The clinical history

Carefully constructed clinical histories underlie successful patient-centred consultations

Osler urged his colleagues and students to: “Listen to your patient, he is telling you the diagnosis”. One of my mentors advised that one must “find out what is wrong with the patient and why they came to see you”. This advice enlightened me and bequeathed improved care to my future patients. Quotable maxims such as these crystallise decades of clinical experience and retain their value to this day. However, increasingly, we can and should improve our consultation skills using contemporary quantitative and qualitative research.

In Osler’s time, doctors necessarily focused on their patients. Patients provided an account of their illness, and physical examination completed the assessment. Physicians were little distracted by test results, of which there were few, or medical theory, of which there was relatively little. In recent decades, clinical leaders have championed a re-orientation to the patient termed “patient-centred” care, which corrects the drift to disease-centred or doctor-centred care that emphasises tests, test results and medical minutiae.

The patient

Doctors practise patient-centred care when they implement two main principles, which have overlapping subcomponents.1 The first principle is that the doctor must attempt to understand the patient’s perspective — including their thoughts and feelings concerning their illness, and their expectation of the doctor. The second principle is that the patient and doctor share responsibility for the consultation and subsequent decision making. In patient-centred care, the history is constructed, not taken. The final construction may be endorsed as accurate by the patient and interpreted diagnostically by the clinician. Techniques for such a construction are increasingly well documented.1-5

Rapport

Patient–clinician rapport develops more rapidly when common experience or values are identified, even when they seem irrelevant to the consultation. This communion can be as simple as the clinician noting that the patient’s scarf indicates their shared support for a specific football club. Patient–clinician rapport and patient-centred care are synergistic, and good patient–clinician rapport is associated with more accurate clinical histories, as well as improved patient satisfaction, adherence and health outcomes.1,5,6

Constructing a history

The patient’s history is given primacy by first asking open questions such as: “What is it that’s most worrying for you at the moment?” or “How can I help you today?” The patient’s answer may be a mixture of facts, interpretations, hopes and fears, with the motive for medical consultation uncertain or unexpressed.1

The clinician can encourage the patient’s narrative with a receptive demeanour, verbal encouragement (“Tell me more about …”), and strategic use of silent pauses.4 Eye contact expresses interest in some cultures and represents an affront in others, so it must be carefully gauged.

Questions should be open-ended, phrased to avoid ambiguity and screened to remove subtle presuppositions.5 Listening should be active (“Why did you stop working at that time?”), sympathetic (“Was that difficult for you?”) and empathic (“I would have felt very sad at such news”), as appropriate. Accurate information, patient affirmation and support are all important, with none being pre-eminent.4 Some histories are completed in a few sentences, while others require more than one consultation.

Meaning

The clinician must uncover what the patient’s words mean to them.5 Is the patient complaining of weakness unable to perform a specific action, such as extending the leg, or experiencing a feeling that previously routine activity no longer seems possible? Common words may be used in idiosyncratic ways, disadvantaging doctors with English as a second language.7 Visiting doctors in areas of rural England must learn that local patients use the term “front bottom” to refer to the vulva or vagina. Patients may emphasise the significance of their symptoms by using words with technical meanings. Hypoglycaemic and dehydrated may simply mean very hungry and very thirsty.

Vocal melody and prosody express the emotional significance of the patient’s narrative. Alert, empathic clinicians use their mirror neurons, which reflect the patient’s facial expression and demeanour, to deepen understanding. Junior doctors may be less confident of these subjective assessments, and are less likely to incorporate them into their formulation of the provisional diagnosis.4,8 It is paradoxical that when clinicians acknowledge and address the patient’s emotional and psychosocial concerns, consultations are shorter than those in which they go unrecognised or are ignored.4

Preconceptions

Patients sometimes pre-empt their doctor’s diagnosis. “Dr Google” is cheap and readily available, is non-judgmental and always willing to offer an opinion. Wikipedia, patient support websites and blogs augment the algorithmic conclusions of “Dr Google”. Therefore, clinicians striving for a successful consultation must assist their patient in parsing the good, the bad and the irrelevant, regardless of whether it originates on the internet or is simply the conjecture of a concerned friend. Success or otherwise may be indicated by something as subtle as the difference between the patient who says brightly: “Well, thank you, doctor”, and the one who says flatly: “‘Well, thank you, anyway, doctor” at the end of the consultation (Dr JD Blackie, Senior Staff Specialist, John Hunter Hospital, personal communication, 2000).

