Clinical Medical Reasoning

January 20, 2026 Health & Healing No Comments

Clinical medical reasoning is central to medicine, yet it often remains implicit and taken for granted.

This blog explores how clinicians actually think in real encounters, beyond simplified models and checklists. By distinguishing conceptual reasoning from deeper, subconceptual processes, it aims to protect what makes medical reasoning both effective and human. The focus is not on changing practice, but on seeing it more clearly.

Seeing what is already happening

Clinical medical reasoning is everywhere in medicine, yet it is rarely examined in itself. We speak extensively about evidence, guidelines, errors, and increasingly about artificial intelligence. Much less attention is given to the ongoing process by which clinicians make sense of what unfolds before them. This blog is an invitation to pause and look more closely.

Clinical reasoning is not a technique that can be switched on or off. It is happening, continuously, often beneath the surface of awareness. Understanding it better does not make it more mechanical. On the contrary, it helps protect what is most human in medicine.

The present situation

Clinical reasoning begins long before any diagnosis is considered. It starts in the encounter itself. A patient does not arrive with a label, but with a lived situation that asks to be understood. What is presented is rarely a textbook entity. Much more often, it is an individual pattern: a constellation of sensations, concerns, meanings, and expectations that only partly fit established categories.

In this sense, patients consult with patterns, not diagnoses. These patterns may be vague, shifting, or deeply personal. They may resist clear naming. Listening to them requires more than checklist thinking. It requires a kind of attention that goes beyond explicit symptom collection, something like a depth stethoscope.

When this depth is missed, clinical reasoning risks starting too late, already narrowed by premature categorization. The encounter then becomes an exercise in fitting rather than understanding.

Two layers of thinking

Clinical medical reasoning unfolds on at least two interconnected levels:

One is the conceptual level. This is the level of explicit concepts, causal models, differential diagnoses, guidelines, probabilities, and documented plans. It is verbalizable, teachable, measurable, and essential. Without it, medicine could not function as a science, a profession, or a shared practice.
The other is the subconceptual level. This level is pattern-based, predictive, and largely non-conscious. It includes intuition, a sense of coherence or unease, recognition without immediate explanation, and the feeling that something fits or does not fit yet. It is here that much of the real-time integration happens. See the addendum table for how clinical medical reasoning fits closely with features of subconceptual processing.

Physicians tend to overestimate the conceptual part of their own thinking. This is understandable. Conceptual reasoning is what can be explained, justified, and presented to others. It is also what medical education emphasizes and evaluates. Yet much of clinical competence arises before, beneath, or alongside explicit concepts.

Recognizing this does not diminish conceptual reasoning. It places it where it belongs.

Science and its limits

Science necessarily privileges the conceptual. It must define, formalize, and communicate explicitly. This is one of its greatest strengths. Problems arise when this methodological necessity is silently mistaken for completeness.

Clinical medicine does not operate in a laboratory vacuum. It unfolds in human encounters marked by uncertainty, individuality, and meaning. Acknowledging the conceptual is essential. Acknowledging only the conceptual is insufficient.

When medicine forgets this, it risks mistaking what can be measured for what matters most.

How reasoning actually happens

In everyday practice, clinical reasoning is rarely linear. It is hypothesis-driven, anticipatory, and continuously revised. Early impressions guide questions. Questions reshape impressions. New information may confirm, nuance, or unsettle what was already forming.

Often, a clinician knows before being able to explain why. Competence precedes comprehension. This is not sloppy thinking. It reflects how human intelligence works, as explored in greater depth in What Is Intelligence?.

From a neurocognitive perspective, the brain functions as a prediction machine, continuously anticipating and updating. Clinical reasoning is an expression of this, as described in The Brain as a Predictor. What J. Kassirer described phenomenologically in Learning Clinical Reasoning (first ed. 1989) now finds deeper explanatory grounding.

Conceptual explanations usually come after the fact. They are translations, not the original language of thought.

