Diagnosis is the process of building the best current explanation for a human health problem from symptoms, context, examination, measurements and tests—while keeping enough uncertainty visible to change the explanation when new evidence arrives.
A diagnosis is therefore more than naming a disease. It is a reasoning process and a communication tool. It helps predict what may happen, determine what evidence is still needed, choose appropriate treatment and organise follow-up.
Diagnosis Starts Before the Test
The patient’s story establishes the initial problem: what changed, when, how severely, from what baseline and in what context. Age, medicines, previous disease, exposures, family history and functional change can all shift which explanations are plausible.
Symptoms Are Many-to-Many Signals
One symptom can arise from many diseases, and one disease can produce many different symptoms. That is why safe diagnosis cannot operate as symptom → one disease → one treatment.
The disease layer is explained in What Is Disease?.
The Differential Diagnosis
A differential diagnosis is a structured set of competing explanations. The goal is not to generate the longest possible list. It is to preserve the important alternatives, especially those that are plausible, dangerous to miss or meaningfully changed by available evidence.
Tests Update Probability
The meaning of a test depends partly on what was believed before it was ordered. A positive result in a high-risk patient can carry a different implication from the same result in a very low-risk population.
probability before test → result → probability after test.
The full testing architecture is in the Medical Tests and Diagnosis Directory.
Sensitivity and Specificity Do Not Tell the Whole Story
Sensitivity describes how often a test detects a condition when it is present under defined conditions. Specificity describes how often it is negative when the condition is absent. Neither number alone tells us the probability that one individual has the disease after a particular result.
Diagnosis Has Levels of Certainty
- Observation: what was seen, reported or measured.
- Possible diagnosis: an explanation still under consideration.
- Working diagnosis: the explanation currently guiding care.
- Confirmed diagnosis: a diagnosis supported strongly enough by the relevant evidence and criteria.
- Excluded or unlikely alternative: an explanation whose probability has fallen sufficiently.
These distinctions matter because a provisional label can otherwise become permanent in a medical record even after contradictory evidence appears.
Urgency Changes Diagnostic Strategy
When delay can cause irreversible harm, clinicians may need to stabilise physiology or begin time-sensitive treatment before the diagnostic picture is complete. The decision depends on probability, consequence, time sensitivity and reversibility.
See the Medical Emergencies Directory.
A Diagnosis Can Be Wrong
Diagnostic error can arise from incomplete information, atypical presentation, measurement error, premature closure, anchoring, communication failures or disease evolution. Medicine reduces this risk by preserving alternatives, following pending results and reopening the model when the patient’s course contradicts expectations.
Failure to Improve Is Diagnostic Evidence
If treatment aimed at the working diagnosis does not produce the expected response, several possibilities reopen: the diagnosis may be incomplete, the treatment may not work for this patient, delivery may have failed, a complication may have appeared or another disease may now dominate.
Diagnosis and Classification Are Not Identical
A clinician’s diagnostic model is richer than a billing, epidemiological or data code. Systems such as ICD and SNOMED CT help preserve and exchange meaning, but the coded label should not replace the clinical reasoning behind it.
See How Modern Medicine Speaks.
Evidence Supports Diagnosis
Diagnostic studies estimate how measurements perform, but their usefulness depends on study quality, population, reference standard and applicability. The broader evidence architecture is explained in What Is Medical Evidence?.
From Diagnosis to Treatment
A diagnosis does not automatically determine one treatment. Severity, baseline risk, comorbidities, contraindications, evidence certainty, alternatives, patient goals and consent still matter.
Continue to What Is Medical Treatment?.
Canonical Diagnostic Routes
- How Medicine Works — complete clinical loop.
- Laboratory & Diagnostics.
- Radiology & Imaging.
- Anatomical Pathology.
- Genetics & Genomic Medicine.
Educational boundary: This article explains diagnostic reasoning. It cannot diagnose an individual or interpret personal symptoms or test results in place of qualified healthcare professionals.