Clinical Trials and Medicines Evidence

Clinical evidence literacy is the ability to understand how a proposed medicine or treatment is investigated, what a trial can establish, and why safety, participant rights and professional review come before persuasive claims. A laboratory finding may suggest a promising mechanism, but it does not by itself prove that a treatment is safe or effective in people. An early clinical study may answer an important question without establishing everyday benefit for a broad population. A positive result does not remove the need to study harms, uncertainties and who was actually included.

Ministry of Education V3.0 should teach the journey from research question to clinical evidence as part of Science, Mathematics, evidence evaluation and health literacy. It should help learners distinguish experiments from observational studies, understand consent and regulatory oversight, interpret proportions without overclaiming, read a trial registry, and recognise the responsibilities of clinicians, researchers, ethics committees and public authorities. This reference contains a twelve-stage evidence-route protocol, a phase-by-phase explanation, fictional numerical cases, safeguards for children, a classroom task and a fair assessment rubric.

This independent eduKateSG educational chapter is not an official Singapore MOE, Ministry of Health or Health Sciences Authority clinical guide. It is not advice to take, stop, test, purchase, prescribe or enroll in any treatment or clinical trial. Real medical decisions and trial participation involve individual circumstances, qualified professionals, ethics oversight and current regulatory rules. Students must not run health experiments on themselves or classmates.

1. A Medical Idea Is Not the Same as an Established Treatment

Scientists may propose that a substance affects a biological pathway. Laboratory observations, computer models or early studies may support that idea. But evidence about the substance under controlled conditions does not automatically predict its effect in human bodies, which differ in physiology, age, conditions, other medicines and exposure. The evidence route therefore involves several kinds of research rather than one decisive demonstration.

The Singapore Health Sciences Authority’s regulatory overview, updated May 2026 defines clinical trials as research intended to investigate relevant effects, adverse reactions, behaviour in the body and safety or efficacy of health products. It identifies regulated product categories and routes under Singapore’s Health Products Act and Medicines Act. The exact regulatory requirements depend on the product and investigation.

2. The Eight Rooms Between a Claim and Clinical Use

Room	Core question	What can go wrong
Research question	Which outcome, population and treatment is being examined?	An attractive slogan substitutes for a testable objective
Foundational evidence	What prior laboratory or other evidence motivates the work?	Early findings are described as proof of human benefit
Protocol	How will participants, comparisons and outcomes be defined?	Results are interpreted without knowing the intended test
Ethics and protection	Were appropriate permissions and safeguards established?	Potential benefit is used to dismiss participant risk
Clinical observation	What happened under the stated research conditions?	Anecdote is mistaken for an appropriately controlled result
Analysis	How uncertain are estimated benefits and harms?	Small or selective numbers are overgeneralised
Regulatory and professional review	Which qualified authorities assess evidence and lawful use?	A public claim is treated as an approval
Post-use monitoring and correction	What new safety and performance information may emerge?	Uncertainty is assumed to disappear after approval

The route is conceptual. Different products and investigations can follow different pathways; some studies are observational and some combine phases. A student should not infer regulatory status from a manufacturer’s marketing description or a social-media headline.

3. Clinical Trial Phases: What They Generally Ask

Phase	General purpose	Important boundary
Phase 1	Early investigation of safety, tolerability and the product’s behaviour in humans	Often small and not generally designed to prove broad clinical efficacy
Phase 2	Explore efficacy, dosing and additional safety in a relevant patient population	Results may be preliminary and depend on study design
Phase 3	Larger studies intended to confirm efficacy and collect further safety information	Evidence is evaluated for suitability of product registration or authorisation
Phase 4	Additional study of safety, effectiveness or optimal use after marketing in relevant contexts	Does not imply every issue was already known before marketing

The HSA guidance for people considering trials, updated March 2026 sets out these common phase purposes and notes that phases may be combined. Phase 1 studies often involve healthy volunteers, but there are exceptions, such as investigations in particular patient groups. It would be misleading to teach that all products must pass one identical four-step ladder with no overlap or adaptation.

4. The Twelve-Stage Clinical Evidence Audit

Stage 1 — Copy the exact treatment claim

Does the source say an intervention is safe, effective, promising, associated with a change or authorised for a specific use? These are different assertions. Record the outcome, population, time period and source before deciding what the claim means.

Stage 2 — Identify the product and intended use

Distinguish a therapeutic product, medical device, supplement or other category under applicable law. A product sold legally for one purpose is not thereby approved for every disease, age group or claimed benefit. In classroom exercises, use invented examples instead of encouraging students to research their own medicines.

Stage 3 — Find the original study or registry record

Search for the study’s original publication or authoritative registry information rather than rely only on a press release. Record the protocol, sponsor, investigators, study dates and outcome definitions where publicly available. If no source can be located, state that provenance is unverified.

Stage 4 — Distinguish experimental from observational design

An interventional trial assigns a particular treatment according to a research plan. An observational study examines health-related outcomes under the relevant observational design. Both can contribute information, but their potential for causal inference and bias differs. A classroom should not assume every observational association proves treatment benefit.

