Chapter 5: Finding Current Best Evidence

Introduction

Searching for Evidence: A Clinical Skill

Searching for current best evidence in the medical literature has become a central skill in clinical practice.^1,2 On average, clinicians have 5 to 8 questions about individual patients per daily shift^3-5 and regularly use online evidence-based medicine (EBM) resources to answer them.^6-9 Some now even consider that “the use of search engines is as essential as the stethoscope.”¹⁰

However, because of the increasing volume of new literature and speed of new research, finding useful evidence efficiently remains challenging. Approximately 2000 new articles are indexed in PubMed every day,¹⁰ and although few of them directly inform clinical practice, as many as 75 are randomized clinical trials and 11 are systematic reviews.¹¹ These numbers explain why searching in PubMed is not the most efficient way to look for evidence-based answers. For example, when typing “stroke prevention in atrial fibrillation” in PubMed, you will see that current best evidence is literally lost in an output of almost 4000 citations, with a mix of trials, reviews, guidelines, and editorials that are impossible to screen for relevance during your daily practice.

Fortunately, numerous EBM resources now provide shorter and more efficient paths. These resources select, process, and organize the evidence; some, however, are more trustworthy than others. This chapter will help you navigate through existing EBM resources, distinguish the trustworthy from the less trustworthy, and maximize your chances of quickly finding answers based on current best evidence.

Start by Clarifying the Question

As we have seen in Chapter 4, What Is the Question? framing the question appropriately is an important prerequisite to any search. An initial distinction to make is whether you are asking a background question (eg, definition or pathophysiology of a syndrome or mechanism of a treatment modality) or a foreground question (eg, targeted questions of therapy, harm, diagnosis, or prognosis that provide the evidentiary basis for decision making). Although some EBM resources also answer background questions, this chapter, and the Users' Guides to the Medical Literature overall, focuses on efficiently finding answers to foreground questions.

Foreground questions often arise in a form that does not facilitate finding an answer (see Chapter 4, What Is the Question?). A first step is to translate and structure the question into its components, using the PICO framework, which accounts for the patient or population, the intervention or exposure, the comparator, and the outcomes (see Chapter 4, Box 4-1). When framing your question, remember to consider all patient-important outcomes. Doing so will guide you in selecting the body of evidence that adequately addresses your patient's dilemma between benefits and harms that matter to your patient's decision.

Structuring the question will not only clarify what you are looking for but also help you ...