The resource began as an Excel spreadsheet with three fields: title, uniform resource locator (URL), and year of publication. To search the spreadsheet, users were limited to the “Find” function offered in Microsoft products. Eventually, the resource became Web-based and received its permanent URL in 2000. With its growing content and increase in use, the developers decided to make the resource available through subscription in 2002 to generate revenue to improve the resource. However, to keep with the spirit of accessibility and in turn to ease users into subscription-based access, resource users were allowed to search five times free per week. On September 1, 2006, the resource was relaunched as a free online resource [1].

One of the nice features of the TRIP database is that results are displayed in categories based on Haynes' work on the 4S approach to current best evidence of studies, syntheses, synopses, and systems [2] (described on the TRIP Website). The TRIP categories are: Evidence-based Synopses; Clinical Questions; Systematic Reviews; Guidelines (North American, Europe, Other); Core Primary Research; eTextbooks; Clinical Calculators; and core general medical journals retrieved from MEDLINE (PubMed). Each of these cate-gories searches a number of re-sources to produce the results. For example, the Evidence-based Synopses searches Bandolier, BestBets, POEMs, Clinical Evidence, Evidence-based Complementary and Alternative Medicine, and many more.

To further understand the Website and resource inclusion process, the reviewer read through the About Us pages of the Website. The Websites and resources included in the TRIP database are systematically selected using a site evaluation tool. TRIP works closely with the Department of Family Medicine at Laval University, Quebec, Canada, in this process. The department has put together a Directory of Clinical Information Websites, which includes over 100 critically appraised Websites that offer systematic reviews, clinical practice guidelines, and critically appraised topics. TRIP looks at these Websites and evaluates them using an in-house evaluation tool and a team of information experts and clinicians. Once a Website has passed the test, it will become part of the “grabbing” process.

The “grabbing” process begins with identifying the clinically relevant material in a preapproved Website. Then the title, URL, and date of publication is extracted and entered into the TRIP database. The TRIP spidering software is then used to visit the URLs entered into the database and grabs the content of that particular page. It is then processed and ready to be searched in the TRIP database. The database is updated monthly.

The reviewer found searchingthe database relatively easy and straightforward. TRIP states that it takes unknown search terms and matches them against a commonly misspelled word database to try to retrieve matches. TRIP also searches for any keywords and/or synonyms matching valid search terms entered. The database can be searched using Boolean operators as well as natural language searching. The reviewer took a question regarding oral rehydration versus intravenous therapy for gastroenteritis in children aged two to five years and tested it using a few different approaches.

The first attempt was to enter search terms into the general search box on the home page combing the search terms with a Boolean “AND”: “gastroenteritis AND oral rehydration therapy AND intravenous therapy AND children.” The search yielded forty-two EBM-filtered results and thirty-two MEDLINE (PubMed) search results. At first glance, the forty-two results appeared relevant. Some seemed irrelevant, but it was a manageable number to look through and the pre-identified systematic reviews, guidelines, and excerpts from electronic textbooks appeared.

The PubMed search that TRIP does automatically retrieved thirty-two articles. The exact same search was run in PubMed and retrieved seventy results. The reviewer wondered why there would be a difference in the number of results and found the answer to this question in TRIP's About Us pages. When TRIP searches PubMed, it uses the clinical queries search filters offered in PubMed that limit the search results to a specific clinical query category (therapy, diagnosis, prognosis, and etiology). This made sense and was consistent with the categorization by question type for users to quickly identify what type of article they require based on the type of question asked.

Another PubMed feature allows users to further focus a search in PubMed by selecting a specific specialty. It is also noted in the TRIP searching pages that the database goes beyond the filters and includes articles from the top five general internal medical journals (New England Journal of Medicine, JAMA, Lancet, BMJ, and the Annals of Internal Medicine) for the past five years, as well as the BMJ Updates.

The second approach to searching the TRIP database involved using the natural language offered by the database. When the reviewer viewed the results, a sentence appeared below the search box: “Do you mean: gastroenteritis oral rehydration therapy intravenous therapy children?” This option was selected to see if it would make any difference, but it did not change the number of results. This is a nice feature for those not familiar with or confident in using Boolean operators for searching. However, the reviewer does not recommend relying on this method of searching. The reviewer entered a question phrase regarding the original example search into the search box and only received a few hits. The results were not relevant and missed the systematic review and current guidelines.

The next step was to rerun the original search string and note whether it was searched in the “title” or the “title and text” of the documents. The search was run again, and, when the results appeared, the reviewer clicked on “Advanced Search.” The screen showed the search string had been run in the “title and text.” When “title” was selected, the search was considerably limited and only retrieved one clinical question, but the same thirty-two results were retrieved from PubMed. The reviewer felt that it would be best to test this approach again with another search to determine whether this was a true representation of the “title” versus “title and text” feature.

The TRIP search page describes the search algorithm used to score the content. It states that TRIP scores content based on three main variables. The first is year of publication, so current material receives a higher score. The second variable is term position/density. This means that the algorithm scores material based on the position of the term in the document (term in the title is given a higher score than if it is in the text). It also scores documents on the number of times a term appears in the document (number of appearances of the term divided by the total number of terms). The third variable is the publication. When a Website or resource is evaluated at the beginning of the process, it is also given a score. The scores are based on clinical usefulness and methodological quality.

The reviewer explored the advanced search feature as well by trying to search each concept separately and combining them using “AND” but only retrieved one result. After reading the Search Help Page, the reviewer revised the strategy based on the suggested methodology. TRIP suggests that the best way to combine search terms is to search the “title” for the condition and then search “title and text” for the intervention.

After users have done this, they should combine the search sets using Boolean operators. This approach received the same forty-two EBM filtered results that the original search yielded. However, it did not retrieve any PubMed search results, which is puzzling. The reviewer assumes that the use of numbers to combine the search is not recognized in the PubMed database because it could not read the search history in the TRIP database.

My Trip is a current awareness– type tool. It allows users to register and select keywords that are automatically searched when the database is updated. If material is retrieved, the results are automatically sent to users by email.

Two important features of the TRIP database are its Medical Images and Patient Information Leaflet options. When users run a search in TRIP, it automatically retrieves any available images and patient information leaflets. This is a fantastic feature. The original search did not retrieve any medical images but did retrieve some great patient information that would be very useful for clinicians to provide to their patients. Some of the resources that appeared with this search were from NHS Direct, eMedicine, and the Merk Manual. More about what resources are searched for Patient Information and Medical Images can be found in the lists under the About Us pages. A few consumer health information resources could be added to the list, but the reviewer applauds TRIP for incorporating patient information features into the database.

Advertising can be found throughout TRIP's Web pages. While advertising is frowned on, TRIP makes it clear on their Website that advertising is simply a way for the company to gain revenue as they no longer rely on subscription revenue. A team of clinicians, consumers, and staff review sponsored advertisements. TRIP also states that they are not by any means influenced by sponsored advertisers and/or pharmaceutical companies. This disclaimer can reassure users that they are entering a resource that is not biased and truly does abide by evidence-based principles.

Because TRIP is moving toward free access, the reviewer wondered what might be next for it. In September, Dean Giustini, reference librarian, Biomedical Branch Library, University of British Columbia, interviewed Jon Brassey, one of TRIP's cofounders. He asked about the move to free access and plans for TRIP. Brassey answered with a few ideas of interest. He discussed becoming compatible for personal digital assistants (PDAs), a related articles feature, and an “ask us a question” option. Although these are still only ideas, it is nice to see TRIP's keen interest in expanding this evolving resource [3].