Stage 5 — Examine comparison and allocation

Appropriate comparison groups help distinguish an intervention’s effects from other influences. Depending on the question, researchers may compare standard care, another intervention or an ethically justified control. Random allocation and masking can help manage certain biases when suitable, but neither word is a magical guarantee of a valid result.

Stage 6 — Ask what participants were told and protected from

The rights and welfare of research participants take priority over a study’s publicity value. Questions include informed consent, appropriate ethics review, qualified clinical oversight, privacy, risk monitoring and lawful authorisation. Students should learn those questions, not obtain or review real participant records.

Stage 7 — Identify the primary outcome and timeframe

A trial may measure symptom scores, laboratory changes, events, quality of life or other outcomes. Improvements in a surrogate measure do not automatically establish longer-term meaningful benefit. Ask whether the outcome was specified in advance and whether it corresponds to what the headline claims.

Stage 8 — Read the population and eligibility conditions

Find the age, relevant condition, setting and inclusion or exclusion criteria. A trial result in one group cannot automatically be transferred to pregnant people, young children, older adults or other populations without appropriate evidence. Absence from a study does not by itself mean a treatment is ineffective or harmful for the excluded group; it means the evidence is limited.

Stage 9 — Interpret benefit and harm together

Separate absolute differences, relative changes and the frequency of adverse events. A benefit may coexist with harms or uncertainties. A responsible summary should not hide adverse results simply because the average primary outcome is favourable. Nor should it treat every adverse event observed during a trial as necessarily caused by the treatment.

Stage 10 — Assess sample size and uncertainty

A difference in small groups can reflect chance, bias or real effects. Confidence intervals, analysis plans, missing data and other methodological details matter. Students should not interpret every numerical percentage as a clinically meaningful difference. Statistical and clinical significance are related but not identical.

Stage 11 — Identify regulatory status and professional boundaries

In Singapore, clinical-trial oversight and therapeutic-product authorisation involve HSA under relevant legislation and procedures. A trial’s permission to proceed is not the same as an authorisation to market a product for a proposed use, and a registered product does not guarantee benefit for every person. A qualified clinician must assess individual care.

Stage 12 — Preserve review and correction

Study findings may be revised after additional research, corrections, retractions or post-marketing safety information. Record the source version, study status and date. Update learner-facing explanations when evidence changes rather than treating the first positive headline as permanent truth.

5. Ethical Permission and Regulatory Approval Are Different Gates

HSA’s participant guidance explains that investigators conducting regulated clinical trials require both appropriate ethics review and regulatory approval before initiating a trial. The relevant Institutional Review Board (IRB) considers ethics and participant protection, while HSA has regulatory responsibilities. These bodies have different mandates; neither process is merely a decorative stamp.

HSA’s September 2026 CTA and CTN guidance distinguishes Clinical Trial Authorisation and Clinical Trial Notification routes for particular investigational products and uses. The precise application route depends on the trial. An adolescent learning the system should not attempt to determine a real protocol’s regulatory pathway from a short summary.

The HSA conducting-trials resource also identifies responsibilities for sponsors, principal investigators, informed consent and essential records. A trial cannot be ethically understood only through the final efficacy graph.

6. Worked Example: A Fictional Treatment Headline

Illustrative numerical case, not actual trial results or medical advice. A fictional study reports that 28 of 40 participants in Group A achieved a prespecified improvement while 20 of 40 participants in comparison Group B did. These are proportions of 70% and 50%, respectively. The absolute difference is 20 percentage points; the relative increase in the proportion achieving the outcome is 40% compared with Group B under these invented figures.

Students should ask which outcome was measured, when, whether groups were comparable, how participants were allocated, who was excluded, what harms occurred and how uncertain the estimated difference is. The percentages alone are insufficient for a conclusion that a real treatment works or should be used. The group sizes are small, and the example gives no statistical uncertainty or clinical context.

The exercise becomes more valuable when a new card states that the outcome was a short-term laboratory value rather than a patient-centred measure. Students must revise the claim instead of defending the initial headline. That is evidence-to-action discipline.

7. A Participant Testimonial Is Not a Controlled Comparison

A patient may genuinely report improvement after receiving an intervention. Their experience can be meaningful, but it does not alone establish what caused the change. Background care, natural variation, expectations and other factors may contribute. Ethical research investigates such questions without dismissing the person or overstating the testimonial.

The school should teach respect for experience alongside respect for reliable comparison. It must not require students to discuss their own illnesses or disclose the medical experiences of relatives. Invented testimonials work adequately for the source-criticism exercise.

8. Trial Registration and Selective Reporting

The World Health Organization’s guidance on trial registration explains why registering interventional trials publicly is a scientific and ethical responsibility: it makes studies visible and can help address publication bias and selective reporting. A trial that remains unpublished or reports only favourable results can distort the available evidence.

Registration alone does not guarantee sound methods, honest results or future approval. A student should compare available registry details with the report where appropriate, check whether outcomes were changed and recognise the possibility of legitimate documented protocol amendments. This is a higher-level research skill, not an invitation for children to audit real patients.

