Emergency physicians need rapid access to the best current evidence on a wide range of clinical topics. This is the goal of BestBETs (best evidence topics).

BestBETs represents a modified version of critically appraised topics (CATs).¹ They originated as a local exercise at the Manchester Royal Infirmary, UK, and since July 1998 have been published as a regular feature in the Journal of Accident and Emergency Medicine (recently renamed Emergency Medicine Journal). The BestBETs web site was created in 2000 and includes several hundred BETs, many of which have not been published in the journal and some of which are still under construction. Access to the site is free.

The process used to produce BETs is clearly described on the web site. A 3 part question is formulated, an explicit search strategy is used (confined to Medline and the Best Evidence CD-ROM), and the retrieved evidence is appraised and summarised.² Studies that provide the highest available level of evidence to answer the particular question are included in the BET and are critically appraised using published check lists.^3–5 The finished BET is displayed in a standard format that reports the clinical scenario, 3 part question, search strategy, search outcome, main results, level of evidence, study weaknesses, and clinical bottom line. Each topic is written by one person and peer reviewed by a second. The peer reviews are generally completed by senior emergency medicine clinicians at Manchester Royal Infirmary who, although they are not clinical epidemiologists, seem to have considerable experience in teaching and practicing evidence-based medicine. The web site also allows readers to submit their own BET using a structured form.

The titles or clinical questions in the BestBETs database can be searched with words or phrases, or the database can be browsed full text. A colour code beside each BET indicates its state of completion and, most importantly, whether it has been peer reviewed. The site's producers state that BETs are rechecked and updated every 6–12 months. No current report on the site indicates that this is happening, but given the site's relative youth, it may be too soon to reasonably expect this.

I found the site easy to use. For example, I was interested in determining whether prophylactic antibiotics prevent infection in patients with dog bites, and within 80 seconds I was able to retrieve an answer (antibiotics prevent infection, number needed to treat of 14). Retrieval times that are this short make the web site suitable for real time clinical practice. Another strength of the site is the provision of search details, allowing interested (or sceptical) users to replicate the search. However, the clinical usefulness of the site is currently hampered by the lack of available evidence for many of the questions posed. These evidence gaps, although frustrating, serve to identify priorities for future research. In addition, the database is not comprehensive, but it is likely to become more so with the passage of time. And the authors might want to consider extending the searches beyond Medline and Best Evidence to include the Cochrane Library.

Although still in its infancy, BestBETs is a good resource for emergency physicians, and I recommend its use.

• Download figure
• Open in new tab
• Download powerpoint

Best BETs(www.bestbets.org)