The advantages of wellbuilt clinical questions

A well-formed clinical question has two strong advantages: it makes finding the evidence easier and forces the clinician to specify the patient populations to which the evidence can be generalised and the outcomes that are clinically important.2,3 A question like Question 4, 5 or 6 above will certainly lead to answers. However, obtaining the answer would require a considerable amount of time in searching and validating a vast amount of literature. Structuring the question as in Questions 1, 2 and 3 above would lead to more specific answers in considerably less time.

For example, consider the difference between searching Medline for Questions 2 and 4 above. Searching Medline to answer Question 4 using the search string "Herpes zoster and treatment" yields 3749 references, many of which are narrative review articles, bench research and case reports. Even limiting the search to randomised controlled clinical trials yields 211 references, many of which are poor-quality evidence. In contrast, searching Medline to answer Question 2 using the search string "Herpes zoster and (corticosteroid* or pred*) and (aciclovir or valiciclovir or famciclovir)" yields one reference that is a randomised controlled trial of the treatment of acute herpes zoster with aciclovir alone for 7 or 21 days and aciclovir plus prednisone for 7 or 21 days.4

What criteria might be used best to specify a question at the dermatology consultation? The answer to this question might vary according to patient attributes such as age, sex, past therapy and allergies. A question can be as only as good as the initial evaluation of the patient, which includes a detailed history and examination in order to obtain an accurate diagnosis. An exploration of which factors are important to the patient in terms of expectation of treatment outcome, willingness to put up with inconvenience of frequent medication, and tolerance of potential side-effects is also crucial at such an initial consultation. Specifying an outcome that means something to the dermatologist and patient is also important. For example, consider a 28-year-old man with psoriasis who is desperate for a remission of the visible plaques on his body because he is planning a once-in-a-lifetime holiday to the coast, where he wants to expose his skin whilst swimming. Finding trials that mention only 20-50% reduction in PASI scores as their sole outcome measure would be of little value. Of more interest to this man would be trials that specify the percentage of patients achieving complete remission after a course of therapy. Although a 30% reduction in PASI could be useful to another patient, it is simply a matter of choosing an outcome that seems clinically relevant to the patient in question.

Asking a clinically relevant and structured question is not as easy as it might appear, but it saves time and it is more likely to provide a useful answer than a vague unstructured one. Following the structure described in this section takes practice but increases the likelihood of ending up with a question that is a good base for the next steps in practicing evidence-based dermatology.

0 0

Post a comment