Come along to watch some clinical teachers in action: 1. A hospital physician rounding with the team’s medical students examines a middle aged woman with upper extremity deep vein thrombosis. After the bedside visit, the attending asks aloud about the frequency of underlying diseases associated with this condition, admits aloud he doesn’t know the answer, and records the question concisely in his PDA.¹ After rounds, the attending invites the students along as he finds and appraises evidence about this topic and “thinks aloud” about how he’ll use this knowledge in planning further testing. 2. An attending physician and 2 learners examine a patient with new onset congestive heart failure. After teaching how to hear a subtle S3 gallop with the stethoscope’s bell, the physician explains the accuracy and precision of this finding as a test for heart failure and provides a reference for further learning.² 3. After starting emergent therapy for a patient’s thyroid storm, an attending physician guides the team through the quantitative results of prognostic studies of this condition. The team then discusses how to use this evidence in counselling the patient.³

These teaching moments share 4 important features. Firstly, notice that the teaching is actually happening, despite the several barriers and disincentives clinical teachers face.^4,5 Secondly, the teaching happens “on foot”—that is, during the course of busy clinical work, rather than off site.⁶ Thirdly, while not exclusively so, these episodes involve teaching-in-context at the bedside.^7,8,9,10 This essay examines a fourth shared feature, the use of evidence from clinical care research. In all 3 episodes, the attending made conscientious, explicit, and judicious use of either the process of evidence-based learning or its yield.¹¹ The subtle but important variations in how this was done can be described as 3 different modes of teaching evidence-based practice, which are summarised in table 1.^12,13

In the first encounter, the attending role-models evidence-based practice (mode 1 in table 1), combining the actual doing with the “think aloud” technique to guide the learners’ attention to what is being done. By using mode 1, we show our learners how we incorporate evidence, along with other knowledge, into clinical decisions, so they come to see the use of evidence as part of good clinical practice. We show by our example that we really do it, and we can give pointers on when, where, and how to do it, too. Given the maxim “actions speak louder than words,” we can expect this first mode of teaching evidence-based practice to be very effective in itself, and it should reinforce what learners derive from other modes.

In the second encounter, the teacher weaves evidence in among the other facts and skills being taught (mode 2 in table 1). When we teach clinical topics with evidence in the mix, we show by example how to integrate evidence with other useful knowledge in the context of preparing to make real clinical decisions, so learners come to see that using evidence is part of good clinical learning. Also, since the evidence and the other knowledge can be learned together, they might be better organised in memory in the elaborated networks of knowledge so important for clinical thinking.¹⁴ By using mode 2, we are “putting our money where our mouth is”—that is, spending valuable teaching time and effort on using evidence in context, so we might expect this mode to be effective as well, although we haven’t seen either mode 1 or mode 2 studied.

In the third encounter, the teacher targets specific skills within evidence-based practice about which to coach the learners (mode 3 in table 1). By using mode 3, we help our learners develop their abilities for evidence-based learning, which in turn should boost their capacity for lifelong learning and professional development. By targeting 1 or a few specific skills at a time, the teacher makes it possible to fit these learning moments into the schedule of busy practice, knowing that the cumulative learning from many such moments could be great. Here, we get somewhat closer to having evidence about the effectiveness of this approach. The Cochrane review of trials of teaching critical appraisal skills found some trial evidence of improvements in participants’ knowledge.¹⁵ A recent review of the trials of teaching EBM to postgraduates concluded that standalone teaching improved knowledge, but not skills, attitudes, or behaviour, while teaching that is clinically integrated into routine practice improved knowledge, skills, attitudes, and behaviour.¹⁶ Although none of the trials appear to have tested the teaching strategies and tactics described above for modes 1–3, table 1 shows these modes would qualify as “clinically integrated.”

