When Assumptions Become Facts

The Hidden Reasoning Errors Shaping Clinical Practice

Welcome to a new edition of the newsletter. I recently encountered a fascinating paper by Weisman and colleagues that examines how reasoning errors in medical literature can lead to assumptions being accepted as established knowledge. The findings are uncomfortable: prestigious experts can make fundamental errors in logic, journal editors sometimes fail to catch them, and entire treatment paradigms can be built on foundations that collapse under scrutiny.

As clinicians, we make dozens of clinical decisions daily, drawing on what we believe to be established medical knowledge. We diagnose conditions, explain mechanisms to patients, and choose treatments based on our understanding of pathophysiology. But what if some of these “facts” we rely on are actually unvalidated assumptions that slipped past peer review decades ago?

Understanding how to construct and critique logical arguments is essential for evaluating the literature and improving clinical decision-making. In this piece, I’ll try to walk you through the key principles of reasoning that every clinician should understand.

The Foundation: Understanding Arguments

In formal logic, an argument is the most basic unit of reasoning. It consists of:

• Premise(s): Starting point(s) that make a claim (either true or false)

• Conclusion: End-point that follows from the premise(s)

For an argument to be sound, two conditions must be met:

1. It must be valid (the conclusion logically follows from the premises)

2. The premises must be true

Example of a sound argument:

• Premise: All patients with bacterial pneumonia have elevated white blood cell counts

• Premise: This patient has bacterial pneumonia

• Conclusion: Therefore, this patient has an elevated white blood cell count

The argument is valid (the conclusion follows logically) AND the premise is true, making it sound.

Example of an invalid argument:

• Premise: Most female patients with knee pain have increased Q-angle

• Premise: This patient has knee pain

• Conclusion: Therefore, this patient's pain may be because of increased Q-angle

This is invalid because the conclusion doesn’t necessarily follow from the premise. “Most” doesn’t mean “all.” Now, let’s look at common logical fallacies in clinical reasoning!

Common Logical Fallacies in Clinical Reasoning

Before exploring the three reasoning processes, let’s identify some common errors that can undermine our clinical thinking:

1. Begging the Question (Circular Reasoning)

This occurs when the premises of an argument assume the truth of the conclusion. You’re using what you’re trying to prove as part of your proof.

Example:

• “This patient has pain because of tissue inflammation”

• “We know there’s tissue inflammation because the patient has pain”

Pain is both the evidence for inflammation and the consequence of inflammation. This is circular reasoning.

2. Reification

Treating an abstract concept or experience as if it were a concrete, physical entity capable of causal action.

Example: “The pain is causing muscle guarding, which causes more pain.”

Pain is an experience, not a physical entity that can cause things. The underlying physiological processes may create both pain and muscle tension, but the subjective experience of pain itself doesn’t have causal power.

3. Post Hoc Ergo Propter Hoc

Latin for “after this, therefore because of this.” Assuming that because B followed A, A must have caused B.

Example: “I performed joint mobilisation, and the patient’s pain improved. Therefore, the mobilisation caused the improvement.”

This ignores alternative explanations, such as natural history, regression to the mean, placebo effects, therapeutic alliance, or other concurrent interventions.

4. Unfalsifiability

Formulating claims in ways that make them impossible to disprove, removing them from the realm of scientific inquiry.

Example: “The treatment works by releasing fascial restrictions. If the patient improves, it proves the treatment worked. If they don’t improve, the restrictions were too severe, or they need more treatments.”

This claim cannot be proven wrong, which means it cannot be tested scientifically.

5. Appeal to Authority

Accepting claims based on the prestige or reputation of the person making them, rather than the strength of the evidence.

Example: “Dr. X is a world-renowned expert with 40 years of experience, so this treatment must be effective.”

Expertise and experience don’t guarantee correctness. Evidence does.

The Three Reasoning Processes in Clinical Medicine

Medical reasoning employs three main logical processes. Understanding their strengths and limitations is crucial:

1. Deductive Reasoning

Definition: Moving from general premises to specific conclusions. If the premises are true and the logic is valid, the conclusion must be true.

Structure:

• General principle (premise)

• Specific case (premise)

• Inevitable conclusion

Clinical Example:

• All patients with diabetic ketoacidosis have elevated blood glucose (>250 mg/dL)

• This patient has diabetic ketoacidosis

• Therefore, this patient’s blood glucose is >250 mg/dL

Strength: Provides certainty when premises are true

Limitation: Requires premises that are actually true and universal. Biological systems rarely follow universal rules without exceptions.

Clinical Application: Deduction is powerful but limited in practice because most of our “all patients with X have Y” statements have exceptions. We use deduction most reliably when dealing with well-established physiological facts (e.g., “all living humans have a heart,” “all diabetic ketoacidosis cases involve insulin deficiency”).

2. Inductive Reasoning

Definition: Generalising from specific observations to broader conclusions. This is the “logic of experience.”

