Diagnostic uncertainty: when we just don’t know

by Carolyn Thomas      @HeartSisters

The image above is all about uncertainty. It’s like a 5-step roadmap that you’d use when traveling an unfamiliar road to a new destination you know nothing about and do not want to visit.(1)  For people experiencing scary symptoms they fear might be heart-related, for example, uncertainty about what’s happening now and what will happen next is pervasive. But a new study published in the journal Patient Education and Counseling reminds us that patients aren’t the only ones facing uncertainty around a medical diagnosis: “Both patients and clinicians experience diagnostic uncertainty, but in different ways.”(2)        .        .  

It turns out that diagnostic uncertainty can have far-reaching effects on patients, on the clinicians who are helping us, and indeed, on entire healthcare systems (the latter through over-testing, increased healthcare costs, hospital re-admissions, and decreased trust in the system).

This study’s Texas-based authors defined diagnostic uncertainty as the “subjective perception of an inability to provide an accurate explanation of a patient’s health problem”.  For patients anxiously waiting for a diagnosis, this perceived inability can lead to missed and delayed diagnoses, treatment delays, and unnecessary tests and/or treatments. And for clinicians, diagnostic uncertainty can lead to stress, anxiety, burnout, or withholding treatments.

As their study reports:

“Throughout the diagnostic process, patients experience uncertainty due to not knowing how the diagnostic process works or how it’s unfolding. It’s often perceived as singularly the clinician’s role to make a diagnosis, but this process can lack transparency for patients due to the high level of medical knowledge needed to understand it. Communication and listening skills are key parts of the diagnostic process. And how clinicians express uncertainty may positively or negatively impact patient satisfaction and decision making.

Clinicians experience uncertainty as well throughout the diagnostic process. However, for clinicians, this uncertainty typically arises due to incomplete mastery of medical knowledge or undeveloped science. This includes incomplete knowledge among  novice clinicians, but also situations in which clinicians face rare diagnoses, or when new diagnoses emerge (e.g. COVID-19) or in well-known medical conditions for which many unknowns remain.”

Researchers also warn that an accurate diagnosis is more than just coming up with the name of a medical disorder that’s causing the health problem. Uncertainty is pervasive throughout the entire 5-step process, starting with the very first symptoms we experience, as this list suggests:

Step 1 – The patient experiences a health problem and engages with the healthcare system: 

In this stage, patients generally feel two key uncertainties: what their troubling symptoms actually mean, and when or even if to seek care. Uncertainty about new symptoms and/or illness can cause anxiety and fear. Patients with negative past healthcare interactions may hesitate to engage at all. Patients are in a vulnerable position: they have to trust clinicians to first make sense of distressing symptoms and test results, and then to arrive at a correct diagnosis.  No correct diagnosis = no appropriate treatment plan. Researchers also warn that patients seeking care in acute care settings may, as I did when my heart attack was misdiagnosed in the Emergency Department, leave their initial visits without accurate diagnoses, and with lingering uncertainty.

Step 2 –  Gather the Information:

Patients now feel uncertain about what to tell or leave out of the health information they’re sharing with a healthcare professional. While patients are experts on their own bodies, they may not know how or if their symptoms are related,  or what to report to clinicians. This may lead patients to under-report symptoms or symptom severity (leading them to be unfairly labeled as “poor historians”) or, conversely, to share information that’s not relevant to their condition. It may also be unclear to the patient how the questions being asked by clinicians around the patient’s medical history or physical exam relate to their present symptoms.

Clinicians, meanwhile, can feel uncertain about whether they’re getting sufficient information from the patient. While collecting diagnostic information, they must continuously assess whether that information seems accurate or not, and if they have enough information to narrow down a working diagnosis into fewer diagnostic possibilities. The Texas researchers describe the “well-calibrated clinician” who knows what an acceptable level of uncertainty is and when that level of uncertainty is too high – thus prompting additional action, e.g. seeking help from colleagues, ordering additional diagnostic tests, or rethinking their diagnostic options.

Step 3 – Integrate and Interpret the Information:

Patients feel uncertain about the whole diagnostic process, about what their test results mean, and about differing opinions – largely because they lack the medical knowledge that clinicians have. Some patients may decide to seek second opinions if they feel uncertain about their clinicians’ trustworthiness, or if their clinicians have not shared enough information about the diagnoses with them.

Clinicians can feel uncertain about the characteristics and even the accuracy of diagnostic test results, the details of the test reports, and if the results are either relevant or significant.  They may face uncertainty about the patient’s current concerns, past medical history, family history, social history, current medications, and other important information that informs their clinical reasoning. And because of consultation time constraints, clinicians often need to make rapid decisions about which information might be more important to consider.

