This issue of the Journal (page 575) features the last article of the EBM in action series,1 conceived to show how clinicians can effectively look for the best available evidence to answer clinical questions. In the current medical climate, clinicians clearly need systems to obtain the best available evidence, and the responsibility for creating these systems falls on both individual clinicians and the organisations for which they work.2 The EBM in action series was based on services that delegate the responsibility of individual clinicians to others. Some prefer this sort of service.3 But these services are few, and expensive. So what is the duty of clinicians to find evidence?

This begs another question: what methods exist to provide evidence for clinicians? The ideal method would deliver the evidence rigorously and fast. But these objectives conflict: fast methods are less rigorous, and rigorous ones less fast. We have to choose the most suitable compromise for each clinical situation. Methods available to all clinicians range from using the expertise of huge enterprises such as the Cochrane Collaboration (which in 10 years has managed to systematically review 2500 clinical questions — only about 10% of the task it set itself) to clinicians answering questions themselves (see Box). Other options include specialist secondary journals, such as Clinical Evidence (www.clinicalevidence.com), which have collected the best evidence for many clinical questions and other non-Cochrane systematic reviews. Predominantly, these limit themselves to intervention questions, leaving questions of aetiology, prognosis and diagnosis aside. All of these methods actually use the same principles: (i) turn the clinical conundrum into an "answerable" question; (ii) search the literature; (iii) evaluate the quality of the evidence found; and (iv) decide how this information might be applied to the current clinical conundrum.4 What differs is the speed and rigour of the results.

There are other factors to take into account for a clinician about to select one of these sources. For "mainstream" questions, there is more likely to be an answer already available through either a Cochrane review (now available free online in Australia at www.update-software.com) or a specialist secondary journal. Finding the evidence from one of the prepackaged sources itself represents a search — which may or may not yield a result and is therefore still somewhat risky for a busy clinician. Searching the literature oneself requires skill and expertise, and not everyone is prepared to invest time to acquire these.3

Before we make disapproving tut-tuttings, it is worth remembering that the medical profession is generally highly specialised. We expect to refer many matters that are outside the mainstream of our expertise to individuals or institutions better equipped. Shouldn't the same be true for finding evidence? We all know the principles of how to take an x-ray or measure biomedical parameters, but few of us actually do these tasks ourselves or even interpret the findings. Shouldn't clinicians wanting evidence-based answers have access to evidence services that might improve the care of their patients?

To meet this challenge, we supplied such a service. This was an interesting venture for several reasons: first, to see if it was feasible (could doctors order such questions, could we answer them, and was it worthwhile?); and second, to look at the sort of questions that doctors ask. Some of these were interesting enough to publish for others to read, and this is what gave rise to the EBM in action series.

There was a side effect that we did not anticipate. Content experts often disagreed with the evidence that we found — a collision between the findings of evidence expertise and content expertise. This often spilled over into the columns of the Journal's Letters to the Editor, generating about two letters for each EBM in action article.

To some extent, it is the need for speed in obtaining information that erodes rigour. Clinical information services represent just one of the compromise solutions (see Box).5 They may have a special place among other sources of evidence because they allow very individual questions to be asked. They also allow clinicians to "check" on questions they have themselves attempted to answer, successfully or not.

In the EBM in action series, we focused more on the process than the outcome, but it was the outcome that generated strong reactions from readers. We think this shows that clinicians are more familiar with the outcomes we described than the process by which we obtained them. We hoped that such "worked examples" would be helpful for clinicians learning and experimenting with undertaking their own searches.

Perhaps the most important message to be learned is not just some of the technical aspects — for example, breaking the question into its "PICO" (population/patient, index variable/intervention, comparison/control and outcome) components; selecting keywords and combining them effectively; understanding the importance of deciding what type of study might give the best answer; and interpreting the information in a manner relevant to the original clinical problem. Rather, it is that these skills do not come naturally to clinicians. Like any other clinical skill, they have to be learned and practised.