Structure:

• Multiple specific observations

• Pattern recognition

• General conclusion (with probability, not certainty)

Clinical Example:

• I’ve examined 50 patients with acute ankle inversion injuries

• 48 of them had lateral ligament involvement

• Therefore, the next patient with an acute ankle inversion injury probably has lateral ligament involvement

Strength: Allows us to learn from experience and recognise patterns

Limitation: Delivers conclusions with probability, not certainty

Critical Insight from the Paper: “What was observed to have happened in the past cannot be used as a guide to what has yet to occur.” This is especially problematic with biological processes that are hidden from view and not necessarily uniform or predictable.

Clinical Application: Induction is essential in clinical practice—it’s how we recognise patterns and make probabilistic predictions. But we must remember we’re dealing with likelihoods, not certainties. The 51st patient may be different from the previous 50. This is why clinical reasoning requires flexibility and willingness to revise our expectations when patients don’t fit the pattern.

3. Abductive Reasoning

Definition: Generating the most plausible hypothesis to explain a set of observations. It’s “inference to the best explanation.”

Structure:

• Observe an unexpected or puzzling phenomenon

• Consider multiple possible explanations

• Select the most plausible hypothesis based on existing knowledge

• Crucially: Test the hypothesis

Clinical Example:

• A patient presents with severe headache, neck stiffness, photophobia, and fever

• Multiple conditions could cause these symptoms (meningitis, subarachnoid haemorrhage, severe migraine, etc.)

• Meningitis is the most plausible explanation given this constellation of findings

• I perform diagnostic tests (lumbar puncture, imaging) to confirm or refute this hypothesis

Strength: Helps generate hypotheses when encountering unfamiliar or complex presentations

Limitation: The “most plausible” explanation isn’t necessarily the correct one

The Critical Error: Stopping at “this is the most plausible explanation” and failing to rigorously test the hypothesis. This is where many clinical constructs go wrong.

Clinical Application: Abduction is extremely useful—it’s how we generate differential diagnoses and clinical hypotheses. But it’s the starting point, not the endpoint. The hypothesis must then be tested using the hypothetico-deductive method.

The Gold Standard: Hypothetico-Deductive Method

Karl Popper articulated the scientific method that should govern medical research and clinical decision-making:

The Process:

1. Start with a hypothesis and a set of given conditions

2. Deduce what facts would follow if the hypothesis were true

3. Test through experiments whether those facts actually hold

4. Refute the hypothesis if observations contradict predictions

Critical Principle: Only attempts at refutation can lead to advances in knowledge.

As Popper stated: “If observation shows that the predicted effect is definitely absent, then the theory is simply refuted.”

What This Means: We don’t just look for evidence that confirms our hypothesis. We actively try to prove it wrong. If it survives rigorous attempts at refutation, we gain confidence in it. If it fails even one crucial test, it must be rejected or revised.

Clinical Example:

Hypothesis: “Anterior cruciate ligament (ACL) injury causes knee instability during cutting movements.”

Deductions:

• If this is true, patients with complete ACL tears should demonstrate instability during pivot-shift testing

• If this is true, patients should report episodes of the knee “giving way” during direction changes

• If this is true, surgical ACL reconstruction should eliminate instability

Testing:

• Perform pivot-shift test (positive in most ACL injuries)

• Assess functional instability reports (positive in many but not all patients)

• Assess outcomes after ACL reconstruction (most patients report stability improvement)

The hypothesis survives these tests, so we have confidence in it. But notice: if a significant number of patients with complete ACL tears showed no instability, we’d need to revise our understanding.

The Role of Assumptions in Clinical Practice

An assumption is anything accepted as true without proof. In formal logic, assumptions are premises that we haven’t validated.

Why Assumptions are Necessary:

As Augustine of Hippo noted 1,600 years ago:

“I began to realise that I believed countless things which I had never seen... Unless we took these things on trust, we should accomplish absolutely nothing in this life.”

“We just can’t investigate everything, and for that reason, we are forced to rest content with an assumption.” We cannot investigate every claim from first principles. Some assumptions are necessary for practical function.

The Problem Arises When:

1. Assumptions are not recognised as assumptions

2. They are not converted into testable hypotheses

3. They are passed off to others as established knowledge

The Paper’s Warning: “A challenging situation arises in Medicine when those who make assumptions neglect to [test them]. They come to believe they are statements of truth and pass them off to others as established knowledge.”

A Practical Example: Examining a Clinical Claim

Let’s apply these concepts to analyse a common clinical scenario and see how reasoning errors can occur.

The Claim: “Poor posture causes neck pain. Correcting postural alignment eliminates the underlying cause and resolves the pain.”

This seems plausible and is widely taught. But let’s examine the reasoning:

Is this deductive reasoning?

Attempted argument:

• Premise: All cases of neck pain are caused by poor posture

• Premise: This patient has neck pain

• Conclusion: This patient’s pain is caused by poor posture

Problem: The first premise is not true. Many people with “poor” posture have no pain, and many with “good” posture have pain. This invalidates the deduction.

Is this inductive reasoning?

Attempted argument:

• I’ve treated 100 patients with neck pain using postural correction

• 65 improved

• Therefore, poor posture causes neck pain

Problem: This commits the post hoc ergo propter hoc fallacy. Improvement after treatment doesn’t prove the theoretical mechanism. We haven’t controlled for:

• Natural history (many neck pains resolve spontaneously)

• Regression to the mean (patients often seek care when symptoms are at their worst)

• Placebo effects and therapeutic alliance

• Other effects of the intervention (movement, attention, reassurance)

Is this abductive reasoning?

