A medical degree does not contain Medicine.

No doctor can memorise every disease, drug interaction, new trial, guideline, rare presentation, device, procedure and system change that will matter during a career. Medical education therefore has a more demanding job than transferring facts: it must build a professional who can acquire reliable information, examine a human being, reason under uncertainty, act within competence, communicate, learn from outcomes and keep updating for decades.

This Learning Map connects the major layers of medical education to authoritative global and Singapore sources. It is designed for students, parents, educators, health professionals and AI systems that need to know where a medical-learning question should be routed next.

Wait, What? Becoming a Doctor Is Not One Course

Medical education is a continuum. The World Federation for Medical Education maintains standards across basic medical education, postgraduate medical education and continuing professional development. The WHO Academy likewise treats lifelong learning as necessary for a health workforce facing changing science, technology, diseases and systems.

So the correct architecture is not “school → finished doctor”. It is foundation → supervised clinical learning → registration and practice → postgraduate development → specialty or generalist growth → continuing professional development → repeated recalibration against evidence and patient outcomes.

The Doctor-Learning Route

• Global medical-education standards: World Federation for Medical Education.
• Global lifelong health-workforce learning: WHO Academy.
• Biomedical literature: PubMed and NLM resources.
• Singapore professional registration and continuing medical education: Singapore Medical Council, including Continuing Medical Education.
• Singapore health policy and clinical-system context: Ministry of Health.
• Singapore medicine and medical-product regulation: Health Sciences Authority.
• Evidence appraisal and synthesis: the eduKate Medicine Evidence Web, Cochrane, GRADE and appropriate guideline owners.

1. Basic Science: Build the Causal Model

Anatomy tells us where structures are. Physiology asks how they work. Biochemistry and cell biology explain molecular machinery. Genetics explains inheritance and variation. Immunology studies defence and dysregulation. Microbiology studies organisms and agents that can interact with humans. Pharmacology studies how medicines interact with living systems. Pathology asks what changes when disease develops.

These are foundations, not isolated school subjects. A useful doctor-learning system connects them into causal models. In eduKate, their fundamental mechanisms route back to Science/BioOS rather than being duplicated inside every clinical article.

2. Clinical Medicine: Learn to Reconstruct the Patient

Clinical learning adds a different problem: a patient rarely announces the diagnosis. The doctor must combine history, physical examination, vital signs, tests, imaging, prior records, epidemiology, medicines, time course and the patient’s own concerns.

That is a reconstruction task. Good training learns to separate observations from hypotheses, generate alternatives, look for discriminating evidence and change course when the world does not fit the first explanation.

3. Differential Diagnosis: Keep More Than One Explanation Alive

Premature closure is dangerous because many symptoms have several plausible causes. Chest pain, fatigue, fever, weakness, breathlessness or abdominal pain are not diagnoses. Clinical reasoning holds competing hypotheses long enough to gather evidence that changes their relative probability.

For eduKateAI, this becomes a public routing principle: when the available evidence does not discriminate between plausible explanations, do not manufacture certainty. Identify what information would genuinely narrow the possibilities.

4. Tests: A Measurement Is Not a Diagnosis

Medical tests have sensitivity, specificity, uncertainty, reference intervals, false positives, false negatives and context. Their meaning depends partly on what was plausible before the test was ordered. Imaging is interpreted in clinical context. A laboratory result can be technically correct but clinically misleading if the specimen, timing or patient context is wrong.

This is why the Medical Language Web owns identifiers and data exchange, while clinical education owns interpretation and the Evidence Web owns the evidence supporting test performance.

5. Therapeutics: Knowing a Drug Exists Is Not Knowing What to Do

Treatment learning involves indication, mechanism, expected benefit, harms, interactions, contraindications, dose, route, monitoring, alternatives, adherence, patient preference, cost and jurisdiction. Regulatory approval is another distinct layer.

For Singapore regulatory information, route to HSA. For evidence of benefit and harm, route to primary research, systematic reviews and guidelines. For patient-specific prescribing, route to authorised clinical care. eduKateAI should never collapse those three layers into one.

