The Search Principle blog

To scan or browse the medical literature for osteoarthritis, try a simple PubMed search, type some keywords into Google Scholar, review what UpToDate has to offer or glance at a practice guideline. What do these simple searches have in common? Beyond the fact that they are the most basic kind of searching clinicians can do, they are obviously not reviews of the medical literature – ergo, caveat searcher.

However, for most clinical questions, physicians do not always have time to do a proper lit review. When a lit review is not your goal, and when Google scholar is unhelpful, use PubMed’s clinical queries.

Sample clinical question:
In geriatric patients with osteoarthritis, what is the best course of treatment beyond cox-2 inhibitors to control pain and discomfort?

Academic, tertiary care hospitals are exciting, cutting-edge places. Health care is, for all its problems, a challenging, rewarding sector to work within – medical librarians can see a direct benefit they provide to clinicians, and patients. It’s measurable.

One aspect of working in health is dealing with the issues around disease and illness, and mortality. From a caregiver-family perspective, Robert Scoble, Microsoft’s Techno-Evangelist, has been blogging about the passing of his mother in recent weeks, and dealing with the health care system. Read some of his posts, they (and the responses) are illuminating reading for all librarians and physicians.

On the passing of his mother, Scoble muses on the meaning and questions of life:

“What do I want to do career wise? I’ve had opportunities thrown at me recently. Should I consider them? For what reason? What’s the value I add to the world? Can I do more? Are there things that I’d love to do more than taking a camcorder around Microsoft and representing Microsoft to developers? In five years, where do I want to be? Who do I want to be? What kind of father should I be?”

Important questions that deserve our attention. Have a good weekend. Dean

Physicians often treat and care for cancer patients, even if they are not oncologists. Cancer is a serious human health issue and a leading cause of death in Canada and U.S. Cancer requires special attention in finding the best evidence.

Some facts about searching in evidence-based oncology:

1. EBM Sources – many sources like Cochrane or Trip lead to best evidence.

2. MEDLINE contains all citations formerly in CancerLit. See the MeSH tree for neoplasms by site. Click this pre-formatted PubMed search on small-cell lung cancer to see how MeSH is used. For best evidence, try PubMed’s clinical queries.

3. U.S. National Cancer Institute – pre-formatted searches (excellent tool)

4. UpToDate – search UptoDate with a subscription, or try a Google browse.

5. BC Cancer Agency Treatment Information for Patients – one of the best oncology sites in Canada, or indeed anywhere. (Browse Google scholar if you must.)

For specific information, ask a BC cancer agency librarian for assistance or see this partial list of cancer libraries.

Postscript: see this patient-oriented pathfinder, by SLAIS student Megan Wiebe, that looks at oncology from several perspectives.

At Microsoft in April, the discussion of UI (user-interfaces) was front and centre. But search companies like Google, Microsoft, Yahoo and Ask need to consider more the underlying structures of their databases as the Web scales up in size. Google Health’s use of social tagging is an attempt to introduce organization, post-hoc. It’ll be interesting to see if end-user tagging will succeed, or not.

Free search engines have traditionally given end-users basic boolean keyword search functionality. As more researchers find themselves dependent on the Web, and as open access to content proliferates, a need to develop more advanced query capabilities will arise. Google Scholar should consider such advanced functionality for medicine and basic science, two subjects which comprise 50% of its database.

Without better organization, academic searching on the Web will remain unruly, frustrating, and little more than a browsing space. Performing a literature review will seem like entering a maze, with no way out. Without some major improvements, it’s simply a matter of time before we have a repeat of the Johns Hopkins’ tragedy.

Podcasting in clinical medicine is on the rise. The Krafty Librarian is presenting a poster on podcasting at this week’s MLA Conference. We presented “Podcasting: the future of staying current in medicine?” at the 2006 CHLA/ABSC Conference this week. (See our RSS & Podcast glossary for words that are new to you.)

Our literature review revealed a paucity of research in the clinical use of podcasts. PubMed has fewer than 10 articles as of May 19th, 2006. Here are some key articles and our top five podcast articles from the health literature. If you are interested in searching for podcasts, or creating your own, see our resources list.

Podcast searchers should use their favorite directories like Podcast Alley or search tools like PodScope. Thus far, there is no Google Podcast search.

