Decontextualised risk information is any information pertaining to diagnosis, which is introduced into a clinical consultation, or a diagnostic thought process, without being requested by the clinician. It can be risk scores, computerised warnings, or lab tests or diagnostic imaging requests ordered by other clinicians. It is a concept which (to our knowledge) has not been studied.

Clinical vignettes have been developed to simulate scenarios where decontextualised risk is introduced into a consultation. These will be presented to primary care clinicians via video, who will then undergo a structured interview to explore how their perception, processing and communication of risk is challenged and changed by decontextualised risk. The findings will be analysed used qualitative methods (Thematic Analysis).

The standard model of a medical consultation involves a patient describing their symptoms to a clinician who then performs a physical examination. At this point a decision is made with the patient whether further information regarding diagnosis is needed: this is often further tests, but can be "risk scores", where various metrics including observations, age etc combine to produce an output which helps clarify the diagnosis in question. This moment, when a clinician decides if further tests or scores (or "risk" information) should be done is significant. In several studies, the eventual discovery of severe disease (such as cancer) has been found to be associated to a greater degree with the decision to do a test, then the actual outcome of the test itself.

If a clinician has not requested this new risk information, this could present a challenge to that professional's clinical thinking. I have called it "decontextualised risk information": risk information that is presented out of sync with when a clinician would normally gather such information, and out of the context of the normal diagnostic thought processes. Examples include the processing of tests a clinician has not ordered, the routine use of scores such as the NEWS2, and computer triggered electronic risk scores.

In primary care the chance of a serious disease being present is less than in A&E or specialist settings. All tests are wrong some of the time and can produce false negative results (when the test is negative but the disease is there) and false positive results (when a test is positive, but the disease is not present). This second situation becomes more prevalent the rarer a condition is to begin with. It means that, in primary care, the decision not to perform a test or diagnostic score is both more common, and more likely to be a reasonable decision.

This study will examine how "decontextualised risk information" affects clinicians' consideration of diagnostic risk and uncertainty. Clinical vignettes have been developed to simulate scenarios where decontextualised risk is introduced into a consultation. These will be presented to primary care clinicians via video, who will then undergo a structured interview to explore how their perception, processing and communication of risk is challenged and changed by decontextualised risk. The findings will be analysed used qualitative methods (Thematic Analysis).

Inclusion Criteria:

• GPs working in primary care in the UK are eligible

Exclusion Criteria:

• No current direct patient contact in a primary care setting

Given the possibility that there may be variable responses from clinicians from different backgrounds, experience and type of practice, the sampling will be purposeful. It will aim to include the following "profile" mix:

Gender Early career and late career clinicians Small and large practice size Part time and full time Self-described high and low risk tolerant GPs

Interview transcripts may be shared, together with the analytical coding framework, on application to the study team.