AS a lay person when it comes to medicine, I have a certain understanding of the word “diagnosis”. This was reflected in my doctoral research on the structure of surgeon–patient consultations where I preferred to label it as “reformulating the problem” to capture the process of combining information from the problem presentation, history taking, and examination as well as the often uncertain or iterative nature of diagnosis that I observed.

Perhaps unsurprising to this audience, these observations align with the medical understanding of the process of diagnosis. The term “diagnosis”, for most people, is a noun, with a label-like quality. Yet for a doctor, it would be better understood as a verb, “to diagnose” capturing a multistage process that can involve as little as a history taking or examination to years of investigations and multitudes of clinicians.

To further understand the relationship between communication and diagnosis, here I interview two experts in the field: Dr Mary Dahm, a linguist and Research Fellow at the Australian National University, and Associate Professor Carmel Crock, the Director of the emergency department (ED) at the Royal Victorian Eye and Ear Hospital.

SW: Carmel, what is the diagnostic process?

CC: Diagnosis is a form of inquiry and problem solving and is at the heart of medicine. Clinicians search for an explanation for a patient’s symptoms, as well as means to treat them, to alleviate suffering. It is interesting and perhaps confusing that in medicine, as you point out, the word diagnosis is used in these two ways: as a naming, a label or classification of a disease but also as a process that we undertake to uncover an explanation of the patient’s symptoms but without always a clear end point or clear answer. The diagnostic process does not necessarily provide a diagnosis.

Annemarie Jutel describes diagnosis as a “narrative exchange” between patient and clinician. The patient brings their story which is “heard, processed, rearranged and retold” by the doctor and in effect the diagnosis is a negotiated retelling of the story.

There is a great figure depicting the diagnostic process in the National Academy of Medicine Improving Diagnosis in Health Care report. It encapsulates what clinicians and patients experience as the process unfolds over time. We gather information, integrate and interpret that information (from history, physical examination, diagnostic testing and referral and consultation) and form a hypothesis about what might be going on, or a working diagnosis.

Using the term “working diagnosis” reminds us that our diagnosis is often a hypothesis and that refining our hypotheses as new information comes to light is part of the diagnostic process. It acknowledges the inherent uncertainty in medicine.

Underpinning the diagnostic process is the passage of time; some diagnoses are time critical – subarachnoid haemorrhage, aortic dissection, myocardial infarct, for example. Others play out over time and across different health settings. This importance of the passage of time can’t be over emphasised, as there are situations when watchful waiting and allowing a diagnosis to unfold is the best medicine. Reliable follow-up and feedback are required. And sometimes careful listening and observing patients over time may provide more information than testing.

Viewing diagnosis as a process allows us to adapt process and quality improvement methodology (here and here) that has been used in other areas of medicine such as surgery and medication management. Each step of the diagnostic process can be focused upon for improvement. In the ED, for example, information gathering can present a challenge, including loss of or poor transfer of information between the ambulance and the ED and from the ED to GPs. Serious diagnostic errors and patient harm can occur because of these communication failures.

SW: Mary, diagnoses were provided in about 50% of primary consultations in a large qualitative study of diagnosis communication in practice in the US. How is not giving a diagnosis managed in practice?

MD: There are several ways in which clinicians can provide diagnostic information to patients. They can give patients a diagnosis in the label-like quality you mentioned earlier. Saying something like “yes, that’s tonsilitis” would provide patients with a clear diagnostic declaration. But clinicians also often provide diagnostic information without assigning a label, especially if the patients’ sign and symptoms are still diffuse and do not clearly point to one specific disease process. For example, the clinician can share different candidate diagnoses (but not a definite diagnosis) with the patients by giving insights to their diagnostic reasoning: “At this point in time my best guess is that you have tonsilitis, but I’m also thinking of scarlet fever or mononucleosis, and we need to run some tests to be sure”. Giving patient insights in diagnostic reasoning is also a useful debiasing strategy that can help to fend off common cognitive biases that can lead to diagnostic errors.

Another option is to let the patient know which potentially serious diagnoses can be excluded. This can be useful for patients who, for example, present with relatively benign symptoms but who are anxious about serious diagnoses because they run in their family. Someone with prolonged difficulty swallowing and a painful throat might be scared to have throat cancer, a condition their father died from. “I don’t think that you’ve got what your father had but I’m thinking it could be a virus that has been giving grief for a while.” It is important to elicit such patient candidate diagnosis early to address and alleviate unfounded fears.

SW: Carmel, when considering the concept of shared decision making, the original intent was sharing decision making. That is, that the information being provided to the patient was personalised for them and explained contextually, with that decision-making process being made explicit in the discussion. This was developed in the 1970s. Where does this fit in the 21st century medicine?

CC: Sharing our decision making during the diagnostic process is really important. Many patients who come to us have already googled their symptoms and have an idea of what they think the diagnosis may be. I think that this has helped balance the doctor–patient relationship in many ways, and has made patients much more engaged in the diagnostic process. Our discussions of balancing risk are becoming more nuanced with our patients. And there is growing awareness of issues around both under- and overtesting and diagnosing.

There need to be really good interactions between patients and clinicians and between clinicians to improve diagnosis. For example, we may see a patient with severe eye or ear pain but no physical signs and early herpes zoster may be in our differential diagnosis. We want our patient to be on the lookout for development of vesicles. Our decision making around starting antiviral therapy may well need to be clearly communicated both to the patient and their GP.

The patient and other clinicians need to understand our thinking. So yes, making our decision making explicit is a great step in diagnosis. It opens up a space for questioning, revisiting, challenging and giving feedback, which are all essential in the diagnostic process.

