How to Read | Class No.0029 | The Health Information Read — Read Risk, Evidence, Benefits, Harms, Labels and Guidance Without Turning General Information into Personal Diagnosis

Class No.0021: The Scientific Read · Class No.0012: The Quantitative Read · Class No.0016: The Uncertainty Read

General health information is not personal diagnosis

A symptom list describes possibilities. It does not tell one reader which condition they have.

A study reports patterns in a group. It does not automatically predict one individual.

A medication guide explains approved use and safety information. It does not replace the prescriber’s knowledge of the patient’s situation.

This boundary should stay visible throughout health reading.

Start by naming the information type

Health texts include:

• patient information;
• medicine labels;
• clinical guidelines;
• research studies;
• systematic reviews;
• public-health advice;
• news reports;
• advertising;
• personal testimony.

Different genres answer different questions.

Read benefits and harms together

Current FDA consumer guidance emphasises that medicines have both benefits and risks, and that patients should understand both when making decisions with health professionals.

A strong reader therefore asks:

• What benefit is expected?
• How large?
• For whom?
• What harms are known?
• How common and how serious?
• What monitoring is needed?

Risk is not certainty

A risk is a probability, not a prediction that an outcome will definitely happen.

“Doubles the risk” sounds large, but the baseline matters.

If risk rises from 1 in 10,000 to 2 in 10,000, the relative risk doubles while absolute change remains 1 in 10,000.

Read both.

Relative and absolute risk answer different questions

Relative risk: how much larger or smaller one risk is compared with another.

Absolute risk: the actual probability difference.

Health communication is stronger when both are available.

Number needed to treat and number needed to harm

Advanced readers may encounter measures describing how many people would need to receive an intervention for one additional benefit or harm to occur over a stated time.

These values depend on population, baseline risk and time horizon.

Time horizon matters

A 1% annual risk is not the same as a 1% lifetime risk.

A treatment benefit measured over six weeks is not evidence of the same benefit over ten years.

Always attach a time period.

Population matters

Ask who was studied:

• children or adults;
• healthy volunteers or patients;
• one sex or mixed population;
• one country or many;
• mild or severe disease.

Generalisation should match evidence.

Association is not diagnosis

If a symptom is associated with several conditions, one symptom cannot identify which condition is present.

Diagnostic reasoning combines history, examination, testing and professional judgement.

Screening is not diagnosis

A screening test identifies people who may need further assessment.

Positive screening can include false positives; negative screening can include false negatives.

Read what the test is designed to do.

Sensitivity and specificity describe different performance

Sensitivity concerns how often a test identifies people who have the condition.

Specificity concerns how often it correctly identifies people who do not.

Neither alone tells an individual the probability that a positive result means they have the condition.

Base rates change interpretation

The prevalence of a condition affects the meaning of test results.

Use Class No.0012’s base-rate discipline.

Read clinical study design

Ask:

• randomised or observational?
• control group?
• blinding?
• sample size?
• duration?
• primary outcome?
• dropout?

Method controls claim strength.

Surrogate outcomes are not always patient outcomes

A treatment may improve a biomarker. That can be important, but the reader should ask whether the biomarker is known to translate into outcomes patients actually experience.

Composite outcomes combine several events

If a study reports one combined endpoint, inspect which components drive the result.

A dramatic composite can be driven by a less serious but more frequent component.

Subgroup findings need caution

When many subgroups are tested, some differences can appear by chance.

Ask whether subgroup analysis was pre-specified and supported across evidence.

Statistical significance is not clinical importance

A tiny difference can be statistically detectable in a large sample.

Ask whether the effect is large enough to matter to patients.

Confidence intervals show precision

Wide intervals indicate more uncertainty.

Do not treat a single point estimate as exact.

Systematic reviews organise bodies of evidence

They can be more informative than one isolated study when they are rigorous and the underlying studies are appropriate.

Read inclusion criteria, heterogeneity and evidence quality.

Guidelines are recommendations built from evidence plus judgement

Clinical guidelines may combine evidence, benefit-harm balance, feasibility and patient values.

They are not identical to raw research results.

Medicine labels are structured documents

FDA consumer guidance for over-the-counter medicines emphasises reading the label each time because active ingredients, dosages and warnings can differ even among similar brand names.

Read:

• active ingredient;
• purpose;
• uses;
• warnings;
• directions;
• inactive ingredients;
• storage and other information.

Brand name is not enough

Two products with similar branding can contain different active ingredients or strengths.

Read the active ingredient every time.

Warnings have action logic

A warning may say:

• do not use if;
• ask a doctor before use if;
• stop use and seek help if;
• do not exceed;
• keep out of reach.

These phrases create different actions.

Dose is amount plus timing plus route

A number without frequency and formulation is incomplete.

Read the whole instruction.

Missed-dose advice is medicine-specific

Do not generalise from one medicine to another.

Use the medicine’s approved information and professional advice.

Interactions are relational

Medicine–medicine, medicine–food and medicine–condition interactions depend on combinations.

One product cannot be read in isolation from the person’s other medicines and conditions.

“Natural” does not mean risk-free

Natural origin tells you where something comes from, not whether it is effective, safe, pure or appropriate.

Read evidence and regulation.

Anecdotes are experience, not prevalence

A personal story can reveal lived experience.

It cannot establish how common an outcome is or whether the intervention caused it.

Testimonials are vulnerable to selection

People with dramatic experiences are more likely to speak publicly.