Although this essay concerns teaching “on foot,” there are other promising ways to teach evidence-based practice that are “clinically integrated” even if they’re not “on foot.” We’ll use the teaching mode structure in table 1 to describe 2 examples. Firstly, in an interactive large group classroom session for students and house officers on the approach to patients with involuntary weight loss, the teacher can include evidence about the frequency of underlying diseases in the material covered. In doing so, the teacher has selected to weave the results of research evidence in with other knowledge to guide explicitly the specific decisions and actions, so we can recognise this as teaching in mode 2 in table 1. Secondly, consider a department-wide, daily conference like Morning Report, wherein residents are asked to bring back to the group the evidence-based answers to questions that arose in the care of their patients, using an Educational Prescription.¹⁷ In addition to facilitating the residents’ discussions of their questions, searches, appraisals, and answers, the teacher can also interject relevant and narrowly focused “slices” about the process of evidence-based practice.¹⁸ Because the teacher is targeting specific skills around which to coach learners, this is teaching in mode 3 in table 1. Most of the clinically integrated forms of teaching we’ve seen or read about emphasise ⩾1 of these modes.

Keep in mind that the 3 modes in table 1 are not meant to be either jointly exhaustive (other possible ways exist to teach the wise use of evidence) or mutually exclusive (as hybrids do exist). Furthermore, these modes are complementary, and the best teachers of evidence-based practice use all 3, moving from one to the next to fit the clinical and teaching situation. If we actually use evidence in our own practice and clinical teaching (modes 1 and 2), we’ll have more realism and legitimacy when we coach our learners on specific skills of evidence-based practice (mode 3).

What can we do to prepare ahead of time to teach in these modes? The following 7 suggestions help us prepare for any mode. Firstly, we should refine our own skills in evidence-based practice, so we’re sufficiently competent to serve as role models. Secondly, we can work to gain access to evidence resources we’ll need, whether through our healthcare system (local or national) or on our own. Thirdly, because we encounter some problems and decisions repeatedly, we can anticipate the questions we’ll have and track down evidence-based answers ahead of time, so they’re ready for use. Fourthly, rather than store the entire text of the evidence, we might assemble and keep available concise summaries of the evidence, such as the 1 page summaries from evidence-based synoptic journals such as Evidence-Based Medicine or ACP Journal Club, or from our own CATs (critically appraised topics)¹⁹ or eCATs.¹ Fifthly, for some of these recurring decisions, investigative teams have assembled evidence-based decision aids to explicitly guide our patients and ourselves through the processes of integrating evidence with values, so we can retrieve these ahead of time and keep them within reach.²⁰ Sixthly, it can be useful to obtain and keep handy some how-to references on evidence-based practice, appropriate to our discipline.^13,21 Seventhly, we should develop our teaching awareness, or reflection-in-action, that allows us to recognise the teaching moments as they occur and helps us choose which to seize and which to let pass.²²

Beyond these 7 suggestions, a few things can be done ahead of time to prepare to teach in each mode. For mode 1, role modelling, many of us feel more comfortable after practising the use of evidence several times without a watching learner, before being observed by learners using evidence in real patient care. Once we feel comfortable with when, how, and how much evidence to add to our patient encounters, we’ll be better able to show others this “flavouring” or “spice” in our practice. For mode 2, we can anticipate that certain bits of evidence will be used alongside predictable aliquots of other knowledge, so we can prepare these mixtures ahead of time, as “teaching vinaigrettes.”²³ For mode 3, it can help to prepare ahead of time some short (ie, 2–5 min) scripts about the specific skills you’ll be targeting, as “teaching slices” of evidence-based practice (not the whole pie),¹⁸ whether your own or from others.^24–27

How can we improve further our teaching of evidence-based practice? As a complex craft built on experience as well as knowledge, excellence in teaching takes time to develop. Table 2 includes 5 suggestions for using that time well. Firstly, to develop excellence or even mastery, we probably need to undertake deliberate practice where we purposefully and repeatedly engage in activities designed to improve aspects of our craft.^28,29 Secondly, we can keep reflective teaching journals to record observations and interpretations of our own teaching, including both successes and failures,^30,31 as well as observations of other teachers, including those outside clinical medicine.³² Thirdly, we can identify respected teachers at our own institutions who can observe our teaching and provide us detailed feedback and coaching.³³ Fourthly, we can attend one of the growing number of workshops on teaching evidence-based practice being held around the world.³⁴ Fifthly, because teaching is so intertwined with learning, we can learn more about how humans learn and how this learning can be facilitated.^35–48

This essay has focused on how to teach evidence-based practice as part of clinical teaching, not on whether to do so.⁴⁹ I have tried to collect here the teaching strategies and tactics I and others have found useful, and I look forward to learning more from many of you.