumsey is at: eric-rumsey AttSign uiowa dott edu and on Twitter @ericrumsey

PubMed Health (PMH) was launched early this year by the National Library of Medicine. As discussed in previous articles, NLM has said very little about this new resource, so I and other medical librarians were hoping that they would clear up some of the mystery surrounding it at the Medical Library Association annual meeting in May. In this article, I’ll report on what NLM sources said about PMH at MLA, which, unfortunately, was not very much.

There were two sessions at MLA where NLM had an opportunity to discuss PMH. The first one was the NLM Online Users’ Meeting, at 7 AM on Monday, which was attended by about 50-75 people. NLM staff presenting at this session were Loren Frant, who talked about Medline Plus, and David Gillikin, who talked about other NLM initiatives. Neither one of the presenters mentioned PMH (see the MLA blog for what was discussed). In the question/answer period afterward, I asked Gillikin about PMH. He acknowledged its existence, but said very little else. He said that NLM Deputy Director Betsy Humphries would talk more about it on Tuesday at the NLM Update “plenary” session.

After the NLM Users’ session I talked more to Gillikin, about NLM’s silence regarding the high Google ranking of PMH, and the controversial claim of NLM-Google collusion that has arisen from that. He indicated that the current version that’s available online is essentially a “beta version,” and that eventually PMH will have a strong emphasis on providing comparative effectiveness information for consumers and healthcare personnel. Gillikin expressed frustration that Google had given PMH pages a high ranking when it was not really in a completed state. When I pushed for him to say why NLM had not anticipated this, he said, frustratedly, “Google is a black box” — Indicating that apparently NLM has had no communication with Google about the high PMH rankings in Google searches. For the record — NLM Associate Director Sheldon Kotzin was also present at this session, although he was not a presenter. The only input he had on PMH was to confirm that there would be more information about it from Humphries at the Tuesday session.

So the stage was set for Humphries at the NLM Update on Tuesday. Although attendance at Monday’s early-bird session was small, word had gotten out, helped along by tweets, that Humphries would have more to say. Here’s the account by conference blogger Alison Aldrich on Humphries’ talk at the NLM Update:

Next came the moment many of us have been wondering about for a long time. What would NLM have to say about PubMed Health, this mysterious new site with such high prominence in Google Search results? In truth, they don’t have much to say… yet. We know its purpose is to provide health consumers with better access to systematic reviews and comparative effectiveness research. We also now know that Google released it in pre-alpha form long before NLM was ready for that to happen. [ER: See my comments on this last sentence below.]

Hopes for more information were thoroughly dashed, then — Humphries talked for less than a minute about PMH, repeating what had been said on Monday. Of the high Google ranking, she said it was a surprise for everyone at NLM, but she said nothing about why NLM has been so silent about this, or why they have not had more to say about the nature of PMH. I waited expectantly until the end of her presentation with the many questions I (and no doubt many others) have about this whole affair, but to no avail — Humphries and the other presenters took NO QUESTIONS!

The Pot calls the Kettle a Black Box?

How ironic that Gillikin called Google a Black Box when NLM itself is being so mysterious! Here was the perfect opportunity to explain their actions in the Google affair to a friendly audience, and they said nothing to answer the obvious questions:

  • Why did NLM release PubMed Health before it was ready for public use, in “pre-alpha stage”? The conference blog report says that “Google released it in pre-alpha form,” but it was not Google that “released it,” it was NLM.
  • There is certainly a precedent for Google putting NLM pages at the top of its ranking for health/disease related searches, with Google Health One Box, so why did NLM not think about the possibility of this happening with PubMed Health?
  • And finally, the most basic question (in two parts) — Does NLM care what the world thinks? Do they care that there is a blog article which is getting high rankings in Google that suggests that NLM is conspiring with the CIA? If they do care, why are they not saying something to clarify the situation? …
  • A subset of whether NLM cares — Do they care what MLA people on Twitter think? Twitter was heavily-used at MLA this year — I and several other conference attenders tweeted throughout the meeting about PubMed Health, with no response from NLM. It seems like it would be a good idea for NLM to have someone communicating on Twitter!

Since the events at MLA reported above happened in May, I’ve found that NLM does seem to be making some slow progress in adding comparative effectiveness information to PMH. It’s unclear how much this happened before MLA, and how much since. If it was happening before MLA, it’s surprising that NLM didn’t talk about it. If it has been happening more since MLA, maybe NLM was motivated by the widespread puzzlement about PMH expressed at MLA. Whichever the case may be, NLM still has a long way to go in clearing up the continuing mystery of PubMed Health.

