Dr Win Huang’s Clinical Reflections（19） | The Hardest Diagnosis to Challenge May Be Your First One
By Dr Win Huang, PhD (Chinese Medicine & Acupuncture)
Recently, I assessed a 39-year-old woman who had experienced left shoulder pain for about six months. Her movement had gradually become restricted, and an ultrasound reported thickening of the subacromial bursa. A specialist had suggested a local injection.
When a patient presents like this, starting with the shoulder itself is entirely reasonable. That was also where I began. I examined the rotator cuff muscles and tendons, performed resisted tests, palpated the surrounding tissues and assessed active and passive shoulder movement.
There were abnormalities. Several areas around the shoulder were tender, movement was limited by pain, and the ultrasound had already identified a structural finding.
At this point, it would have been easy to form a coherent explanation: shoulder pain, local tenderness, restricted movement and an abnormal ultrasound. The pieces appeared to fit together.
A clinical loop had begun to close.
But sometimes that is exactly when we need to be most careful.
We Naturally Look for Evidence That Supports Our First Impression
The longer I practise, the more I recognise a simple tendency in clinical reasoning. Once we form an initial diagnosis, we naturally begin looking for evidence that supports it.
If I believe the rotator cuff is the problem, I pay more attention to rotator cuff findings. A tender point becomes another piece of supporting evidence. An imaging abnormality strengthens the impression further. Gradually, new information is absorbed into the original explanation.
There is nothing inherently wrong with forming a diagnostic framework. Clinicians cannot remain indefinitely between dozens of possible diagnoses. At some stage, we have to decide which explanation best fits the available evidence.
The important question is what happens next:
Once the clinical loop begins to close, are we still willing to notice evidence that does not support it?
Sometimes the Most Valuable Findings Are the Ones That Do Not Fit
What made me continue examining this patient’s shoulder pain was not an absence of shoulder findings. She did have local abnormalities.
The problem was that some findings did not fit particularly well.
Local tenderness did not reliably reproduce the pain she normally experienced. Resisted tests produced little of her characteristic pain. Her movement was restricted, but passive examination did not feel like a true mechanical block.
Each inconsistency could have been explained away. Clinical findings are rarely perfect, and it would have been possible to keep the original shoulder diagnosis.
But I asked a different question:
Why don’t these findings fit?
I therefore widened the examination to include the cervical region, relevant peripheral nerves and the scalene–supraclavicular region. When this area was stimulated, her familiar anterior shoulder pain could be reproduced much more clearly. More importantly, while maintaining that stimulus and reassessing her shoulder, her pain and movement changed immediately.
The original local shoulder explanation was no longer sufficient.
Clinical Reasoning Is Not Only About Finding More Supporting Evidence
Medical training rightly teaches us to look for evidence. Does the history support the diagnosis? Do the physical findings support it? Does imaging support it?
But with experience, I have come to believe there is another equally important clinical skill:
Can I recognise what my diagnosis fails to explain?
A diagnosis may explain 80% of a patient’s presentation. That can feel convincing. But what about the remaining 20%?
Are those findings irrelevant background noise, or are they telling us that our diagnostic framework is too narrow?
Sometimes the finding that advances a diagnosis is not another positive test. It is the one symptom or sign that stubbornly refuses to fit the existing explanation.
A Larger Clinical Framework Can Still Become Another Trap
Different clinicians may construct very different diagnostic frameworks around the same shoulder pain.
One clinician may focus mainly on muscles and tendons. Another may consider the rotator cuff, joint capsule, bursa, cervical spine and peripheral nerves. A broader knowledge base generally allows us to consider more possibilities.
But a larger framework is not automatically a correct one.
In this case, I moved beyond a purely local shoulder explanation and considered the relationship between the scalene region, neural mechanical sensitivity and the patient’s shoulder symptoms. This framework explained more of the clinical findings and was further supported by the immediate change following treatment.
Even so, I cannot say that it represents the final truth.
If new symptoms appear, further examination produces contradictory findings, or treatment no longer produces the expected response, I must be willing to question this framework as well.
Otherwise, I have simply escaped one small diagnostic loop and entered a larger one.
Do Not Explain Away Contradictions Too Quickly
Paradoxically, clinicians may be most cautious when they know very little about a condition. The greater challenge comes when the history, examination and imaging already seem to form a convincing story.
Experience allows us to recognise patterns quickly. That is one of its greatest strengths. But the same experience can also make us stop thinking too early.
So after forming a diagnosis, I increasingly try to ask myself two questions rather than one.
Not only:
What evidence supports my diagnosis?
But also:
What does my diagnosis still fail to explain?
When something does not fit, we should not always rush to dismiss it as an irrelevant finding or force it into our existing theory. Sometimes it is worth leaving that contradiction unresolved for a while.
It may be the clue that eventually reopens the entire diagnosis.
Clinical reasoning requires us to build coherent explanations because eventually we must make decisions. But it also requires the ability to break those explanations apart when the evidence no longer fits.
Disease does not exist to conform to the clinician’s theory.
Perhaps one sign of clinical growth is not becoming faster at making diagnoses, but becoming less likely to be trapped by the first one.
中文微信：nzacupunctureclinic
Leave a reply