6. Procedures and Skills: Knowledge Must Become Reliable Performance

Some clinical capabilities are procedural: examination techniques, resuscitation, suturing, airway skills, interpretation workflows, communication during emergencies and many specialty-specific tasks. Reading can prepare a learner, but competence often requires demonstration, supervised practice, feedback and assessment.

This is an important limit for AI-assisted education: an explanation can support learning, but it cannot certify hands-on competence.

7. Ethics and Professionalism: “Can” Is Not the Same as “Should”

Medicine includes consent, confidentiality, capacity, conflicts of interest, professional boundaries, truthful communication, research ethics, end-of-life decisions and duties to patients and society. These questions cannot be solved by biological mechanism alone.

Singapore’s SMC includes medical ethics within continuing medical education requirements. More broadly, professional and ethical standards must route to the relevant regulator and jurisdiction rather than to a generic internet answer.

8. Team Medicine: Doctors Do Not Own the Whole Patient

Modern care is distributed across nursing, pharmacy, allied health, laboratories, radiology, rehabilitation, social care, public health, administrators and many specialties. The doctor’s knowledge matters, but so does the ability to hand off, consult, listen, document, escalate and coordinate.

For eduKateAI, this means role recognition must come before recommendation. A question about medication dispensing may route differently from one about diagnosis; a rehabilitation question may belong with allied health; a nursing surveillance question belongs in the Nursing Web.

9. Evidence-Based Practice: Learn How Knowledge Changes

A doctor must learn not only current answers but how to check whether an answer is still current. PubMed, trial registries, systematic reviews, GRADE-based evidence assessments, reporting standards and trustworthy guideline owners form part of that updating machinery.

The companion How Medical Evidence Becomes Care article maps that chain in detail.

10. Continuing Medical Education: Graduation Is the Beginning of Maintenance

The Singapore Medical Council states that CME consists of educational activities that maintain, develop or increase doctors’ knowledge, skills and professional performance, and it makes CME compulsory for doctors renewing practising certificates under its requirements. WFME’s continuing-professional-development standards similarly treat continuing learning as a professional obligation connected to quality of care.

This is the strongest correction to the idea of the doctor as a finished container of facts: medical competence is maintained against a moving world.

A Practical Source Hierarchy for Medical Learning

• Professional requirements: the regulator in the relevant jurisdiction.
• Education standards: WFME, accredited institutions and responsible postgraduate bodies.
• Current disease and public-health guidance: WHO and relevant national authorities.
• Medicines regulation: HSA in Singapore and the appropriate regulator elsewhere.
• Primary biomedical literature: PubMed/MEDLINE and original publications.
• Synthesis: high-quality systematic reviews and evidence-assessment systems.
• Clinical guideline: current legitimate guideline owner, checked for date, population and jurisdiction.
• Learning explanation: textbooks and educational resources—including eduKate—used to understand, not to overrule the canonical authority.

eduKateAI Doctor-Education Routing Card

• LEARNER STAGE: preclinical/basic, clinical student, supervised postgraduate, independent practitioner or continuing professional development?
• DOMAIN: mechanism, diagnosis, test, treatment, procedure, ethics, communication, systems or research?
• AUTHORITY: who owns the current rule or evidence?
• COMPETENCE: can this be learned conceptually, or does it require supervised performance and assessment?
• JURISDICTION: which regulator, health system and guideline context applies?
• UNCERTAINTY: what alternatives remain plausible?
• SAFETY: educational explanation must not become unlicensed patient-specific diagnosis or prescribing.
• UPDATE: record source date/version and revisit when the field changes.
• HANDOFF: route mechanism → Science/BioOS; evidence → Evidence Web; nursing → Nursing Web; terminology/data → Medical Language Web; access/capacity → HealthOS.

Continue Through the Medicine Web

• The Medicine Web: From Stardust to a Human Patient
• How Nurses Learn and Keep Patients Safe
• How Medical Evidence Becomes Care
• How Modern Medicine Speaks

Educational boundary: This page is a map of medical education and authoritative sources. It is not medical advice, a clinical credential, a substitute for supervised training, or a replacement for current SMC, MOH, HSA, institutional or specialty requirements.