The Texas researchers also make this observation around interpretation:

“Additionally, during the physical exam, there is uncertainty around the words that both clinicians and patients use. Clinicians often use medical terminology during history-taking and it’s not always clear that the patients understand.

“Since many of these questions take the form of yes/no questions, patients may not speak up about not understanding.  Patients use non-clinical terms and they may not always translate to clinicians in the same way. For example, different patients use different words or focus on different aspects of their symptoms (e.g., gender differences in the way patients describe symptoms of angina). Clinicians must also translate the language patients use into corresponding clinical meaning. This issue will be exacerbated when there are language, educational, or cultural barriers between clinicians and patients.”

(Speaking of gender differences in the words we use to describe symptoms, consider the research of Harvard’s Dr. Catherines Kreatsoulas, a linguist who studies how women communicate to Emergency Department clinicians during a heart attack. She reports that women, for example, are significantly less likely to use the word “pain” to describe their cardiac chest symptoms, instead using words like pressure, heaviness, fullness, burning, ache, etc. See also:  Words Matter When We Describe Our Heart Attack Symptoms).

Step 4 – The Working Diagnosis:

Patients may not feel certain if their test results are pointing to the actual final diagnosis or not. Even when diagnostic test results appear “normal”,  patients may fear something serious was missed. And they may not know the degree of uncertainty experienced by their clinicians.

Clinician uncertainty involves the accuracy and adequacy of all the information collected so far.. They have to decide how much certainty is enough to avoid the risk of diagnostic error.  The Texas researchers found that more experienced clinicians display more comfort with uncertainty, and are correspondingly less risk-averse when managing their patients. This increased comfort can come with its own risk, however; it may result in less information-gathering during diagnosis, or increased willingness to make a conclusion even when some data do not fit, or what academics who teach critical thinking skills call anchoring bias (“locking on to a diagnosis too early and failing to adjust to new information”).  Additionally, anxiety and concern related to uncertainty tends to be higher in less experienced clinicians. This heightened anxiety could lead to excessive data collection and overuse of resources, but as the Texas team found, can also lead to a more thorough diagnosis.

(Speaking of conclusions that are made despite data that doesn’t fit: this is precisely my take on how the older, male  Emergency physician – clearly comfortable with uncertainty – swiftly misinterpreted my textbook heart attack symptoms of central chest pain, nausea, sweating and pain down my left arm: “You are in the right demographic for acid reflux!” he pronounced confidently. When I asked him directly “But what about this arm pain?”  – I’m not a doctor, but even I knew that pain down your arm is NOT a sign of indigestion –  he did not revisit his misdiagnosis. Instead, a few minutes later, an Emergency nurse came over to my bedside (no doubt having had an earful from her colleague about the pain-in-the-neck patient in Bed 8) and sternly scolded me:

“You’ll have to stop asking questions of the doctor. He is a very good doctor, and he does NOT like to be questioned.”

Step 5 – Communicate the Diagnosis:

The Texas researchers observed that patients may be uncertain about what this diagnosis means, and sometimes even uncertain if they want to know the diagnosis.  They may not know whether they should have a follow up appointment, or seek additional care if symptoms get worse or change – unless clinicians make that clear to them. 

(I recall feeling that precise uncertainty on the morning I returned to the same Emergency Department where my cardiac symptoms had been misdiagnosed two weeks earlier. When the cardiologist said the words, “significant heart disease” to describe my new diagnosis, I asked him if this meant I was going to have a heart attack.  His reply: “You’re having one right now!”  But my odd reaction to that was to ask if I should make an appointment to come back for treatment at a later date. “No!!” he snapped: “We’re sending you upstairs to Cardiology NOW!”  Clearly, I felt a wee bit uncertain about everything after hearing that diagnosis).

The researchers also report that, when clinicians communicate the most probable diagnosis to the patients or their families, their uncertainty in this important step often lies in how and what to communicate.  While this study recommends that clinicians should discuss uncertainty with patients, others have argued that communicating diagnostic uncertainty with patients is not a skill that’s well taught during medical training, and patients may in fact rate physicians negatively if they admit uncertainty. Other studies have found that clinicians are more likely to share general uncertainty with patients who are better educated, those who ask more questions, and those who request more information.(3)

Communicating diagnoses does not guarantee that patients understand the diagnoses, hence, the importance of clinicians providing support and information to help patients understand and cope with uncertainty.  Trust and confidence in healthcare professionals’ ability to make diagnoses can decrease uncertainty, while a lack of confidence or a lack of information may increase uncertainty. 

