Evidence Based Medicine

Step 1 - Asking the well-built clinical question.

In research, the scientist asks a very specific question and tests a specific hypothesis. Broad questions are usually broken into smaller, testable hypotheses or questions. Clinical medicine is no different than research. The well-built clinical question is like the well-built research question.

Early in our training, our questions are usually broad. What is the pathophysiology of asthma? What is the pharmacology of the drugs used to treat asthma? As you progress in your training and begin dealing with individual patients, your questions tend to become more specific. What is the best method to remove dust mites from the home? Are inhaled steroids better than inhaled cromolyn for asthma? To make it easier to search for the answers to our clinical questions, our questions must be specific and focused.

The well-built clinical question consists of three parts: target disorder, intervention, and outcome. This is shown in the table below.

An example:

A 36-year-old male presents with a one-month history of chest pain and shortness of breath with extreme exertion. There is no radiation of the pain, syncope, lightheadedness, diaphoresis or palpitations. The patient has no significant past medical history or surgeries. His family history is significant for a father who died of an acute Myocardial Infarction at the age of 46. Two older brothers have angina, and the patient smokes 1-2 packs per day. On physical examination, his BP is 150/100 with a pulse of 70 and respirations of 16. General appearance is that of a thin male, in no acute distress, and the remainder of the examination is unremarkable. A work-up including blood work, EKG, and stress testing shows no active coronary artery disease. However, his cholesterol and LDL are elevated and his HDL is low. You ask yourself how vigorously should you be in treating his abnormal lipids.

A well-built question for this patient may be "In a middle age male with 4 risk factors (gender, hypertension, smoking, family history) for coronary artery disease, does a statin agent lower the risk of developing symptomatic coronary artery disease (angina, MI or sudden death)?" The target disorder is the description of this middle age male with multiple risk factors for coronary artery disease. The intervention is the use of a statin agent to lower his LDL. In this case a comparison was not given. We could have compared diet with a statin. The outcomes of interest were sudden death, development of angina or myocardial infarction. In epidemiologic terms, this is a question of whether primary prevention is effective.

Another example

You are seeing a moderately obese, 56-year-old male with Type 2 diabetes with no complications of his diabetes. His blood sugars are controlled with oral agents. He has mildly elevated BP (158/94). He is reluctant to take more medication. He wants to know how much benefit he can expect from optimally controlling his blood pressure.

A well-built question for this patient may be "In a middle age male with diabetes mellitus, will optimally controlling his hypertension lower the risk of developing stroke, renal failure or myocardial infarction." The target disorder in this case is that of a middle-aged male with Type 2 diabetes and elevated BP. The intervention may be the use of life style changes (such as lower salt intake or stress reduction) or exercise or antihypertensives. You could be even more specific and ask compare the different types of interventions or compare specific classes of antihypertensives. The outcomes of interest to you and the patient would be the development of complications of hypertension.