Absence of unsuccessful stories does not prove universal success.

Before-and-after images need context

Lighting, angle, timing, selection and concurrent interventions can change appearance.

Visual persuasion is not controlled evidence.

Health headlines often overcompress

A study may show association in one population while a headline says “X prevents Y.”

Trace headline → article → original study.

Read the source institution

Prefer authoritative public-health bodies, regulators, professional organisations and peer-reviewed evidence for consequential claims.

Then still read the evidence, scope and update date.

Recency matters

Health guidance changes as evidence, products and outbreaks change.

Check publication and update dates.

Consensus can change

Scientific revision is part of medicine.

A changing recommendation does not automatically mean earlier advice was dishonest; it may reflect new evidence.

Conflicts of interest deserve context

Funding source and financial relationships can matter.

They should trigger scrutiny of method and independent replication, not automatic dismissal.

Risk communication should preserve uncertainty

A health text should distinguish known, probable, possible and unknown.

Readers should do the same.

Do not turn mechanism into outcome

A plausible biological mechanism is not enough to prove a treatment works in people.

Clinical outcomes require clinical evidence.

Do not turn animal or laboratory evidence directly into human treatment claims

Early-stage evidence can be scientifically valuable while remaining far from proven human benefit.

Healthy scepticism is not blanket distrust

Ask evidence questions without assuming every institution is lying or every claim is true.

Calibration matters.

Kindergarten: body information versus diagnosis

Teach children to name simple body states—tired, thirsty, sore—and to tell a trusted adult rather than diagnose themselves.

Primary 1: labels and safety words

Use non-medicine examples first: warnings, expiry dates, instructions.

Primary 2: source recognition

Compare a school-health poster with an advertisement.

Primary 3: fact versus testimonial

Ask whether one person’s experience proves what happens to everyone.

Primary 4: risk language

Teach possible, likely, rare and common conceptually.

Primary 5: medicine label structure

Use an educational mock label to identify active ingredient, warnings and directions.

Primary 6: absolute versus relative risk

Use simple safe examples with counts.

Secondary 1: study versus headline

Compare cautious study language with stronger media phrasing.

Secondary 2: screening and diagnosis

Teach false positives and false negatives conceptually.

Secondary 3: benefit-harm table

Read one intervention through benefit, harm, uncertainty and population.

Secondary 4: evidence hierarchy

Compare anecdote, observational study, trial and systematic review by what each can establish.

JC: full health-claim architecture

Map:

condition → intervention → population → comparator → outcome → effect size → uncertainty → harm → applicability

Then separate population evidence from personal decision-making.

The Health Information Read protocol

1. Name the information type.
2. Separate general evidence from personal diagnosis.
3. Identify population and time horizon.
4. Read benefit and harm together.
5. Translate relative and absolute risk.
6. Read method and uncertainty.
7. Check active ingredients, warnings and directions on labels.
8. Trace dramatic claims to original evidence.
9. Check recency and authoritative guidance.
10. Escalate personal decisions to qualified professionals.

The Health Claim Card

Claim: ______
Population: ______
Intervention/exposure: ______
Comparator: ______
Benefit: ______
Harm: ______
Absolute risk: ______
Uncertainty: ______
Personal decision? professional guidance needed

Common failure: symptom list becomes diagnosis

Repair: preserve differential possibilities and seek professional assessment when needed.

Common failure: relative risk without baseline

Repair: absolute-risk translation.

Common failure: testimonial becomes treatment evidence

Repair: separate experience from causal evidence.

Common failure: label read by brand only

Repair: active ingredient and warning scan.

Common failure: one study becomes medical consensus

Repair: wider evidence and guideline context.

Three learners, three health-reading failures

Maren reads health articles carefully but turns population associations into personal conclusions. Her repair is applicability control.

Iona sees every side effect and treats all risks as equally likely. Her repair is absolute probability and severity reading.

Leonie trusts labels but skips active ingredients when buying a familiar brand. Her repair is ingredient-first reading.

A 60-minute Health Information Read tutorial

1. 10 minutes: information type and source.
2. 10 minutes: population evidence versus individual decision.
3. 10 minutes: relative and absolute risk.
4. 10 minutes: study design and uncertainty.
5. 10 minutes: label anatomy.
6. 10 minutes: compare headline with original evidence.

The Health Information Read audit

Ask:

• Is this general information or individual medical advice?
• Who was studied?
• What benefit was measured?
• What harms were measured?
• What is the absolute risk?
• How certain is the evidence?
• What does the label actually say?
• Is this the current guidance?
• Is the source authoritative and transparent?
• Am I making a personal health decision that needs professional guidance?

Reference anchors

• FDA — Managing the Benefits and Risks of Medicines
• FDA — The Over-the-Counter Drug Facts Label
• FDA — Patient Labeling Resources
• Cochrane — Evidence-Based Health Care

Continue through SETC

• How to Read | Class No.0021 | The Scientific Read
• How to Read | Class No.0012 | The Quantitative Read
• How to Read | Class No.0016 | The Uncertainty Read

Final lesson

Health reading is calibrated caution.

Read the population before applying the result to one person. Read benefit beside harm. Read absolute risk beside relative risk. Read the label every time. Read uncertainty as information rather than inconvenience.

And when the reading becomes a decision about an actual person’s diagnosis, medicine or treatment, move from general information to qualified professional care.

Return to the English Learning Library · Advanced English / General Paper

Continue with the How to Read collection.