Module 2

MODULE 2: Using an electronic index, a guideline index, and a web search engine to retrieve information relevant to your clinical problem, compare and contrast your results. Which resources were useful/ not useful for your information retrieval task, and why? Identify some alternative strategies for retrieving relevant information - would context relevant information retrieval be useful?

As I once again date myself, I remember back to the time when I would sit in the library, pad and pencil in hand, combing through the hard bound volumes of MicroMedex searching for journals that matched my interest. Feverishly scribbling the information I needed I would then proceed to the “stacks” looking for the journal in print. If I was fortunate, the library would have the journal in their collection. Then, the decision…to read it all or spend money on copying the article. Now at the touch of keypad from the comfort of my own home I can access a wealth of information far exceeding what those old dusty volumes could ever hope to contain. We live in a marvelous time where the world is literally at our fingertips.

Recently, I had the pleasure of observing clinical practices in an intensive care unit in Taiwan. One noticeable difference which struck me immediately was the lack of sequential compression devices and the minimum use of low molecular weight heparin to prevent deep vein thrombosis. With a DVT rate similar, or less than incidence of DVT compared to that in the US, the question begged to be asked….what are the benefits of these two therapies in the prevention of DVT?
Three major sources of information retrieval were used to search the answer to this question. Each of the information sources comes with advantages and disadvantages. For example, when using Google or Yahoo, a general search using the search term “DVT prevention” will net the reader over 75,000 results with one click of a button. While this certainly is a wealth of information, it would take months to shift through the important pieces separating them from the irrelevant pieces of information. If the topic one searches for is not highlighted in the titles within the first few pages of the search the researcher will often abandon using this strategy. Both Google and Yahoo searches do allow for “advanced search” strategies but they are not specific as those found in PubMed. For example, the most you can limit a search strategy is by using “all of the words, the exact phrase, any of the words or none of the words”. These can be applied to any part of the page, in the title of the page or in the URL of the page. Using this strategy limited the results to still a non-manageable number of 9000. One can also distinguish between .com and .gov domains. Google also allows a “wildcard” feature which essentially tells the system to look for any unknown terms and then finds a best match when you use a * as a placeholder. Beyond these strategies the advanced search mode is not very useful. The credibility of the information obtained from Google or Yahoo must also be evaluated by the investigator, whereas electronic indexes and guideline indexes have typically undergone more rigorous peer review processes.

Electronic indexes such as PubMed and MEDLINE allow for what Poynton (2003) calls high-precision searches directed toward the minimization of irrelevant and less useful citations. Bibliographic databases such as PubMed have more advantages over generalized search engines like Yahoo and Google in that they offer more selective search filters, e-mailing of search updates and storage of search histories and records. I found PubMed, MEDLINE and UpToDate very valuable resources when searching for EBM. The articles come from peer reviewed journals and have “limits” or “filters” that help one conduct a high-precision search.

When I entered my query into the National Guideline Clearinghouse (NGC), the search tools were clear and easy to navigate but led me to guidelines that were too specific and did not really address the clinical question. Instead of examples of RCTs to support clinical practice, the NGC is more devoted to guidelines and practices. I found it difficult to identify the studies upon which these guidelines are based. While each of the above strategies have value, use of the systems will be based on what question is being asked, how much time one has to devote to the exploration of the answer and how thorough one needs to be in assuming they have the answer to their question.

Time is a valuable commodity when seeing patients. Maviglia, Yoon, Bates and Kuperman (2006) note that if a provider has to spend more than 30 seconds trying to find the answer to a clinical question, the information will most likely not be adopted into practice. Sources of information need to be readily available to the provider instead of conducting hours of online searching. Context sensitive information retrieval may be one approach to help solve this challenge. As with any great endeavor, a well thought out plan can be advantageous. One alternative strategy is outlined by Poynton (2003), where she counsels the researcher to strategize with paper and pencil, outlining the topic needed and orienting yourself to what it is you exactly want. It is easy to click buttons one after another, but like any road map you have to orient it first or you can easily get lost.