Attempted argument:

• This patient has neck pain

• Poor posture could explain it (it’s a plausible mechanism)

• This is a reasonable explanation

• Therefore, I’ll treat the posture

Analysis: This is where abduction should transition to hypothesis testing, but often doesn’t.

Questions we should ask:

1. Has “poor posture” been clearly defined in operational terms?

2. Does poor posture reliably precede the development of neck pain in longitudinal studies?

3. Is there a dose-response relationship (worse posture = worse pain)?

4. Does correcting posture reliably reduce pain in controlled trials?

5. What is the proposed mechanism by which static alignment causes dynamic pain experiences?

Current Evidence: Research has largely failed to find consistent relationships between postural measures and neck pain. Studies show:

• Cross-sectional studies find weak or no associations between postural measures and pain

• Longitudinal studies don’t show posture predicting pain development

• Interventions targeting posture don’t consistently outperform other treatments

Additional Concerns

Is this reification? Are we treating “posture” (a static measurement of joint angles) as if it were a concrete entity capable of causing pain, when pain is a complex neurobiological and psychological experience?

Is this circular reasoning?

• “Poor posture causes pain”

• “We know the posture is poor because the patient has pain”

Is the claim falsifiable? What evidence would disprove the posture-pain connection? If we can’t specify this, we’re not making a scientific claim.

What Should Happen

The posture-pain hypothesis should be:

1. Clearly formulated with operational definitions

2. Tested in well-designed studies that attempt refutation

3. Accepted, rejected, or revised based on evidence

This doesn’t mean postural interventions are useless—it means the theoretical rationale needs scrutiny. Maybe they work through other mechanisms (movement, self-efficacy, attention). But we should be honest about what we know versus what we assume.

Why These Errors Persist: The Snowball Effect

Beyond the initial logical failures, a secondary problem that compounds the issue is the citation snowball effect.

How It Works:

• An influential author publishes a plausible-sounding hypothesis

• Other researchers cite it (often without critically evaluating the reasoning)

• The paper accumulates citations

• More researchers cite it because it’s highly cited

• High citation counts create the appearance of validity

This creates an “illusory truth effect”—repeated exposure to an idea makes it seem more true, independent of actual evidence.

Clinical Impact: Ideas become entrenched in clinical practice not because they’ve been proven correct, but because they’ve been repeated frequently. Citation count becomes confused with evidence quality.

Three Principles for Scientific Rigour

To prevent assumptions from being passed off as established knowledge, we should apply three logical principles rigorously:

1. Assumptions Should Lead to Testable Hypotheses

• Not circular arguments

• Not unfalsifiable claims

• Clear predictions that could be proven wrong

2. Repeated Observations Don’t Validate a Hypothesis

• Confirmation requires attempting refutation, not accumulating confirmatory anecdotes

• 100 patients improving after treatment doesn’t prove the mechanism

• We need controlled studies that could prove the hypothesis wrong

3. Assumptions Become Facts Only After Rigorous Testing

• The burden of proof lies with the proponent; it is their responsibility to demonstrate that a claim is supported by robust evidence, not for others to disprove it.

• “This seems plausible” is not sufficient

• Hypotheses must survive attempts at refutation

Conclusion

The Weisman paper reveals an uncomfortable truth: prestigious experts can propagate significant reasoning errors, peer review sometimes fails to catch them, and entire fields of practice can be built on unvalidated assumptions.

This matters because:

• Patients undergo treatments based on faulty reasoning

• Clinicians make decisions based on “facts” that may be assumptions

• Research builds on foundations that may be unsound

• Resources are spent on interventions that lack scientific justification

The lesson for clinicians:

Critical thinking is not optional; we must actively question the assumptions, evidence, and reasoning that underpin our clinical decisions.

• Scrutinise both content AND reasoning in the literature

• Distinguish between what is assumed and what is known

• Recognise that authority and prestige don’t guarantee correctness

• Remain open to the possibility that cherished concepts may rest on fallacies

As Christopher Hitchens stated: “What can be asserted without evidence can also be dismissed without evidence.”

The question is whether we have the intellectual courage to apply this principle to our own clinical beliefs.

Here are a few questions to reflect on, and I would genuinely value your thoughts and perspectives as you consider them in the context of your own clinical practice

• What clinical concepts in your field might rest on unvalidated assumptions rather than rigorous evidence?

• How can you distinguish between plausible explanations (abduction) and established facts in your clinical practice?

• What role does prestige bias play in your acceptance of clinical concepts? How can you guard against it?

• How should we communicate uncertainty to patients when much of what we “know” may be an assumption rather than a fact?

• What systems could we create in clinical practice and research to catch logical fallacies before they become entrenched?

Reference

Weisman A, Quintner J, Galbraith M, Masharawi Y. Why are assumptions passed off as established knowledge? Medical Hypotheses 2020;140:109693.

Thanks for Reading!

Ammar