9. Good Evidence Includes Results That Are Not Exciting

A study that does not demonstrate its hoped-for benefit can still protect future patients and improve scientific understanding when conducted ethically and reported transparently. Not finding sufficient evidence of an effect does not always prove the effect is exactly zero. The question is what the study was sufficiently capable of detecting and what uncertainty remains.

Educators should avoid teaching that the purpose of a clinical trial is to ‘prove the medicine works’. A well-designed study tests a defined claim, informs benefit–risk evaluation and can change a hypothesis. An honest negative or inconclusive result is not wasted education.

10. Worked Example: A Study With an Age Boundary

A fictional article reports promising findings among adults in one study setting and advertises that the intervention ‘works for everyone’. Students inspect the stated participant population and identify the unsupported expansion. The source may support a narrower claim about the studied adults, subject to other limitations, but cannot automatically establish safety or efficacy for children.

The appropriate learner response is to correct the scope of the claim and identify what qualified review would be necessary before an individual decision. The lesson does not ask children to test the intervention. This complements Reference 22 on Health Information and Risk Literacy.

11. What Does Post-Marketing Evidence Add?

Evidence continues to develop when a health product is used more broadly under lawful arrangements. Rare effects, long-term patterns and differences among populations may emerge. Appropriate surveillance and professional review can lead to revised guidance or other regulatory action. The presence of post-marketing monitoring is not evidence that all approved products are unsafe; it is an acknowledgement that scientific knowledge remains open to correction.

Students should distinguish an individual adverse-event report from proof that the product caused it. Assessment of causality and significance belongs to qualified regulators and health professionals using the relevant evidence. The educational route is report responsibly, investigate through proper channels and update claims transparently.

12. A Classroom Clinical-Evidence Card

Field	What to ask	Quality control
Claim	Exactly which treatment, population and outcome?	Quote the claim faithfully
Original source	Where is the protocol, registry or study report?	Issuer, date and status
Research design	Is it interventional or observational, and what comparison exists?	Do not assume causation
Safeguards	What ethics and regulatory permissions are relevant?	Qualified authority, consent and confidentiality
Participants	Who was included and excluded?	Avoid untested generalisation
Outcome	Which specified benefit and harms were assessed?	Patient relevance and timeframe
Numbers	What are denominators, absolute changes and uncertainty?	Correct statistical interpretation
Limitations	What important question was not answered?	Confidence calibrated to evidence
Return	What future evidence could change the assessment?	Review owner and source version

13. Forty-Five-Minute Classroom Evidence Lab

Minutes	Teaching activity	Student output
0–7	Compare an advertising headline with a fictional study summary	Bounded claim
7–15	Identify study design, comparison and participants	Design-and-population table
15–23	Calculate two outcome proportions and absolute difference	Correct numeracy
23–31	Check ethics, trial phases and professional roles from official sources	Authority map
31–38	Reveal a missing adverse-event or population detail	Revised evidence conclusion
38–45	Apply the same method to a different fictional health claim	Independent transfer response

The exercise must remain educational. No student should be assigned a real treatment, asked to enrol in a trial, requested to disclose medical records or asked to advise a peer about clinical care. For younger learners, replace detailed trial analysis with the basic distinction between credible information and qualified professional help.

14. Age-Appropriate Progression

Stage	Appropriate learning	Safeguard
Early primary	Recognise that trained adults provide medical care	No clinical case studies involving peers
Upper primary	Distinguish an advertisement from a trusted general-health explanation	Fictional, non-distressing materials
Lower secondary	Identify research question, participants and outcome	No actual health experiments
Upper secondary	Compare percentages, control groups and evidence limitations	No individual medical advice
Pre-university	Study study designs, trial registration, ethics and uncertainty	Professional approval boundaries
Adult learning	Use authoritative trial summaries in evidence discussions	Consult qualified clinicians for personal decisions

15. Performance Rubric

Dimension	Beginning	Proficient	Advanced
Claim interpretation	Treats ‘promising’ as approval	States the exact population, outcome and study status	Recognises unsupported extrapolation
Design	Assumes all studies prove treatment effects	Distinguishes observational and interventional evidence	Evaluates comparison, bias and protocol limitations
Numeracy	Repeats a percentage without denominator	Computes meaningful absolute and relative differences	Explains uncertainty and clinical relevance limits
Ethics	Treats participants as data points only	Recognises consent, IRB and HSA roles	Designs a rights-protecting evidence question
Correction	Defends the original headline	Revises after new evidence	Explains what further study or monitoring is needed

16. The Gold Standard

A clinically evidence-literate learner understands that promising research, lawful investigation, credible proof and individual care decisions are different stages. They can explain why human trials require qualified oversight, interpret a limited numerical result without exaggeration and recognise that patient safety and informed participation are part of scientific quality.

The larger educational achievement is respect for both inquiry and uncertainty. Medical progress depends on excellent research, truthful reporting, professional judgement, public trust and the ability to correct a conclusion when better evidence appears.

Ministry of Education V3.0 Reference 69 | Medicines, Clinical Trials and Evidence-to-Care Literacy: From Promise to Proof