How Can a Layman Determine Whether a Medical Diagnosis Is Legitimate?
FndNope
Moderator
How can we know that anything we believe is true?
Even in areas where we have genuine expertise, by what standard do we measure our claims? What separates something we have defined from something we have actually demonstrated?
This is a question I have been assessing and reassessing since my early teenage years, when I first became involved in apologetics and, eventually, lost my faith.
It was during that period that I was introduced to terms such as ontology, epistemology and axiology, and to how easily these concepts can become confused in ways that obscure rather than clarify what we actually know.
Ontology, Epistemology and Axiology
Ontology deals with existence and the nature of reality. It asks what kinds of things exist, how we define them and how we categorise the things we observe.
Epistemology asks how we know. What is the evidence for our claims? How strongly does that evidence justify the conclusion?
Axiology asks why any of this matters. What value does the concept have, and what are the consequences of applying it?
I would argue that FND is ontologically strong, epistemologically weak and axiologically variable, with the last of these depending heavily on how the diagnosis is used by the clinician involved.
The central problem, as I see it, is that FND can blur ontology and epistemology. Characteristics used to define the disorder can begin doing double duty as evidence that the disorder itself has been demonstrated.
To simplify the problem, reduce it to three words:
Definition. Evidence. Value.
Definition
FND has a definition. Certain clinical patterns are recognised as characteristic of it, including the so called positive signs used to rule in the diagnosis.
The signs themselves are observable phenomena. That is not really the contentious part.
The epistemological question is:
What do those signs actually demonstrate?
A clinician may observe inconsistency, distractibility, entrainment or a positive Hoover sign. But moving from "this pattern occurred" to "this pattern demonstrates Functional Neurological Disorder" requires an additional inferential step.
That step deserves far more scrutiny than I believe it usually receives.
A sign can demonstrate that a particular phenomenon occurred during examination. Establishing why it occurred, what mechanism produced it, and how specifically it identifies one disorder rather than another are different questions.
This is where I believe the epistemic bar for FND becomes remarkably low.
We still do not know, in any comprehensive sense, why all of these phenomena occur. There are hypotheses, models and proposed mechanisms, but substantial uncertainty remains about the causal picture.
So the observation may be real while our explanation for that observation remains uncertain.
That distinction matters enormously.
Classification Is Not the Same as Explanation
At its strongest, an FND diagnosis may achieve something reasonably defensible: clinical classification.
A clinician observes a recognised pattern and concludes that the presentation fits what medicine currently defines as an FND phenotype.
But that is not automatically the same as establishing the cause of the patient's symptoms.
There is an important difference between saying:
"This patient demonstrates a recognised FND pattern."
and saying:
"We therefore know what is causing this patient's symptoms."
The first is a classificatory claim.
The second is a much larger explanatory claim.
And this, I think, is where much of the epistemological confusion around FND arises.
FND may be considerably stronger at identifying and classifying a pattern than at explaining why that pattern exists.
The epistemic problem begins when confidence in the classification is allowed to become confidence in the explanation.
Evidence
This is why calling something a positive sign or a rule in sign does not settle the epistemological question.
Those terms describe how the signs are used diagnostically. They do not, by themselves, establish how specific the signs are, what biological processes can produce them, or whether every competing explanation has been adequately characterised.
Consensus can tell us how medicine currently interprets an observation.
Consensus is not the same thing as demonstrating that the interpretation is ultimately correct.
This becomes especially important when the underlying causal picture remains incomplete.
If we observe phenomenon X, define X as characteristic of disorder Y, and subsequently use the presence of X as strong evidence that Y explains the patient's symptoms, then the strength of that conclusion depends on something beyond the definition itself.
We need to know how reliably X distinguishes Y from other possibilities.
We need to know its limitations.
We need to know what else can produce X.
And we need to know what evidence would cause us to reconsider Y.
Otherwise, classification begins masquerading as explanation.
The Risk of Circularity
Conceptually, we can define a category by grouping observations together. That does not make the category meaningless or useless. Medicine does this routinely, particularly where mechanisms remain incompletely understood.
The danger arises when the characteristics defining the category are subsequently treated as though they independently establish the explanatory construct attached to it.
At that point, there is at least a risk of circularity:
We know this is FND because the patient demonstrates the signs we define as FND signs, and we know those signs indicate FND because FND is defined by those signs.
That may establish that a patient satisfies a clinical definition.
It does not necessarily establish why the patient demonstrates those phenomena, whether our current explanation for them is correct, or whether another process could ultimately account for them.
This is the question I keep returning to:
Fine. Rule it in. But what, precisely, have you ruled in?
Have you ruled in a phenotype?
Have you ruled in a diagnostic category?
Have you ruled in a mechanism?
Have you ruled in a cause?
Or have several very different epistemic claims quietly been compressed into one label?
Value
And that brings us to axiology.
FND can have genuine clinical value.
When medicine does not yet have a complete causal explanation, recognising patterns can provide patients and clinicians with a framework for treatment rather than simply saying, "We don't know."
But the value of that framework depends enormously on how its uncertainty is handled.
