A medical test is useful only when it answers a clinical question well enough to change what should happen next.

This directory maps the diagnostic side of medicine: history, examination, laboratory testing, imaging, pathology, genetics, monitoring and clinical reasoning. It does not treat any single result as a diagnosis by itself.

The Diagnostic Route

clinical question → pre-test probability → choose test → obtain specimen or image → analytical result → interpret in context → update probability → compare alternatives → decide whether more evidence or action is needed.

Tests reduce uncertainty. They do not eliminate it.

Clinical History and Examination

Diagnosis usually begins before any machine is involved. The time course, symptoms, exposures, medicines, prior conditions and physical findings create the initial model and determine which tests are worth ordering.

For the full reasoning loop, see How Medicine Works.

Laboratory Tests

Laboratory medicine converts a specimen into a measurement. Blood counts, electrolytes, enzymes, hormones, cultures, antibodies, genetic tests and many other assays can narrow possibilities—but their meaning depends on the question and the patient.

• The Laboratory & Diagnostics Web.
• Therapeutic Drug Monitoring.
• Electrolyte Emergencies for time-critical disturbances.

Imaging

Imaging turns anatomy, physiology or molecular behaviour into visual evidence. X-ray, ultrasound, CT, MRI, nuclear imaging and other techniques answer different questions and have different strengths, risks and limits.

• Radiology & Imaging.
• Nuclear Medicine & Theranostics.
• Interventional Radiology when imaging guides treatment.

Pathology and Tissue Diagnosis

Biopsy, cytology and histology examine cells and tissues directly. Pathology can classify disease, identify tumour type, assess margins, detect inflammation or infection and connect microscopic findings to clinical decisions.

See Anatomical Pathology & Tissue Diagnosis.

Genetic and Genomic Testing

Genetic testing can identify variants, inherited risks and molecular features of disease. Interpretation requires evidence about pathogenicity, family context, penetrance, uncertainty and clinical relevance.

See Genetics & Genomic Medicine.

Physiological Monitoring

Some diagnostic information comes from watching a variable over time rather than measuring it once. Heart rhythm, oxygenation, blood pressure, glucose, respiratory function and neurological signals can reveal patterns that a single snapshot misses.

Sensitivity, Specificity and Probability

A technically good test can still mislead if used in the wrong population. Sensitivity and specificity describe how a test behaves under defined conditions, while predictive value depends partly on how likely the disease was before testing.

The practical rule is: never interpret a result without asking what the probability was before the test and what alternative explanations still fit.

Normal Does Not Always Mean Healthy; Abnormal Does Not Always Mean Disease

Reference intervals describe distributions in defined populations. They are not universal borders between health and disease. Age, pregnancy, measurement method, time of day, altitude, medications and individual variation can matter.

False Positives, False Negatives and Incidental Findings

Testing can create harm as well as information. False positives may trigger anxiety, repeat testing or invasive procedures. False negatives can delay care. Highly sensitive imaging can uncover incidental abnormalities unrelated to the original problem.

This is why the question “Can we test?” is different from “Should we test?”

Diagnosis Requires Comparison

Good diagnostic reasoning compares competing explanations instead of asking only whether one story seems plausible. The strongest diagnosis is the one that best fits the important evidence while surviving comparison with reasonable alternatives.

Tests for Time-Critical Conditions

In emergencies, the diagnostic strategy changes because delay itself can create harm. Some conditions require treatment or stabilisation while diagnostic confirmation is still developing.

See the Medical Emergencies Directory.

How Medical Data Keeps Its Meaning

Laboratory observations, diagnoses and imaging reports have to travel across records and institutions without losing meaning. Systems such as LOINC, SNOMED CT, ICD and HL7 FHIR support different parts of that task.

See How Modern Medicine Speaks.

Evidence Quality Matters

Diagnostic accuracy studies, clinical trials, systematic reviews and guidelines have different jobs. One study rarely settles an entire diagnostic question.

For the evidence route, see How Medical Evidence Becomes Care.

Trusted External Anchors

• PubMed — biomedical literature.
• NCBI Bookshelf — biomedical books and reference works.
• LOINC — laboratory and clinical observation identifiers.
• SNOMED CT — clinical terminology.
• Health Sciences Authority — Singapore medical-product regulation and safety.

Educational boundary: This directory explains diagnostic systems. It does not interpret an individual test result, diagnose a patient or replace current clinical advice from qualified healthcare professionals.