Related articles:

Eric Rumsey is at: eric-rumsey AttSign uiowa dott edu and on Twitter @ericrumsey

There’s been much questioning in the medical library community about the nature of PubMed Health (PMH), the new source from the National Library of Medicine. When it first came out, in early 2011, it seemed to be pretty much a copy of the ADAM Health Encyclopedia that’s part of Medline Plus. Before the Medical Library Association annual meeting in May, NLM staff had said little to clarify how PMH would be different.

At the MLA meeting, NLM staff said that the unique feature of PMH would be the addition of Comparative Effectiveness Research (CER) information to help consumers make health decisions. They didn’t say that adding CER information has already started, and it was only after I got back from MLA that I saw that indeed it has. I tweeted to see if anyone else had noticed this, and got only one response — PMH watcher Alisha Miles replied, saying that “limited articles” in PMH have had CER links since it launched, and that NLM is now adding more (Thx, Alisha!). I did a small random sample of 26 PMH pages and found that 6 (23%) of them have CER links.

In this article, I’m providing screenshots of PMH pages that have CER data and that don’t. Below the screenshots, I have a few comments and data on the random sampling that I did.

Here’s the PMH page for Asthma, which does have CER links, under “Evaluating your options” (lower right):

For comparison, a screenshot the Asthma page when PMH first launched in Feb, that doesn’t have CER info, is in this article.

Here’s an example of a current PMH page that does not have CER links – Sleep Disorders:

A few quick observations:

It appears that, if a page has more than five CER links (like the Asthma page above), there’s a link for “See more” CER links — For Asthma, the “See more” page has 22 total links, with a URL that ends with this:
… term=Asthma%20AND%20subject_comparative_effectiveness[sb]
This indicates that the CER links are obtained by AND’ing together the subject (Asthma) and the CER subset. Searching the CER subset by itself <subject_comparative_effectiveness[sb]> gets 252 links. This seems like a fairly small number in comparison to the total number of subjects in PMH, so I suspect that there will be considerably more CER data added.

In some cases, the indexing for CER sources may be questionable — For example: The PMH page for Brain Surgery, under Evaluating your options, has 4 links. Two of them seem doubtfully relevant – Fact sheet: Period pains and Low back pain: Can massage help?

I did the random sample by looking at the first stand-alone link (not a cross reference) in the middle column for each letter of the alphabet of the PMH alpha index.

NO – Pages do NOT have CER links – 20 pages on June 8, 2011

YES – Pages DO have CER links – 6 pages on June 8, 2011

It’s encouraging to see that NLM is beginning to give PMH an identity separate from Medline Plus. As discussed in another article, however, I hope NLM will do more to communicate the timetable of this process, and other CER features that they plan to add to PMH.

Related articles:

Eric Rumsey is at: eric-rumsey AttSign uiowa dott edu and on Twitter @ericrumsey

In February, Google began giving prominent placement to articles in NLM’s PubMed Health. As I discussed in a previous article, NLM and Google have been strangely silent about announcing this new feature, with no discussion of it anywhere that I can find.

It’s especially surprising that Google hasn’t said anything about this because — coincidentally with the NLM boost — Google’s ranking system has been under attack recently, with charges that doctom sites (most notably JC Penney) have used Search Engine Optimization (SEO) tricks that cause Google to give high rankings to their product pages.

As I’ve discussed before, although using SEO techniques to get high rankings in Google is widely discussed in the doctom world, it’s an almost unknown subject to most librarians. This is unfortunate, because without a background understanding of SEO, the next step in the “NLM-CIA conspiracy” story seems completely bizarre …

As I discussed in my previous article, soon after Google began giving prominent ranking to NLM, Jeff Hamilton, who blogs about ADHD, raised questions in his short article PubMed Health Who?:

Where the heck did these guys come from? Try and Google “ADHD” and these guys are the #1 search result!! PubMed Health is a new online resource under development at [NLM-NCBI] … Hmmmm, CIA? Secret Government agency? How does an organization go from not being on the radar to the #1 search engine result for ADHD overnight? … You’ve heard stories about how much control and power the Government has over the Internet…….I wonder if this secret SEO organization would be interested in doing some site optimization for me …

As a member of the medical library community, it seems laughable that anyone would suggest underhanded dealings between NLM and Google, and in my previous article I described Hamilton’s idea as “hare-brained.” But thinking it over I realize that to a non-librarian who’s been reading about the recent Google-SEO controversy, Hamilton’s speculations seem more reasonable. As he says, it does indeed seem surprising that PubMed Health pages suddenly began appearing at the top of Google’s rankings, with no explanation from Google or NLM about why this is happening.

The story gets more meta-interesting because Google’s ranking of Hamilton’s SEO story itself becomes part of the story — If the article had stayed on his blog, it probably wouldn’t have gotten much attention from Google and hence the eyeballs of the world. But instead it got copied on the Psychology Today blog, and that brought it a high ranking (generally between #1 and #6 in the last two weeks) in a Google search for PubMed Health — So, let’s say you’re a health-information-seeking consumer who comes across a PubMed Health page in a Google search — You like the page, so you do some googling to find out more about PubMed Health — And what do you find? Hamilton’s NLM-CIA conspiracy article.

