How and why do doctors communicate diagnostic uncertainty: an experimental vignette study

Why it matters

It’s not unusual for doctors to be uncertain about the exact cause of a patient’s symptoms. However, there hasn’t been much research into how, if at all, doctors should convey this uncertainty to patients.

Because diagnostic uncertainty is so common, it follows that doctors need to learn how to manage it, and how best to communicate their thoughts to patients. Doctors use various methods to manage uncertainty, including ‘ignorance-focused’ strategies (e.g. requesting further investigations) which aim to reduce uncertainty, and ‘relationship-focused’ strategies (e.g. communicating uncertainty to patients), which try to mitigate the effects of the uncertainty on patients, as opposed to trying to reduce it. There have been calls from researchers and regulatory bodies for diagnostic uncertainty to be communicated more openly, but the recommendations aren’t particularly evidence-based, and it’s been suggested that more systemic empirical research is required. In this study, we wanted to address the issues around diagnostic uncertainty, illustrate some of the ways that doctors communicate it to patients, and explore their reasoning for doing so.

Our approach

We created four written vignettes that featured clinical scenarios involving significant diagnostic uncertainty. These vignettes were developed using the clinical expertise of the authors, established clinical guidelines, and expert input from consultants in relevant specialties.

We recruited participant doctors from five hospitals and asked them to read the vignettes online, in a randomised order, and then tell an interviewer exactly what they would tell a typical patient.

We followed this up with semi-structured interviews which delved further into the reasons for communication choices in each scenario. We asked the participants why they decided to communicate the diagnostic uncertainty in the way they did, what teaching they had received in communicating diagnostic uncertainty, and how realistic they thought the vignettes were.

What we found

This study shows that diagnostic uncertainty is communicated differently by different doctors, even when they are presented with identical clinical information in a controlled setting. There was a marked difference in the ways that they approached discussing possible diagnoses, to what extent they communicated their uncertainty about a possible diagnosis, and how much they acknowledged diagnostic uncertainty when they were ‘safety-netting’ – that is advising patients about what to do if their condition failed to improve, changed or they were worried about it in the future. Safety-netting was common across all vignettes but was often carried out without any explicit reference to the risk of diagnostic error.

It was more common for expressions of uncertainty to be implicit than it was for them to be voiced explicitly. Participants expressed different communication goals (including reducing patient anxiety, building trust, empowering patients, and protecting against diagnostic errors), and different perspectives on how to best achieve these goals.

Doctors had different ideas about what the impact of communicating diagnostic uncertainty to patients might be. For example, although some felt that communicating uncertainty would cause patients to worry, others felt that open discussion about the diagnosis – including any related uncertainty – would help to relieve patient worry.

Although we asked participants to imagine they were talking to a typical patient, in the follow-up interview some participants said in practice that they would often be guided by the patient themselves in deciding what to communicate. Some doctors said that a lot depended on the individual’s perceived ability to handle the information they were being given. For example, it was felt that older patients with more experience of the healthcare system might be more likely to accept an uncertain diagnosis than younger patients. Although there may be logic in this reasoning, there is potential here for assumptions to be made that could magnify existing inequalities if doctors make presumptions about what patients want to be told based on characteristics such as their age, ethnicity, or perceived education level.

We found that most of the doctors reported they had not had any specific training in the communication of diagnostic uncertainty, and that many doctors felt that some training in this area would be useful.

This study emphasises the importance of research into patient perspectives, and how uncertain diagnoses might affect them. This information may help in finding the right balance in communication of uncertainty in diagnosis, and it may contribute to further ethical and legal discussions about when and how much information doctors should disclose regarding uncertainties in the diagnostic process.