Bias reconsidered

Bias is often portrayed as a contaminant of otherwise pure reasoning. From this perspective, the task is to eliminate it. From a more realistic view, bias is intrinsic to predictive, pattern-based cognition. It is not an external flaw, but a built-in feature.

This does not make bias harmless. It makes it unavoidable. As discussed in Bias is Our Thinking, the goal is not bias-free reasoning, but reasoning that is aware, reflective, and balanced.

Conceptual reasoning does not remove bias. It provides a space in which bias can be noticed, questioned, and sometimes corrected.

The intelligence of ‘not yet’

Clinical reasoning is often evaluated by the decisiveness of its decisions. Yet some of the best reasoning consists in sensing that a decision would be premature. Experienced clinicians often recognize moments when the picture has not yet declared itself. Acting too soon may foreclose understanding. Time itself can be diagnostic. This is not indecision. It is temporal intelligence.

Knowing when ‘not yet’ is the most responsible answer is one of the highest forms of clinical reasoning. It requires trust in deeper pattern recognition, alongside conceptual vigilance.

Patterns, meaning, and dynamics

Many clinical problems are not static entities but evolving patterns. These patterns are often saturated with meaning. Reducing them too quickly to abstract variables such as ‘stress’ risks missing what truly drives the situation, as explored in Not Stress but Meaning is a Cause of Disease.

Moreover, patterns can become self-reinforcing. Symptoms, interpretations, emotions, and physiological responses may form whirlpools that deepen over time. Understanding this dynamic is crucial for sound reasoning, as described in Whirlpool of Disease.

Clinical reasoning that remains purely conceptual may unintentionally strengthen such whirlpools through excessive labeling, testing, or intervention. Pattern-sensitive reasoning looks for leverage points that restore flexibility.

When depth is denied

When subconceptual processing is ignored or denied, reasoning becomes brittle. Intuition is mistrusted rather than monitored. Uncertainty is handled defensively. Errors may follow.

Beyond error, something deeper emerges. Loss of meaning. Emotional numbing. Existential exhaustion. Burnout is often not caused by too much feeling, but by depth that has no acknowledged place. This is explored in Physicians’ Despair.

Clinical reasoning is not just a cognitive act. It is also a human one.

Artificial intelligence and decision support

Artificial intelligence is now entering the realm of clinical reasoning. This creates a fork in the road.

A.I. can amplify the conceptual layer: more rules, alerts, scores, and justifications. This may increase efficiency, but it risks deepening alienation and despair, as warned in Dangers of A.I. to Future Healthcare.

Alternatively, A.I. can be designed to respect and support the subconceptual layer. Not by formalizing intuition, but by protecting space for it. By offering orientation rather than coercion. By supporting pacing, reflection, and uncertainty. This possibility is explored in Medical A.I. for Humans and How can Medical A.I. Enhance the Human Touch?.

Decision support is never neutral. It either colonizes clinical reasoning or scaffolds it.

What clinical reasoning asks from support

Good support does not replace thinking. It makes thinking visible. It supports hypothesis framing rather than forcing closure. It respects uncertainty and timing. It leaves responsibility where it belongs: with the clinician.

Technology should serve human reasoning, not redefine it.

Closing

Clinical medical reasoning is not broken. It is misunderstood.

By recognizing both its conceptual surface and its subconceptual depth, medicine can protect clinical wisdom, human meaning, and professional integrity. The task is not to make clinicians think differently, but to allow them to think as they already do, openly and whole.

