eduKateSG · Why Science?
Turn evidence into accountable health decisions—connecting product claims, quality, clinical studies, benefit-risk reasoning, policy, monitoring and public trust
A fact-checked guide to the Duke-NUS Centre of Regulatory Excellence and the science used to strengthen health-product and policy decisions.
Duke-NUS describes the Centre of Regulatory Excellence as an Asian platform for regulatory capability, policy innovation and leadership serving health regulators, biomedical industries and health-product organisations.
Its current strategic approach combines education and capacity building, think-tank activity and advisory support; its stated focus areas include patient engagement, digital health, vaccines, innovative therapies and innovative trials.
Did you know? Regulation is a lifecycle: evidence before market entry and monitoring after use answer different but connected questions.
This guide does not interpret the legal or regulatory status of any real product. Relevant authorities control current decisions. Sources were checked 11 November 2026.
Chapter index
Chapters 1–9: regulatory science, intended use, product claims, quality systems, preclinical evidence, clinical studies, statistics, benefit-risk, patient engagement
Chapters 10–17: digital health, vaccines, innovative therapies, innovative trials, manufacturing, monitoring, policy, communication and decision receipts
Regulatory science develops and applies evidence methods for health decisions. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 1 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Name the product, jurisdiction, question and decision owner. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
Regulation is broader than a checklist. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Define regulatory science” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Write a decision statement. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 1 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
Intended use defines who, what, where and why a product is used. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 2 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Connect claims to users, settings and instructions. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
The same technology can carry different risks in different uses. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Specify intended use” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Compare two intended uses. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 2 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
Evidence develops from discovery through development, use and monitoring. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 3 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Mark decision points and evidence updates. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
Approval is not the end of learning. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Map the evidence lifecycle” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Draw a lifecycle map. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 3 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
Standards support consistency while authorities make context-specific decisions. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 4 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Record applicable editions and deviations. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
Compliance with one standard does not prove overall benefit. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Separate standards and decisions” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Build a standards matrix. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 4 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
| Evidence move | What to record | What it cannot prove alone |
|---|---|---|
| Regulatory question question | Population, variable, unit and time window | A definition does not establish a result |
| Comparison | Baseline, comparator and uncertainty | Association alone does not establish cause |
| Decision boundary | Owner, trigger, review and escalation | A model does not replace accountable judgement |
Quality systems aim for consistent identity, strength, purity and performance. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 5 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Trace materials, processes, controls and release tests. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
A passing batch test cannot prove clinical benefit. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Evaluate product quality” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Map a quality system. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 5 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
Laboratory and animal studies explore mechanism, performance and hazards. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 6 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Match models and exposures to the intended claim. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
Nonclinical success does not establish human benefit. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Read nonclinical evidence” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Build an evidence ladder. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 6 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
Clinical studies test safety, performance or effectiveness in people. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 7 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Review design, endpoints, comparators and representativeness. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
Statistical significance may not be clinically meaningful. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Assess clinical studies” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Critique a fictional trial. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 7 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
Decisions weigh expected benefits, harms and uncertainty for an intended population. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 8 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Use structured evidence and stakeholder perspectives. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
A numerical score cannot remove value judgements. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Reason about benefit and risk” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Create a benefit-risk table. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 8 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
| Evidence move | What to record | What it cannot prove alone |
|---|---|---|
| Evidence question | Population, variable, unit and time window | A definition does not establish a result |
| Comparison | Baseline, comparator and uncertainty | Association alone does not establish cause |
| Decision boundary | Owner, trigger, review and escalation | A model does not replace accountable judgement |
Experience and preferences can improve relevant questions and communication. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 9 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Define when and how input affects decisions. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
Consultation is not consent or automatic agreement. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Engage patients and publics” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Design an engagement plan. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 9 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
Software can change through data, updates and workflow integration. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 10 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Track versions, validation, cybersecurity and human oversight. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
A high model score does not prove safe deployment. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Regulate digital health” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Audit a fictional algorithm. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 10 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
