OVERVIEW

Diagnostic Questions
- Radiologic-Pathologic Correlation
- Comparison of imaging methods in a specific clinical condition
- Details about one imaging method in a specific clinical condition

Interventional Questions
- Should a newer technique replace an older method? This is a benefit / harm analysis and should, ideally be from a study in which patients undergoing a newer technique have been compared with similar patients undergoing the older, alternative technique.
- Safety of a technique: EBM methods allow us to estimate if a study is large enough to prove that a technique is safe. Case series data can be used.
- Detail about the technical aspects of a technique in a specific clinical condition.

Why not just use Google?
For many people not used to searching the literature, it is very tempting just to type a few words into Google or (even better) Google Scholar and take advantage of the fact that Google generates useful links quickly. Analysing the value of Google Scholar to clinicians in 2005, Henderson stated "Google, however, emphasizes Web sites that are popular, as measured by the number of links from other sites, and does not weigh quality or date. Searchers are often frustrated by the large number of links Google will generate for common topics." To read his analysis, click here [1].

The Evidence Pyramid:
The evidence-based approach has been used much more widely to date in medicine and surgery than in radiology. In these specialities, it is common to find that someone else has already done the work by performing an evidence-based review that includes the material of interest. In EBM terms, this is a form of evidence-based review (‘secondary’ literature), to be distinguished from original published journal articles (‘primary’ literature). In Medicine, practical resources to support evidence-based health care decisions are rapidly evolving. This evolution is based on the concept of the ‘Evidence Pyramid', described by B. Haynes [2].’ The higher up the pyramid we go, the more reliable the information.

PRIMARY LITERATURE forms Level 4 (the base) of the pyramid. We commonly search this using PubMed or the 'Search by Clinical Study Category' section of B. Haynes Clinical Queries link from the PubMed sidebar.

SECONDARY LITERATURE forms the upper 3 levels and includes:

- Level 3: Evidence-based reviews. These are performed to strict methodologic criteria and provide more reliable information than traditional ‘expert (or 'narrative')’ reviews. The best of these, Systematic Reviews, are increasing in number in the medical literature. A quick search for systematic reviews can be performed using the 'Systematic Reviews' section of the Clinical Queries link from the PubMed sidebar. As well as reviews, secondary sources of information include Databases, Guidelines, Gateways and some Search Engines that filter evidence according to quality and relevance. If you would like to sample a few of the most popular medical subscription sites without leaving this page, have a look at The Cochrane Library, DynaMed, or TRIP; and try SumSearch, an example of an 'evidence-based' search engine.
Some organisations compile online databases of 'critically appraised topics' (CATs) produced by their clinical members. To see and search examples of these CATbanks, click here

- Level 2:Synopses: These combine systematic reviews and recent primary literature. They comprise online subscription-only journals like ‘Evidence Based Medicine’and the ‘ACP Journal Club’ which extract the best current internal medicine literature using impeccable methodology and present it with a one-line, usable title and contextual comment from an expert in the field.

- Level 1: Information Systems. These are at the apex of the pyramid. They integrate and summarize all relevant and important research about a clinical problem. For an example of one, see the 'evidence-based' online medical textbooks, Clinical Evidence and Up to Date.

You can build search strategies for Diagnostic and Interventional questions around these concepts. If you are interested in using handheld technology for literature searching, click here to go to the relevant page of the 'EBR in Practice' section of this website.

Summary

• The best search strategy varies with the question and the operator.
• The subscription sites listed here have advantages over the free sites. We recommend you explore them.
• If you can get training, do so. Librarians are the best resource. Getting a librarian to search your question after you have done so defines the current limits of your knowledge and strategy, identifying your ‘knowledge gaps’ about searching. EBM courses, held in Oxford and McMaster, include training in literature searching. Check their websites [3, 4] for further details.

References

1. Henderson J. (2005). Analysis: Google Scholar. A Source for Clinicians? Can Med Assoc J 172(12): 1549-50. [ link ].

3. Centre for Evidence Based Medicine. Oxford University. [ link ].

4. Health Information Research Unit (HIRU): Evidence-Based Health Informatics. McMaster University, Hamilton, Ontario, Canada. [ link ].