A POEM a week for the BMJ

A POEM is Patient-Oriented Evidence that Matters

From now the BMJ will publish every week a POEM—a summary of a valid piece of research that carries information that is important to patients and so to their doctors. Unfortunately most research does not provide information that matters to patients. The POEMs will be published beside Editor's Choice. POEM stands for Patient-Oriented Evidence that Matters, and the concept was developed by David Slawson and Allen Shaughnessy, academics in family practice from University of Virginia in the United States. 1 2

The concept has its origins in a formula developed by Slawson and Shaugnessy: where U=usefulness of the information to doctors, R=relevance of the information to doctors, V=validity of the information, and W=work to access the information. In words, the most useful information for doctors is information that is relevant to their practice, valid, and does not take too much work to access. After listening to a presentation by Maria Musoke, a researcher from Uganda, on the usefulness of information to rural health workers in Uganda I added “interactivity” to the top line of the equation.3 The information is still more useful if you can interact with the source and interrogate it.

The formula provides a test of the ways in which doctors look for information they need. Traditional journal articles, although usually valid, are rarely directly relevant to a practitioner and are hard work to read—and they cannot be interrogated, although rapid responses (electronic letters to the editor) provide a possible means of getting answers from authors. The usefulness of original articles might thus be categorised as low. Textbooks should be relevant, although it's disturbing how often they fail to provide an answer to a direct question, and are comparatively easy to access. Their validity is questionable because they are rarely based on a systematic review of the literature and are often out of date, and they cannot be interrogated. They are thus of medium usefulness. In contrast, expert colleagues will give a direct and relevant answer to a question, should be little work to access, and can be interrogated. They are thus a highly useful source of information, although sometimes the validity of their answers may be low—”the blind leading the blind.” The formula thus explains why doctors use colleagues most commonly to answer questions and journals least often.4

Doctors suffer from what Muir Gray, director of the National Electronic Library of Health, calls “the information paradox”: they are overwhelmed with information, many receiving their own weight in journals and newspapers ever month, and yet cannot find the information they need when they need it. At least two questions arise during the average consultation between a doctor and patient.4 Most of those questions can be answered but few are. When I asked a sample of doctors to give me the one adjective they associate with their information supply, 90% gave a negative answer—overwhelmed, crushed, despairing. More than half of doctors feel guilty that they don't read more. Information has negative connotations for doctors.

Doctors are in a “knowledge business” and yet have severe information problems. The electronic age allows the possibility of a solution,4 but it hasn't been found yet. POEMs are a step forward. The box shows the three criteria that POEMs have to meet. Very importantly they have to provide information that will matter to patients. Will they live or die? Will they feel sick? Will they have pain? Will they be able to do what they want to do? A great many studies in medical journals give information on mechanisms of disease, aetiology, prevalence, pathophysiology, and pharmacology—studies that may be important but don't matter to patients. Faced with far more material than they can ever hope to master doctors might find it useful to concentrate on the studies that provide evidence that will matter to patients. They will discover that it is a minority of studies.

POEMs are selected by searching the current issues of 100 journals looking for relevant studies, potential POEMs, which are then evaluated for validity. The valid POEMs are summarised, and the summary is then reviewed and revised. The service is provided by InfoRetriever, who have kindly allowed us to publish a POEM each week. Those who would like to subscribe to their full service should access their site at http://www.infopoems.com/index.cfm