So what’s wrong here? Why is a standard resource by large government site like NLM not able to outrank a blogger’s speculations about its validity in a Google search? Normally, Google does a good job finding “the real thing,” the site itself. The problem, I think, is that there has been nothing for Google to link to for “PubMed Health” — It didn’t even have a home page until last week, when it was announced by NLM/NCBI in Twitter. And there still hasn’t been a press release or longer announcement by NLM or Google. If these sources existed, they and medical library bloggers discussions of them would soon dominate Google’s top ten, and leave wild NLM-CIA conspiracy speculations in the dust. I’d guess that sooner or later, NLM and/or Google will make some sort of announcement. But I’d predict that the longer they wait, the harder it will be to displace Hamilton’s article from its high ranking — In my experience, Google has a persistent memory, and it often holds on to links after they have been obsolesced by events.

I think this is an excellent example of why librarians should learn more about SEO — If people at NLM and in the wider medical library community were paying more attention to SEO, it would have been clear that the sudden appearance of a new resource from NLM at the top of Google searches needs to be explained.

Learning more about SEO — If you google for SEO be ready for a fire-hose of sites offering to help you get a Google ranking. You might want to start out with Wikipedia’s lengthy SEO article or a book on SEO in the Dummies guide series.

Eric Rumsey is at: eric-rumsey AttSign uiowa dott edu and on Twitter @ericrumsey

Earlier today I tweeted about blogger Jeff Hamilton’s hare-brained idea that the recently implemented top-of-the-search links to NLM’s PubMed Health is some kind of government-Google conspiracy. I tweeted about Hamilton’s tweet, which had a link to his article on his own blog — Ha Ha, funny, right?

The way I came across Hamilton’s article, however, gives it a bit more seriousness — I found it when I was searching in Google for pubmed health, as in the screenshot at left — The first 6 hits are links that are well-known to the medical library community. But #7 is Hamilton’s article — that seemed so laughable on his own blog — in PageRank-powered Psychology Today, which means Google takes it seriously!

The lesson here, I think, is that NLM needs to say something about PubMed Health! As I discussed in my earlier article on it, and as Nikki Dettmar has discussed, it’s very strange that PubMed Health has been launched and assumed automatic #1 rankings in Google searches with no announcement or discussion of any of it by NLM or Google — If it had been talked about, assuredly it would be reflected in the Google search results in the screenshot. Instead, as these results show, there’s sort of a “vacuum” of information about the whole situation — which is just waiting to be filled by “passing spectators” like Hamilton ;-)

For the record, I’m including a screenshot of Hamilton’s article in Psychology Today that’s linked in the Google search:

Eric Rumsey is at: eric-rumsey AttSign uiowa dott edu and on Twitter @ericrumsey

I just noticed last week that Google is now ranking pages from the National Library of Medicine’s PubMed Health encyclopedia at the top of search results for disease words (accompanied by an attention-getting red icon) as shown at left for asthma. Other disease examples putting NLM at the top of search results are diabetes, migraine, lupus, and chickenpox. So, Congratulations, NLM! — A great new opportunity to lead people to the wealth of information at your site!

There are questions though — I was surprised to find this prominent placement for NLM (which replaces less prominent placement in the Health OneBox group of links at the top of Google searches that I wrote about in 2009). There apparently has been no announcement of the change, either from NLM or from Google, as far as I can find in searching. Beyond that, I also can’t find that NLM has announced the launching of the PubMed Health (PMH) encyclopedia that’s linked from Google. It was mentioned as being in development in summer, 2010, but there’s been nothing since saying that it was completed and ready to use.

There are also questions about the PubMed Health pages that are linked from Google (first screenshot below). These are from the ADAM Health Encyclopedia, and the same content is also part of NLM’s MedlinePlus (second screenshot below) — Why is NLM maintaining two different versions of the same content? Also, the PMH page that’s linked from Google (in the first screenshot below) has no link to MedlinePlus (MLP). The MLP version of the ADAM content, on the other hand (in the second screenshot below), is tightly integrated into the wealth of other information in NLM’s flagship MLP resource.

Below is the PubMed Health page that’s linked at the top of the Google search for asthma. This page has no link to MedlinePlus, in contrast to the MLP version of ADAM content, in the second screenshot below.

Here’s the MLP page for asthma, that’s well-integrated into other resources in MLP. So, NLM, how about asking Google to link to MLP instead of PMH? Otherwise, if the Google link continues to go to PMH pages, make a link from those to MLP!

Eric Rumsey is at: eric-rumsey AttSign uiowa dott edu and on Twitter @ericrumsey