CHAPTER 2
Evidence-based History and Examination

John Frain

OVERVIEW

• The history and physical examination are essential diagnostic tools
• To formulate a differential diagnosis, the history should identify the patient’s key presenting symptoms and associated symptoms
• The patient’s background and the natural history of a condition are essential contextual factors in making a correct diagnosis
• An incomplete physical examination increases the risk of diagnostic error
• If available, likelihood ratios are a measure of diagnostic accuracy which are easy to understand and apply at the bedside
• Clinical teachers should be familiar with evidence-based history and examination

Introduction

Evidence-based history and examination surveys all data from the clinical encounter, compares it to available evidence, including recognised diagnostic standards, and quickly identifies those variables with the greatest diagnostic accuracy.

Around 80% of diagnoses are made from the history alone [1]. Its importance is such that Hampton et al. recommended more emphasis should be placed on teaching students accurate history-taking and more emphasis should be placed on research into communication between patients and clinicians [1]. The purpose of the history is the generation of a differential diagnosis broad enough to include the actual diagnosis but focussed enough to be tested by an appropriate physical examination, and investigations when necessary. The differential diagnosis consists of a leading hypothesis and two or three other hypotheses including any ‘must not miss’ diagnoses, all justifiable by the data gathered, and taking into account the patient’s own concerns. Generating a differential diagnosis guards against premature closure (see Chapter 9). For the learner, it aids in developing their clinical reasoning ability. Generating a differential diagnosis needs to be taught alongside the history and physical examination, not separately. This chapter will introduce the idea of evidence-based history and physical examination, an important component of clinical reasoning.

The Evidence Base

Most of our knowledge of history and physical examination predates the development of evidence-based medicine, and from a time when patients presented later in the course of their illness, usually with more ‘classic’ features. Nowadays, especially in developed countries, clinical problems present earlier and are often less well differentiated from one another. We know that many single features in the history and physical examination are limited in their usefulness (e.g., whether someone’s chest discomfort is ‘central and crushing’ or not; or whether Kernig’s sign is present).

Nonetheless, good quality evidence now exists, including systematic reviews of symptoms and diagnostic accuracy studies of physical signs. This is being assimilated into textbooks (see further resources) that provide epidemiological and qualitative data as well as diagnostic accuracy studies including sensitivities, specificities, and likelihood ratios. Where statistical evidence is lacking, authors provide diagnostic guides based on the best available epidemiological evidence. Evidence-based principles and knowledge can be reinforced for learners through reflective coursework (see Box 2.1) which can further enable reflection on clinical reasoning during workplace training.

Box 2.1 Reflective coursework: the presentation of chest pain in women

‘I encountered a female patient in her 50s presenting with a possible acute coronary syndrome (ACS). We had recently been discussing in clinical skills about possible differences in the presentation of acute coronary syndrome in women and men. I decided to examine the evidence for this.

I undertook a systematic search of the literature using the 6S evidence pyramid we had been taught. I identified 16 relevant studies using the SORT criteria.i There was heterogeneity of data recording and analysis across the studies. However, I was able to draw the following conclusions:

• Women are more likely to present with ACS atypically compared to men
• Men are more likely to present with chest pain than women
• Chest pain is the most common symptom presentation for both sexes
• Younger women are more likely than older women to present with typical symptoms
• No difference exists between sexes for prevalence of chest pain and/or other typical symptoms of ACS
• Women reported more associated (non-chest pain) ACS symptoms than men

Continued development as a clinician requires three things: reflective practice, unwavering curiosity, and maintaining an open mind to new evidence and ideas. Writing this essay has highlighted the importance of looking to the highest level of evidence for guidance, while maintaining a healthy level of scepticism for the recommendations by analysing the primary literature and systematic reviews behind them.’

• Ebell MH, Siwek J, Weiss BD et al. (2004). Strength of recommendation taxonomy (SORT): a patient-centered approach to grading evidence in the medical literature. Am Fam Physician; 69(3): 548–56.

This is an excerpt from a patient-based piece of coursework undertaken by graduate entry medicine students at the University of Nottingham, UK. Students examine the evidence underlying a clinical feature encountered in practice.

Exploring Symptoms: Combining Process and Content

The process of history-taking cannot be separated from knowledge of epidemiology, anatomy, and physiology. Nor can it be separated from effective communication with patients [1]. Irrespective of the cause, each patient seen by a clinician is essentially saying: ‘I’ve noticed some changes in my anatomy and physiology. Can you tell me, is this normal, part of getting older or is it pathology? If it’s pathology, will I return to my previous health, will it leave me with some loss of function or could it even shorten my life?’ Clinical reasoning and decision-making is required to assess all these concerns.

Models for clinical communication have been developed, most notably the Calgary-Cambridge model [2] consisting of 70 skills which facilitate accurate history-taking (see Box 2.2). Teaching this model involves deliberate practice with detailed, specific feedback by observers. It is used either one-to-one with a tutor and patient, with patient actors, or at the bedside with real patients. An adapted form of the observation guide includes feedback on elements of clinical reasoning [3, 4].

Box 2.2 The Calgary-Cambridge model: the process required for accurate history-taking

• The patient’s opening statement (usually 30–120 seconds)
• Identifying a problem list
• Agreeing an agenda for the interview
• Exploring each symptom experienced by the patient
  • Patient and clinician agreeing on definition of each symptom present
  • Gathering data for each symptom using open questions
  • Completing details using closed question
  • Establishing a sequence of events
  • Attentive listening
  • Picking up cues
• Exploring the patient’s relevant background information
  • Relevant systems review
• Ensuring all the patients concerns have been addressed
• Explanation and planning
• Shared decision-making

Adapted from Silverman J, Kurtz SM, Draper J. Skills for Communicating with Patients, 3rd edn. CRC Press, 2013.

When taking a history, key symptoms emerge within the relevant system (see Box 2.3). While symptoms may overlap different systems (e.g., chest pain could be cardiac, musculoskeletal, or respiratory in origin) or be challenging for both patient and doctor to define (e.g., dizziness), many diseases present with varying configurations of key symptoms within the relevant system. Provided the clinical setting of an individual symptom is clearly defined (e.g., nausea and vomiting in patients with suspected intestinal obstruction, or chest pain in patients with suspected myocardial infarction in the emergency department) it is possible to reason its significance and thus its usefulness as evidence of the presence of the target condition. Course content can be developed to teach students the evidence-base for each symptom alongside the process of exploring symptoms with the patient (Box 2.4).

Box 2.3 Summary of key symptoms by body system