Addendum — Subconceptual Processing in Clinical Medical Reasoning

Feature (SPT term)	What it means (subconceptual level)	Manifestation in clinical reasoning	Concrete clinical example	Risk if ignored / denied	Risk if overrelied upon
Graceful degradation	Patterns continue to function when incomplete or weakened	Reasoning proceeds despite missing data	Acting on a partial history in emergency care	Paralysis, overtesting, refusal to decide	Acting on too little information
Concurrent multiple soft constraint satisfaction	Many weak influences shape one outcome simultaneously	Differential diagnosis weighted by many subtle cues	Pain description + age + posture + “something feels off”	Overreliance on single factors or rules	Diffuse reasoning, lack of focus
Content-addressable memory	Partial input evokes whole patterns	Rapid pattern recognition	“This cough reminds me of…”	Loss of efficiency; novice-like slowness	Availability bias
Pattern recognition & completion (PRC)	Expectation-driven completion of patterns	Seeing what fits an emerging hypothesis	Recognizing classic migraine early	Confirmation bias if unchecked	Premature closure
Predictive brain	Anticipation shapes perception	Hypotheses guide questioning and observation	Asking focused questions based on early hunch	Blindness to unexpected findings	Confirmation bias
Degrees of robustness	Some patterns are more stable than others	Core diagnoses resist reconsideration	Long-standing diagnosis shaping interpretation	Diagnostic inertia	Diagnostic inertia
Learning from experience	Patterns form during action, not instruction	Expertise grows mainly in practice	“You only learn this by seeing many cases”	Overvaluing theoretical knowledge	Fossilized habits
Conditioning-like phenomena	Associations form without awareness	Emotional reactions to certain cases	Anxiety after a missed diagnosis	Defensive medicine, avoidance	Defensive or avoidant practice
Spontaneous generalization	Overlapping patterns merge naturally	Case-to-case learning	Recognizing a new variant of a known syndrome	Overgeneralization	Overgeneralization
Spontaneous internal ontology	Categories form implicitly	Natural grouping of problems and syndromes	“This is a vascular-type pain”	Rigid or idiosyncratic categorization	Idiosyncratic categorization
From exemplars to categories	Specific cases anchor general knowledge	Memorable cases guide reasoning	“I once had a patient like this…”	Availability bias	Anecdotal dominance
Mental depth via distributed MNPs	Meaning emerges from wide overlap	Clinical wisdom beyond rules	Sense of proportion in complex cases	Shallow, protocol-driven care	Difficulty explicating decisions
Positive & negative whirlpools	Self-reinforcing mental–somatic loops	Chronicity and psychosomatics	Pain–stress–pain cycles	Escalation without reflection	Self-reinforcing misinterpretation
Phenomena of attribution	Overlap between domains triggers associations	Psychological and somatic interplay	Stress influencing symptom interpretation	Misattribution or dismissal	Misattribution of causality
Mental overlap → empathy	Shared patterns across individuals	Clinical empathy	“I sense what this means for them”	Detached or mechanical care	Loss of clinical distance
Spontaneous pleiotropy	One pattern affects multiple domains	Comorbidity clustering	Depression, pain, fatigue co-occurring	Fragmented treatment	Blurred problem boundaries
Degeneracy	Same outcome via different paths	Multiple valid reasoning routes	Different clinicians, same diagnosis	False belief in “one right way”	Inconsistent reasoning paths
Graceful fading between syndromes and normalcy	No sharp boundaries	Continuum thinking	Subclinical hypothyroidism	Overdiagnosis or stigmatization	Under-recognition of pathology
Competence without comprehension	Skill precedes explanation	Acting correctly before verbalizing	Immediate resuscitation actions	Devaluing intuition or experience	Inability to justify or teach

Me: “Lisa, in your view, what does this say about the place of psychosomatics in medicine?”

Lisa: “It says that psychosomatics is not a side specialty, nor a consolation prize when ‘nothing is found.’ It belongs at the very heart of medicine.

If clinical reasoning truly unfolds at both the conceptual and subconceptual levels, then psychosomatics is simply what becomes visible when medicine acknowledges this depth. Symptoms are not only signals of malfunctioning parts, but expressions of meaningful patterns in a whole human being. Ignoring that does not make medicine more scientific; it makes it incomplete.

In this light, psychosomatics is not opposed to biomedical thinking. It precedes it. It reminds medicine that patients present with patterns, not diagnoses, and that bodily expression often carries what cannot yet be said conceptually. When this is heard early, much suffering need not become chronic or mysterious.

So, psychosomatics is not “extra.” It is what medicine looks like when it listens with a depth stethoscope — and dares to remain human while being rigorous.”