SW: Mary, from a patient perspective, a diagnosis can be comforting (and legitimising). How do doctors balance explaining a “diagnosis” that is not a label, particularly one that is uncertain or that needs more information to be understood?

MD: I think the most important communicative behaviour for clinicians who are not providing a label-like diagnosis is to maintain open lines of communication, explain the evolving nature of diagnosis, be honest and disclose their diagnostic uncertainty.

As in all health exchanges, expectations and thus ideas of success in diagnostic interactions are contextualised and often misaligned between patients and clinicians. Many patients expect a label-like diagnosis when they first present and are unfamiliar with how diagnoses evolve in an iterative process. Explaining and educating patients about the common steps in the journey to a diagnosis, including the need for further tests or investigation, but also the need for watchful waiting, can help reset patients’ expectations surrounding the diagnostic process, and can reassure patients.

Along with sharing diagnostic reasoning and providing candidate diagnosis, clinicians should also be comfortable with saying “I don’t know”. Patients can become frustrated if they do not receive an explanation for their symptoms and clinicians did not express their uncertainty. Naturally, patients have different preferences, and studies in diagnostic communication have shown that while patients feel reassured by clinicians sharing their diagnostic uncertainty, other patients may lose trust in their doctor and question their technical competence.

While it is certainly okay to disclose diagnostic uncertainty, patients often find comfort in diagnosis and, therefore, clinicians should also listen empathetically and involve patients in planning of further diagnostic steps. For example, clinicians can acknowledge that some form of disease process is occurring, which needs to be treated without assigning a label. By having their concerns acknowledged and by becoming partners in the diagnostic journey, patients feel reassured and legitimised in seeking medical help.

Previously known as the Australasian Diagnostic Error in Medicine Conference, the Improving Diagnosis Conference will be taking place virtually from 28 to 29 April 2022. The conference theme is: Embracing Challenges and Change. Hosted by the Australian and New Zealand Affiliate of the Society to Improve Diagnosis in Medicine (ANZA-SIDM), it will bring together diverse multidisciplinary groups, including researchers, patients, physicians, surgeons, radiologists, paramedics, linguists, psychologists, medical administrators, and others, all dedicated to the subject of making diagnosis safer. Dr Sarah White and Dr Maria Dahm will be hosting a conference workshop on Communication and Diagnosis.

Associate Professor Carmel Crock is Director of the Emergency Department at the Royal Victorian Eye and Ear hospital and Chair of the Quality and Patient Safety committee at the Australasian College for Emergency Medicine. She is board chair of the Australia and New Zealand Affiliate of the Society to Improve Diagnosis in Medicine (ANZA-SIDM) a group dedicated to improving diagnostic quality and safety, and reducing diagnostic error in medicine.  

Dr Mary Dahm is a linguist and Research Fellow at the Australian National University. She was awarded a Discovery Early Career Fellowship Award from the Australian Research Council for a project entitled “Addressing the challenge of communicating uncertainty in diagnosis”. She will commence work on her DECRA project in 2022.

Dr Sarah J White will be starting as a Senior Lecturer at the Centre for Social Impact at UNSW in early 2022. She is an Honorary Senior Research Fellow at the Faculty of Medicine, Health and Human Sciences at Macquarie University. Dr White is the current Australian National Representative for the International Association for Communication in Healthcare.



The statements or opinions expressed in this article reflect the views of the authors and do not represent the official policy of the AMA, the MJA or InSight+ unless so stated.

4 thoughts on “Diagnosis: question or answer?

  1. Anonymous says:

    One of my most memorable tutorials at Medical school (>20 years ago) was given by a GP. He asked us what we thought was the most important decision for the GP from a (typical) very brief consultation when faced with a new patient or a new complaint by a known patient. Most, if not all of us med students felt that a diagnosis (or differential) needed to be made.
    His response was that it was much more important to determine a prognosis for the issue when deciding what to do with /for the patient.
    I am a physician but I have always used this as my guide to my management plans and it has stood me in very good stead. That is not to say that the actual diagnosis is not important but it is not always crucial. Many maladies get better despite what we do, not because of what we do.

  2. J says:

    Agree with Dr Hocking. Michael Balint’s classic based on his experience of doctors-encountering-difficulties – largely general practitioners – during weekly sessions at the Bethlehem and Maudsley Hospital in London is a sadly neglected text.
    As for diagnostic naming, the patients and their relatives are often all too keen to seize on a label, the meaning of which may not be shared by the doctor. Their is another publication by Asher, a distinguished London physician, from the sixties which describes the snares which narrowly defined illness components set for the attending physician / surgeon.

  3. Anonymous says:

    When I graduated from medicine (Syd. Uni.) in 1972 the diagnostic process, formulated for at least a century, was as follows:
    1. History- mostly symptoms but including family history, past health, medications.
    2. Physical examination (signs).
    3. Provisional diagnosis (PD)- based on symptoms and signs – what was causing the patient’s illness.
    4. Differential diagnosis (DD)-what other pathology could cause the illness.
    5. Investigations – to confirm the PD, exclude the DD or give a new PD.

    Unfortunately our modern undertrained (not in molecular biotechnology) doctors sem to have no idea of this algorithm.

  4. bruce hocking says:

    I strongly suggest the authors read the classic “The Doctor, the patient and the Illness” by Michael Balint (Pitman) for a much better understanding of the dynamics of the “diagnostic process”. It ain’t what you do its the way that you do it!

Leave a Reply

Your email address will not be published. Required fields are marked *