Vaccine evidence spans quality, immunogenicity, safety, efficacy and effectiveness. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 11 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Preserve population, schedule and follow-up context. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
An immune marker alone may not establish protection. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Evaluate vaccines” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Map vaccine evidence. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 11 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
Cell, gene and other advanced therapies create novel manufacturing and follow-up needs. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 12 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Connect product variability, mechanism and long-term monitoring. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
Novelty does not lower the evidence threshold. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Consider innovative therapies” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Build a therapy risk map. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 12 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
| Evidence move | What to record | What it cannot prove alone |
|---|---|---|
| Lifecycle question | Population, variable, unit and time window | A definition does not establish a result |
| Comparison | Baseline, comparator and uncertainty | Association alone does not establish cause |
| Decision boundary | Owner, trigger, review and escalation | A model does not replace accountable judgement |
Adaptive and other innovative designs can answer questions efficiently. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 13 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Predefine rules, simulations and operating characteristics. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
Flexibility without control can inflate error. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Design innovative trials” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Sketch an adaptive decision tree. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 13 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
Post-market evidence can reveal rare harms and performance in broader populations. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 14 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Define data quality, signals and confirmation routes. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
A spontaneous report alone does not prove causation. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Monitor real-world use” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Create a signal workflow. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 14 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
Effective systems need law, institutions, expertise, data and cooperation. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 15 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Assess capability, governance and feedback. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
Copying one country’s model may not fit another. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Strengthen regulatory systems” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Build a systems map. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 15 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
Public explanations should connect evidence, uncertainty and conditions of use. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 16 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
State what changed and what remains unknown. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
Confidence must not erase uncertainty. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Communicate decisions transparently” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Rewrite a decision summary. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 16 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
The receipt links claim, evidence, standards, uncertainty, decision and monitoring. Regulatory science connects scientific evidence with public decisions about health products and systems. Duke-NUS says its Centre of Regulatory Excellence builds capability, convenes policy dialogue and provides advisory support across Asia. Chapter 17 keeps product claims, evidence standards, benefit-risk reasoning, manufacturing quality and accountable decisions connected, so regulation appears as applied science rather than paperwork.
Archive versions, advice, rationale and review triggers. A defensible regulatory record states the product, intended use, target population, evidence question, study design, comparator, quality controls, uncertainty and decision owner. Teams trace every claim to data and every data set to its method and version. Because evidence evolves after approval or deployment, monitoring, reporting and reassessment belong in the same lifecycle.
Documentation does not make a decision infallible. The boundary protects trust. Regulatory education does not authorise a learner to approve a product, and a submission, designation or clearance in one jurisdiction does not prove universal safety or effectiveness. Scientific advice, policy development and formal regulatory decisions have different owners. Current authorities, laws, standards and qualified professionals govern real products and clinical care.
Explore “Keep a regulatory decision receipt” with a fictional health technology dossier. Define the claim, evidence, uncertainty, affected groups and possible decision options, then write what additional evidence would change the decision. Assemble a one-page receipt. Do not evaluate a real person’s treatment or market an unapproved product. The classroom goal is transparent benefit-risk reasoning.
Did you know? Regulators can reach different decisions from the same evidence when intended use, population, comparator, uncertainty or acceptable risk differs. Chapter 17 shows why a decision receipt should preserve both the scientific record and the policy context instead of reducing regulation to a simple pass or fail label.
Capabilities include biology, medicine, pharmacy, engineering, statistics, law, ethics, economics, quality systems, communication and public policy. Current programmes and regulatory roles set their own qualifications. Students can build foundations by checking sources, separating observations from claims, comparing alternatives and explaining uncertainty without hiding the decision that must be made.
| Evidence move | What to record | What it cannot prove alone |
|---|---|---|
| Decision question | Population, variable, unit and time window | A definition does not establish a result |
| Comparison | Baseline, comparator and uncertainty | Association alone does not establish cause |
| Decision boundary | Owner, trigger, review and escalation | A model does not replace accountable judgement |
Contents · Previous chapter · Continue in the Science Learning Hub
Primary sources and verification date
Source status for Why Science? | Centre of Regulatory Excellence was checked 11 November 2026. Research priorities, programmes, recruitment, facilities and participation routes can change; use the linked official owner for the current edition.
- Duke-NUS Centre of Regulatory Excellence — Official current centre overview and programme areas.
- About CoRE — Official institutional purpose and regional context.
- CoRE strategic goals — Official approaches and current focus areas.
- Duke-NUS research centres — Official current centre listing.
Continue the evidence journey
Regulatory science earns trust when claims stay connected to evidence, decisions preserve uncertainty and monitoring continues throughout the product lifecycle. Families considering science tuition or a science tutor can use the same evidence questions: what will the learner practise, what feedback will they receive, and how will progress be shown?