Context

Events in the patient’s history unfolded in a milieu of hopes, fears and relationships unique to that individual.4 Therefore, all history must be understood in its temporal and psychosocial context.2,9 Often, the solution to a puzzling diagnostic problem lurks in the time preceding the patient-defined beginning of their story, or in the psychosocial events surrounding it. Thus, the clinician’s understanding should encompass the patient’s important relationships, as well as significant life changes in preceding years.10 It is inevitable that patients will omit factors they judge to be irrelevant, and information so basic or intrinsic to their life that it is invisible to them.

The doctor

At the same time as they construct the history with the patient, experienced clinicians will identify one or more provisional diagnoses — often within 30 seconds of commencing the consultation.11 Doctors then test these diagnostic hypotheses in parallel with the history construction, using the patient’s spontaneous narrative, as well as direct but still open questions. The trajectory of a consultation will differ with each clinician and each new patient.

Diagnostic skills

It is not surprising that expert and novice clinicians differ in several ways, including the sum of their biomedical knowledge, the structure and integration of their clinical knowledge, and the specific problem solving techniques used. Experts arrive at diagnostic hypotheses more quickly and accurately than novices, due to differences in each of these elements. Although the timing and use of these diagnostic elements varies with each clinician and each patient, there has been little study of this variation among clinicians.12 While serendipity, in the form of a chance remark or observation, sometimes supplies the missing piece in a diagnostic puzzle (author’s observation), Pasteur was correct in noting that: “Chance favours the prepared mind”.

Knowledge structures

Experts have, over time, encapsulated their knowledge more effectively than novices. As a consequence, cardiac failure or raised intracranial pressure may be used as single components in diagnostic thinking, without the need to parse component symptoms and signs or refer directly to underlying pathophysiological knowledge, except when it is specifically required.

Probability

Brains are metabolically expensive tools for predicting the future, including likely diagnoses. Nonetheless, most humans, including clinicians, are notoriously inept when confronted with formal problems in probability.13 Despite this difficulty, expert clinicians adjust the likelihood of their diagnostic hypotheses with knowledge of population characteristics pertinent to their patient (eg, malaria and tuberculosis are much commoner infections in Port Moresby than in Melbourne). While these adjustments can be succinctly described using Bayesian algebra (ie, Bayes’ theorem calculates the probability of disease A, given fact B; the next iteration calculates a new probability of disease A given the additional fact C and so on), probabilistic decision trees or probability matrices, few clinicians use these tools consciously. In fact, exactly how clinicians determine or adjust likelihoods, and consciously or unconsciously use probabilistic thinking, is yet to be clarified.12

Hypothesis testing

Expert clinicians test diagnostic hypotheses using illness scripts, which are story-like narratives that condense the range of different presentations of a specific disease into a general mental representation that can then be compared with the patient’s history.14 For example, loss of consciousness is a starting point for at least two illness scripts: generalised seizure and syncope. Targeted questions concerning prodromal and postictal symptoms will often allow the clinician to identify one script as being more likely than the other in explaining the patient’s presentation.

Experts are more likely than novices to employ semantic qualifiers to test the match between a specific illness script and the patient’s history. Semantic qualifiers are usually binary, antithetical, abstract descriptors, such as fast/slow, distal/proximal or simple/complex.15 These qualifiers assist in formulating illness scripts by clarifying, simplifying and generalising the patient’s history. Both illness scripts and semantic qualifiers are more likely to be used by expert diagnosticians and are associated with more accurate diagnosis.14

Problem solving

Experts are more likely than novices to employ pattern recognition and deductions arising from illness scripts when generating and testing diagnostic hypotheses. While it is unknown if that difference is coincidental or causal, experts still default to the slower hypothetico-deductive approach when their initial strategies founder.13

Consultations are unique

The trajectory of history construction and diagnostic thinking varies widely, even among expert clinicians. This is because every patient is unique and each will interact with a clinician whose experience, style and habitual thought patterns differ from those of equally competent peers. Despite 35 years of research in clinical diagnostic thinking, understanding of the process and of the acquisition of diagnostic skill by novice clinicians continues to resist simple explanation.13

Conclusion

Most junior doctors have at some time been embarrassed on consultant rounds when the patient contradicts their careful summary of the history. They know their consultant regards the history as extremely important, but they must also learn that no one true history exists. Novelists such as Lawrence Durrell (The Alexandria quartet), and Amy Tan (The kitchen god’s wife) have emphasised the lack of any canonical authority adjudicating competing views of events. In light of this conundrum, how might the value of a consultation be assessed?

A patient-centred consultation may be judged successful if the constructed history has enabled the identification of a most likely diagnosis, the patient understands how and why that point has been reached, and given that understanding, has decided with the clinician on the most appropriate course of action for that patient at that time. I trust that Osler would approve.