If the diagnosis remains open to revision when the clinical picture changes, progresses or produces contradictory evidence, the category may remain useful.
If instead it becomes a concrete fact through which every subsequent symptom is interpreted, it risks closing diagnostic doors in the name of preserving the original diagnosis.
At that point, its value can collapse.
The ontology has swallowed the epistemology.
We have confused having a name for something with knowing what it is.
When Practical Certainty Outruns Evidential Certainty
This is where the problem becomes more than philosophical.
A relatively modest claim such as "this presentation resembles a recognised functional pattern" may, in practice, become operationalised as the much stronger claim "we now know what is wrong."
Those are not equivalent.
And if the stronger claim changes what happens next, then the distinction matters clinically.
If an FND label encourages clinicians to investigate less, lowers suspicion of alternative pathology, or causes subsequent abnormalities to be interpreted through the original diagnosis, then practical certainty may have outrun evidential certainty.
That, to me, is one of the most important problems in the entire debate.
Why This Feels Familiar
Strangely, this is where my experience with apologetics starts feeling familiar.
In apologetics, I repeatedly encountered arguments in which the proposition requiring justification gradually became incorporated into its own justification. Assumptions generated conclusions that were then used to reinforce the original assumptions.
The analogy is not that FND is a religion or that clinicians are simply acting on faith.
The similarity I see is methodological:
Once a framework begins supplying part of the justification for itself, its evidential foundations become increasingly difficult to interrogate from within that framework.
And when people become heavily invested in a framework, criticism of its evidential foundations can start being interpreted as criticism of the framework's very right to exist.
But those are different questions.
Something can exist as a useful clinical category without every claim made about that category being equally well established.
A definition is not a mechanism.
A pattern is not necessarily an explanation.
Consensus is not demonstration.
A diagnosis is not necessarily a final truth.
And uncertainty is not a failure of medicine.
Sometimes it is simply the most epistemically honest answer available.
So What Can a Layman Actually Do?
None of this means that a layperson should pretend to be a neurologist.
You do not need to know more medicine than your doctor to ask whether the reasoning behind a diagnosis is sound.
The questions are much more fundamental.
What Exactly Was Observed?
Separate the observation from its interpretation.
"Your leg behaved differently during these two manoeuvres" is an observation.
"Therefore your symptoms are functional" is an inference.
Those are not the same thing.
How Specific Is the Evidence?
Does the finding occur almost exclusively in the condition being diagnosed, or can it occur for other reasons?
A finding can be associated with a diagnosis without uniquely establishing it.
And importantly, ask whether claims about specificity have themselves been adequately demonstrated rather than merely repeated within the diagnostic framework.
What Would Prove the Diagnosis Wrong?
This is perhaps the most important question.
If progression occurs, new objective abnormalities emerge, investigations contradict the original interpretation, or a better explanation becomes available, can the diagnosis be reconsidered?
A diagnosis does not need to be discarded merely because something unexpected happens. But there must be some conceivable evidence capable of lowering our confidence in it.
If effectively nothing can do so, we have moved away from an ordinary scientific hypothesis and towards an unfalsifiable framework.
How Much Is Known, and How Much Is Inferred?
There is nothing inherently wrong with uncertainty.
Medicine operates under uncertainty constantly.
The problem begins when uncertainty about mechanism and causation is communicated as certainty about what is happening to the patient.
"This pattern is recognised in FND" is not epistemically equivalent to "we know FND is the reason this is happening."
The confidence of the language should reflect the strength of the evidence.
What Happens When New Evidence Appears?
A good diagnosis should organise evidence, not protect itself from evidence.
New findings should be capable of increasing or decreasing confidence in the original diagnosis.
If every new symptom supports FND, while contradictory findings are simply incorporated into an ever expanding FND framework, we should reasonably ask what evidence could ever count against it.
That is not uniquely an FND problem. It is a basic problem of scientific reasoning.
But it becomes especially important when dealing with diagnoses capable of influencing how subsequent symptoms are interpreted.
What Are the Consequences of Being Wrong?
This is where epistemology meets axiology.
Not every diagnostic error has the same consequences.
The evidential threshold should matter enormously when a diagnosis substantially changes what happens next.
If a label encourages clinicians to stop investigating, reinterpret subsequent abnormalities through that label, lower suspicion of alternative pathology, or discourage further investigation, then the cost of being wrong becomes much greater.
Conversely, if the diagnosis is treated as a revisable working model that helps guide treatment while remaining responsive to new evidence, its potential value is considerably greater.
That distinction matters more to me than the label itself.
The Question That Ultimately Matters
A layperson therefore does not need to determine whether the neurologist is medically correct.
They can ask something more fundamental:
What was observed?
What has actually been demonstrated?
What remains inferred?
What would change the conclusion?
And do the consequences of accepting that conclusion justify the level of certainty being expressed?
Ultimately, it comes down to one question:
Does the Confidence of the Conclusion Actually Match the Strength of the Evidence Supporting It?
You do not need a medical degree to ask that question.
That question belongs to everyone.
