Heuristics and Axioms in Clinical Reasoning

This essay is by Jim Meadows, BScPT, MCPA, FCAMPT. It is the third in a series on clinical reasoning in physical therapy.

---

Clinical reasoning rarely happens in a vacuum. Before any formal process begins, most clinicians are already working with quick mental shortcuts — rules of thumb drawn from experience, pattern instincts, and guiding principles absorbed from training and practice. These shortcuts have names: heuristics and axioms. Understanding what they are, how they help, and where they can mislead is a foundational part of reasoning well in the clinic.

Definitions

A heuristic (from the Greek word meaning “to discover”) refers to an experience-based technique for problem solving that finds a solution which is not guaranteed to be optimal, but good enough for a given set of goals. Heuristics are practical shortcuts — rule of thumb, common sense, intuition — that reduce the complexity of a problem without requiring exhaustive analysis.

An axiom is a self-evident truth that requires no proof — a premise so widely accepted that it functions as a starting point for reasoning rather than a conclusion to be argued toward. In practice, many statements we treat as axioms in physical therapy are actually opinions that have simply achieved widespread enough acceptance to feel self-evident.

Heuristics in Clinical Practice

Heuristics reduce diagnostic complexity to manageable proportions. Using lateral elbow pain as an example: common sense suggests the pain is local rather than referred, and based on prevalence, tennis elbow is the most likely diagnosis. This heuristic gets you to a working hypothesis quickly, without an exhaustive examination.

The problem is clear immediately. Completely trusting a heuristic without verification invites error. This is generally how humans approach everyday problem solving — and unfortunately, how many clinicians approach diagnosis. Heuristics are useful starting points, but effort must be made to confirm their conclusions and guard against the cognitive biases they can introduce.

Key Axioms in Clinical Reasoning

Think Horses, Not Zebras

Two old clinical maxims capture the same idea: “If you hear hoof-beats, think horses not zebras” and “If it looks like a duck, waddles like a duck, and quacks like a duck, it is a duck.” Both are reminders to keep the common and obvious hypothesis at the front of clinical thinking. The formal version is the “common disease, common variant” principle.

Worth noting: zebras do beat their hooves. If you were in Africa, the saying would be reversed. Uncommon conditions exist and will occasionally present in your clinic. The point is not to dismiss them — it is to assign them appropriate probability rather than pursuing them first.

Occam’s Razor

Occam’s Razor is arguably the most useful axiom in clinical problem solving. The loose clinical translation: all things being equal, the simplest explanation is probably the correct one. More precisely, when choosing between competing hypotheses, make as few assumptions as possible.

Using lateral elbow pain again: it is possible that both a tendinopathy and cervical referral are contributing. It is possible that radiohumeral joint dysfunction is also involved. But the probability that two or three distinct pathologies are simultaneously producing the same presentation is low. Do not pursue multiple explanations until the evidence forces you to. When a patient reports that gripping is painful but so is sitting quietly watching television, more information has come in — and now it is worth reconsidering.

Clinicians tend to resist Occam’s Razor. We are drawn to rare and complex diagnoses. We find the common and straightforward unsatisfying. That instinct is worth recognizing and countering deliberately.

Hickum’s Dictum

Hickum’s Dictum pushes back on Occam’s Razor: “A patient can have as many diseases as he damn well pleases.” It is a reminder that forcing every presentation into a single unifying diagnosis is not always correct — and that the requirement for diagnostic elegance can cause its own errors.

In general, a single condition explaining all symptoms is preferable — it simplifies treatment planning and reduces the risk of over-treating. But if three common conditions are each independently likely to produce three different symptoms, three diagnoses may be more accurate than one rare condition forced to account for all of them. Hickum’s Dictum exists to prevent Occam’s Razor from becoming a straitjacket.

The Uncertainty Principle: Treat the Patient, Not the Numbers

Statistics describe averages. They tell you what happens across populations — not what is happening to the individual in front of you. A condition that occurs in 1 in 1,000,000 adults does not mean it will never present in your clinic. It means it probably will not present today. But somebody will see that patient, and it might be you.

Clinicians and researchers who deal with homogeneous systems — particles, molecules, falling objects — can rely on statistical prediction at the individual level because the variables are consistent. Human patients are not homogeneous. The same condition presents differently from person to person, sometimes dramatically so. Use population statistics to inform probability, not to dismiss individual variation.

Sutton’s Law

Willie Sutton was a bank robber who, when asked why he kept robbing banks, reportedly said: “Because that’s where the money is.” Sutton’s Law applied to clinical reasoning: do only the tests that are appropriate for your current hypothesis.

Clinicians tend to test everything available, justified by the goal of not missing anything. In practice, indiscriminate testing is often counterproductive. An MRI on every patient with low back pain will reveal disc bulging, degeneration, and other incidental findings in a significant proportion — most of which are entirely unrelated to the presenting complaint. The investigation shapes the diagnosis, and not always for the better.

If you know what the problem probably is, use only the tests that move you to the next decision point. This applies equally to questions asked during the subjective examination. Testing costs time, generates noise, and introduces opportunities for cognitive bias. Sutton’s Law keeps the process focused.

Putting It Together

These axioms appear to contradict one another — and they do. Occam’s Razor and Hickum’s Dictum point in opposite directions. The uncertainty principle and the horses-not-zebras heuristic create tension by design. They are not rules to follow mechanically; they are frameworks for thinking that must be applied with judgment.

Taken together, they reduce to two working principles:

First: base everything on probability, not possibility. Choose the solution that best fits the available facts. Start with the most probable hypothesis. Uncommon conditions and presentations do occur — but they are probably not occurring right now. Occam’s Razor, the duck and zebra heuristics, Hickum’s Dictum, and the uncertainty principle all inform this judgment without replacing it.

Second: ask only the questions and do only the tests that are immediately relevant to the immediate problem. If the goal is diagnosis, focus there. If the goal is establishing etiology, focus there. Do not import questions about etiology into a diagnostic workup, or vice versa. Sutton’s Law keeps the reasoning clean.

---

Also in This Series

• Script Focused Deduction: Mimicking the Expert
• Clinical Reasoning: Methods and Tools
• Methods of Clinical Reasoning
• The Pathoanatomical Diagnosis
• Locking and Specificity in Spinal Manipulation