Synergy (some say, tension) exists between the new openness on the Web in medicine, and access to the evidence. See the latest additions to my list of OAM Sources On the Web:

OPEN ACCESS TO EVIDENCE-BASED MEDICAL INFORMATION –
OPEN VS. CLOSED (i.e. LOCKED & SUBSCRIPTION-ONLY)

Let me know if you would like me to add anything. Dean

“If we could only sleep, sleep well… sleep with that perfect unconsciousness we experience on nights when we are thoroughly fatigued, we would sleep without dreams”
– Guy De Maupassant (1850-1893)

The 2006 CHLA/ABSC Conference wrapped up this morning, and Vancouver has passed the responsibility to Ottawa for 2007. As one of the program chairs, CE instructors, posters 1, 2, I am happy to say that we had a record numbers of attendees, the highest number of student librarians in attendance, and beautiful Vancouver weather. We have nothing but praise for the conference hotel, the Marriott Pinnacle.

Tonight, now that the pressure is off, I should sleep without dreams.

Diane Helmer (Thompson) and I gave our workshop today entitled “Finding the hard to finds: searching for grey literature” at the 2006 CHLA/ABSC Conference in Vancouver, Canada.

Here are the learning objects – powerpoints and manual and a synopsis of our discussion and brainstorming session at the end on Top Skills Needed to Find Grey Literature in Health:

1. Expertise in conducting exhaustive literature reviews in health, especially where grey literature is used to address bias (ie. clinical trials, systematic reviews, meta-analyses). Strong web skills.

2. Knowledge of publishing trends: 1) traditional publishing vs. 2) everything else, including self-published materials on the web, blogs, podcasts, wikis and other non-traditional, non-indexed materials.

3. Knowledge of sources of information beyond MEDLINE, EMBASE, CINAHL and directories, gateways, and portals. Know how to apply principles of evidence-based practice to assist with grey lit search strategies and checklists.

4. Knowledge of major U.S. and Canadian agencies that produce grey literature, information specialists at the Canadian Agency for Drug and Health Technologies, and the NYAM, Grey literature report.

5. Expertise in project management and librarian advocacy. Manage large amounts of information. Know trends in self-archiving and repositories (ie. PubMedCentral, BioMedCentral, institutional). Advocate for open access.

Finally, we looked at the possibility of my starting our very own Grey literature wiki for Canadian health librarians to share pathfinders, and checklists.

Stay tuned. Dean

Google’s annual press conference would have been raucous yesterday. I wish I had been invited. Post-facto, Sergey and Larry – do you realize how confusing Google Coop-Health is?

A disappointment. This is Web 2.0 alright, a portal concept that combines gadgets, tagging and channel surfing. But physicians don’t need gadgetry to practice medicine – they need evidence. Stick to your day job, Google. Leave medical classification and document analysis to librarians. Sure, have fun with social tagging and folksonomies – just don’t mislead physicians by implying anarchic tagging comes close to what they need to find things, and to make clinical decisions. This is not what I had in mind when I wrote about Google Medicine.

Despite a roster of U.S. health partners working with Google Inc. (including the National Library of Medicine), Google Health is “channel” searching among four main health domains: conditions, drugs, research, info-type; with up to five specific categories in each domain. Are any standards used to populate these categories? It’s one thing to use social tagging for automobiles, and stereos – but for health? Even consumers are advised to use more targeted search techniques and best content websites. Rather than rely on social tags to lead to trusted content, use good consumer sites in Canada here, or U.S. suggestions here.

Look for a review, complete with various tests, and comparisons next week.

In the meantime, Google Health gets a thumbs down. I won’t be recommending it to docs or patients at UBC or Vancouver hospital. For now, stick with PubMed, Scirus and plain Google, readers. Anurag, where does Google Scholar fit into this mess?

Top Five Reasons Why We Still Need Google Medicine

1. Medical information needs contextualization and personalization. Medical librarians gauge the scope and expertise level of information they provide, for example. Will Google Health provide a range of materials to address the many contexts needed for medicine like a true vortal/portal?

2. Is Google Health integrated with Google Scholar and regular Google? The more Google appliances we have, the more places searchers have to look. This creates problems in terms of findability and recall.

3. Will Google Health oversimplify and dumb down medical research? Consumers may find it helpful, but will doctors? Google stands to lose mindshare by diluting its business by over-verticalization.

4. Does it make sense to search across the *entire* Internet and all 50 billion pages for medical information, when the wheat should be searched apart from the chaff. From a librarian’s perspective this is ill-conceived and a waste of time. Information demands classification. Computer power cannot – does not – replace organization.

5. Findability in medicine improves markedly by creating specific channels for searching. (This is actually true across human knowledge.) Algorithms that lead to popular or oft-cited materials is only a small sliver of what’s relevant and valuable. The most popular items also introduce bias if overly emphasized.