In textbooks, guidelines and letters from specialists to generalists, life can seem clear cut. Symptoms, signs and tests all mesh together; a diagnosis appears. Everyone is relieved, except the poor patient who must wrestle with the consequences.

In Primary Care, it is not like that; patients may come with some, any, none or all of the symptoms of not just one, but several conditions. The ANALYSING quadrant of the Map of Uncertainty in medicine can help here. This addresses the WDYDWYDKWTD moments that arise when one doctor and patient are working together and they still don’t have a working diagnosis.

Agnes is a 36 year old kick boxer. She often competes and accepts the bumps and bruises that follow. She ignored an achy pain in her left jaw for a few days, putting it down to a fall in a bout she lost. It didn’t get better, she started to feel unwell, with a slightly sore throat. A friend noticed a swelling in Agnes’ neck and recommended a visit to the GP.

This unusual set of symptoms may give the clinician a WDYDWYDKWTD moment! Is her jaw broken? Is the swelling an abscess, a haematoma, the start of a lymphoma? Why does Agnes feel vaguely ill? It is tempting to take rapid “dysfunctional ways out” here.  As she is not that unwell, reassurance that all will settle with time may seem appropriate.  An X-ray in the A and E department to rule out a fracture uses referral as a quick way out. When the X-ray is normal she will go home none the wiser about her problem. Avoiding a diagnosis and saying vaguely, “it’s probably a virus”, or giving an antibiotic “just in case” are other “dysfunctional ways out”.

Better to use the skills of accurate hypothesis generation, thinking rationally about the differences between ruling things in or ruling them out. Marshall Marinker famously characterised the work of the generalist as being to “marginalise danger, explore probability and tolerate or manage uncertainty” (reference 1)

Could the worst case scenarios here be ruled out?  The trauma cannot be discounted; is there a head injury, jaw fracture? Could the swelling be something serious, an early lymphoma perhaps?

As for probabilities, a fracture was not that likely; although stoical, Agnes would probably have sought help immediately if she had pain on eating, with swelling or bruising over the jaw. The neck swelling may indicate inflammation; is there a local infection? Or are there, as yet unnoticed, swellings elsewhere, in the axillae, groin or liver?

Clinical reasoning can help in deciding where to go next Doctors are keen to rule things OUT, because it reduces worry about serious issues, if they can be reasonably ruled out. Patients are naturally keener for answers and would like to know what has been ruled IN. Examination helped here.

Agnes was fully functioning and alert and had had no period of unconsciousness or head trauma. With no bruising or localised tenderness over the jawbone it was reasonable to “rule out” trauma as the cause for her current problems.

The neck swelling felt like a reactive lymph node and the absence of other enlarged glands, spleen or liver, made a lymphoma less likely. This left some inflammatory possibilities. Careful examination of Agnes’ face revealed a few subtle pinky red spots, all on the same side of the cheek, her skin being otherwise healthy. The ears and throat appeared normal and Agnes repeated that she was not her usual fit and active self.

Could anything be ruled IN? The spots were nonspecific. Pinkish, small and perhaps a bit tender. Was the gland in the neck reacting to an infection there and if so what kind?  The malaise and sore throat could indicate some kind of mild viral process, but that would not explain the spots nor the unilateral lymph nodes. Perhaps the pain was early shingles.

So in the WDYDWYDKWTD moment, what could the clinician do next? When things are not clear cut, follow up and reassessment are powerful diagnostic tools for managing uncertainty proactively. When handled correctly, this can build a strong doctor patient relationship and therapeutic alliance.

When these thinking processes were articulated explicitly with Agnes, she was happy to accept some pain relief and an early telephone review. 36 hours later her face was looking awful, “as you said doctor, things might change quickly”. When she returned, the diagnosis became obvious; there was a patch of shingles across her cheek.

Using explicit reasoning skills and timely reassessment, unnecessary X-rays, antibiotics and an Accident and Emergency attendance were avoided. Uncertainty is an important sign in clinical practice; holding uncertainty and working through it methodically can yield effective results.

• Marinker M. Medical Audit and General Practice. London 1990