Finally, the three take-away points to remember from this study:

  1. Uncertainty occurs throughout the diagnostic process.
  2. Both patients and clinicians experience uncertainty, albeit in different ways.
  3. Educating patients about the diagnostic process and communicating any uncertainty throughout can help manage uncertainty.
1. Image: ©National Academies of Sciences, Improving Diagnosis in Health Care, The National Academies Press, Washington, DC (2015)
2. Ashley Meyer et al.Patient and clinician experiences of uncertainty in the diagnostic process: Current understanding and future directions”.  Patient Education and Counseling, Volume 104, Issue 11, November 2021, pages 2606-2615
3.G.H. Gordon et al, “Physician expressions of uncertainty during patient encounters”. Patient Educ Couns, 40. 2000, pp. 59-65, 10.1016/S0738-3991(99)00069-5

NOTE FROM CAROLYN:   I wrote much more about the many kinds of uncertainty women face through diagnosis, treatment and recuperation in my book, A Woman’s Guide to Living with Heart Disease. You can ask for it at your local library or favourite bookshop, or order it online (paperback, hardcover or e-book) at Amazon, or order it directly from my publisher, Johns Hopkins University Press (use their code HTWN to save 30% off the list price).


Q:  At what point along that diagnostic process did you feel most uncertain?

5 thoughts on “Diagnostic uncertainty: when we just don’t know

  1. A month ago I went to the hospital in the middle of the night with a high fever and very violent chills in an ambulance.

    Now picture this: I had the widowmaker 5 years ago with minor damage and no more symptoms. Last year I was diagnosed with severe anemia and I have gotten iron infusions. I also have Type 2 diabetes with insulin under control and no side effects. I have no other issues. I am 72 and don’t even get colds and never had the flu. I got both COVID vaccines last spring.

    I give the ER docs credit for all the testing they did. Troponin was very high. But the monitors showed normal heart rate. I did not have COVID. My blood sugars were normal and low. BP was low. I had an iron infusion the day before.

    In the end, it turned out to be sepsis from a bladder infection with no symptoms and no blood in urine, E-coli was the culprit. I didn’t lose my mind with fever and was able to walk to the bathroom after my fever went down.

    Once on a floor, they gave me a walk around monitor for my heart. So every 4 hours I got an IV with fluids and antibiotics but in between I was free to wear my own PJ’s. I kept getting fevers at night, so I stayed for a week. No other patient on the acute medicine ward could get out of bed.

    They said my tests showed severe systemic reaction, yet after the first day I was up and walking. In these days of overcrowded hospitals, you don’t get to stay if you are not critical.

    If I had not gotten treatment so fast, they said I could have died within hours. Sometimes they do get it right!!!!

    Liked by 2 people

    1. Hello Chris – what a month you’ve had! It turns out that some studies suggest that cardiac troponins are high in about 85% of patients with sepsis, even with no other signs of acute coronary syndrome. You dodged a bullet there – sepsis is nasty. So many things could have gone south very fast – I’m glad you pulled through, and especially glad you called 911 in the middle of the night. Trusting your gut probably saved your life. I hope you are feeling much better now.

      I think you’ve officially reached your maximum lifetime quota of high-drama medical crises by now – you can relax and just live life from now on, okay? 😉

      Take care, stay safe. . . ♥


  2. I’ve been through so many diagnoses and processes. . . but the two most disconcerting were ones left open with no absolute sureness about anything.

    1) A persistent night time cough, fully researched but no “apparent” cause. I was given tessilon pearls, a medicine to suppress the cough. I am still concerned there is something pathological going on that wasn’t found.

    2) I was recently diagnosed with a specific kidney disease that is associated 30% of the time with a malignancy somewhere in the body. So the doctors sent me for CT scans of my entire body, colonoscopy and mammograms. Nothing was found. I still wonder if they missed something due to the limits of our testing?

    I often tell people, you know bacteria existed before we invented the microscope. Isn’t quite possible we have just not invented yet some of the tools we need for accurate diagnoses?

    Liked by 2 people

    1. Hello Jill – I hear ya! As the Texas research team also wrote, “… when diagnostic test results appear “normal”, patients may fear something serious was missed.” Of course they do – an understandable and common response.

      When an obvious culprit is not identified via testing, as in your cough, it can mean going down a rabbit hole, so the next most expedient move from clinicians is often to address the symptoms only, in the hope that things will somehow settle down on their own. Sometimes this does happen, sometimes we never do solve the mystery of what caused this.

      The kidney disease/malignancy scenario is more complicated and worrying, because of that lack of sureness you mention. Even with the range of sophisticated diagnostic tests already done, it’s hard to feel reassured. There’s very little we can be truly sure of in life, but what you do know so far is that 70% of the time, this kidney diagnosis does not result in a malignancy. Perhaps holding onto that – although not a sure thing – is better than focusing on the 30%?

      Good luck to you – stay safe. . . ♥


Your opinion matters. What do you think?

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s