The evidence based clinician: part 1, asking answerable questions

It's quite shocking to realise that there is an awful lot of rubbish behind a lot of clinical practice. Many doctors will swear that their choice of management is effective. "But," asks Christopher Ball, "where is the evidence?"

Clinicians suffer from information overload. Hundreds of medical articles and textbooks are published weekly, but most are of poor quality and already out of date. The internet contains opinion intermingled with scientific fact, usually with no clear distinction between the two. Experts cite selectively and may twist facts to suit their viewpoint.

Evidence based medicine offers a way of cutting through the garbage to good quality material, but many clinicians do not have time to learn critical appraisal and statistical analysis in detail. Fortunately numerous resources are now available that have done the hard work already, allowing clinicians to move rapidly from question to answer and improve the quality of care that they provide for their patients.

This series of articles will focus on how to ask answerable questions, where to look, and, finally, how to customise the information to your patient, without spending hours learning complex techniques or reading long winded papers.

Using evidence in your practice will make you better at diagnosing conditions, increase the chance that you start the right treatment and avoid harm, and help you better advise your patients on likely outcomes. Plenty of studies have shown that patients who receive evidence based care have better outcomes, and patients who don't have worse ones, so it is worth the effort. The goal is to improve your patients' care, so you won't find any statistics or research methodology mentioned, only practical strategies for frontline clinicians.

Let's get real

Most on-call shifts are so crazy that doctors have no time to think about improving their diagnostic or therapeutic skills. Surveys have shown that over 70% of house officers read nothing about medicine, and senior house officers read on average only 20 minutes a week. (If you hope consultants do better the answer is, not much. They read on average for only 45 minutes a week, and this falls with increasing age.) Combined with over three million medical articles published a year, this means that doctors soon get out of date. Dave Sackett, a leading expert on evidence based medicine, once reckoned that physicians would need to read 30 articles a day, 365 days a year to stay up to date in just internal medicine. He didn't--and no one else does either.

So what does this mean? Firstly, there is little point in reading standard journals--you are unlikely to find or read the important articles, and even if you do, they probably won't be relevant to your patients. Secondly, the only way to stay up to date is to use resources that have already selected the best research. Thirdly, with so little time you need to learn to ask questions that relate to your clinical practice and develop efficient searching strategies to hunt down the answer rapidly.

Asking answerable questions

This article focuses on asking questions. Sounds easy--after all, doctors ask hundreds every day--but we're interested in answerable clinical questions that matter to you and your patient.

Scott Richardson (see further reading list) has pointed out that there are two main types of questions that clinicans ask--background stuff typically focusing on the basics of a condition ("What's pancreatitis?" "Why does pancreatitis cause hypercalcaemia?") and foreground stuff concentrating on the practicalities of managing a patient with that condition ("If I give my patient with pancreatitis octreotide, will it stop her dying?", "In patients with suspected pancreatitis, what clinical features best diagnose it?").

Students enter medical school asking lots of background questions but quickly learn that life is easier if they stop questioning and simply accept everything they are told. Sound familiar? IQ levels are said to fall through medical training as curiosity is replaced by medical facts. Unfortunately this leads to two dangerous assumptions--that senior doctors always know the right answer and that admitting ignorance is a flaw. Consequently some trainees simply never learn to ask foreground questions.

The result is that many house officers know lots of medical science but little about smart clinical management such as which signs and symptoms best predict complications, or how to determine which patients will benefit most from a treatment. Fortunately, once you recognise that asking foreground questions is the best way to improve clinical skills and maintain knowledge, this deficit is easy to fix and can make patient care more fun.

Try an experiment--on your next ward round, note how often someone asks a foreground question about patient management. Once you ignore people asking questions, demonstrating how smart they are, or consultants grilling you on basic management you might be disappointed.

Next, identify how many assumptions the medical team makes on the ward round about treatment, diagnosis, or prognosis (some folk reckon clinicians need on average five bits of evidence per patient), and make a list. When I first tried this it felt weird, because I had become brainwashed into believing everything I was told. Just keep asking yourself "What's the evidence for that?" whenever the team makes a decision or your consultant offers a pearl of wisdom, and you will soon recognise how much guesswork is going on.

Back to the clinical scenario

The table shows some of the assumptions made in the clinical scenario, and I've used a technique to extract suitable foreground questions from each one. Sometimes called the PICO formula (see box), it is a useful way to construct focused clinical questions, makes searching more structured, and increases the chance of finding an answer.

Choosing the best questions

• If you make a list, you will quickly discover every ward round generates far more questions that anyone could hope to answer. So, which questions are worth answering? Although many clinicians delight in knowing more about incredibly rare diseases than their colleagues, I'd encourage you to concentrate on questions that
• Deal with common conditions. These usually have more evidence available, so you are more likely to find an answer
• Deal with emergencies
• You can answer within your limited time. In general, avoid poorly focused questions or ones dealing with rarities
• Keep arising. You will discover that over time you start asking similar questions for some conditions--start with these, since you are most likely to remember and apply the evidence back to your patient
• Could make a big difference to your patient
• Interest you the most

If you develop the skill of asking questions about your patient (and finding the answer), you will stay up to date and relevant for your patients, rather than depending on a postgraduate tutor or a journal editor to decide what you should learn. The next article will look at some efficient search strategies, where to find answers to your questions, and what to do if you get stuck.

Christopher Ball, project director, Centre for Evidence-based Medicine, Oxford OX3 7JX
Email: chris.ball@doctors.net.uk


studentBMJ 2002;10:303-352 September ISSN 0966-6494

Further reading

Richardson WS. Ask, and ye shall retrieve. Evidence Based Medicine 1998;3:100-1

Sackett DL, Straus SE, Richardson WS, Rosenberg W, Haynes RB. Evidence-based medicine: how to practice and teach EBM. London: Churchill-Livingstone, 2000.

Thanks to Dave Sackett, Kilgore Trout, and Bob Phillips who did lots of the hard work behind this.