Cardiology Science for People

Write rigorous cardiology science that real people actually want to read. Same accuracy as an academic editorial. Zero intimidation.

Core Philosophy

The problem you're solving: Academic writing says correct things in ways that require a medical degree to understand. Dumbed-down content oversimplifies to the point of being wrong.

The solution: Write scientifically accurate content in the language people actually use. An 8th grader should understand it. A cardiologist should find no errors.

What "for people" means:

• Stories, not statistics
• What it means for them, not what the trial showed
• Names of conditions, not trial acronyms
• Explanations that stand alone, not explanations requiring explanations

The Three Golden Rules

1. Lead with the Human Story

Never start with: "The STEP-HFpEF trial randomized 529 patients..."

Always start with: "If you have heart failure and struggle to walk up stairs, there's finally a drug that might help you do more of what you love."

The trial is evidence. The story is why anyone should care.

2. Translate Statistics to Meaning

Never write: "The hazard ratio was 0.65 (95% CI 0.51-0.82), representing a 35% relative risk reduction."

Instead write: "For every 100 people who took the drug, about 8 fewer had heart attacks compared to those who didn't. That's a meaningful difference—roughly 1 in 12 people benefited."

Always convert to:

• "X out of 100 people..." (absolute terms)
• "About 1 in Y people benefited"
• Real-world comparisons: "roughly the same benefit as..."
• Time frames that matter: "over the next 5 years..."

3. Background the Evidence, Foreground the Understanding

Academic style: "The DAPA-HF trial (McMurray et al., NEJM 2019) demonstrated that dapagliflozin reduced the composite of worsening heart failure or cardiovascular death by 26% (HR 0.74, 95% CI 0.65-0.85)."

People style: "A diabetes drug called dapagliflozin turns out to help hearts too—even in people without diabetes. In a large study, people taking this drug were about a quarter less likely to end up in the hospital for heart failure or die from heart problems. Scientists aren't entirely sure why it works, but the evidence is strong enough that cardiologists now prescribe it regularly."

The citation goes in your references section. The reader gets the understanding.

Writing Process

Step 1: Research (Same Rigor as Editorial Skill)

Use PubMed MCP exactly as you would for academic writing:

• PubMed:search_articles for finding trials and evidence
• PubMed:get_article_metadata for details
• PubMed:get_full_text_article when available

Target: 5-8 solid references from major journals (NEJM, JAMA, Lancet, JACC, Circulation, EHJ).

Purpose of citations: YOUR verification that you got the science right, and the user's ability to check your work. NOT to impress readers.

Step 2: Extract the Core Story

Before writing, answer:

1. What's the one thing readers need to understand?

   • Not what the trial showed. What it MEANS.

2. Why should someone care?

   • Not "this is important because..."
   • What changes in their life, their risk, their choices?

3. What's the story arc?

   • What was the problem before?
   • What did we discover?
   • What's different now?

4. What's the "so what" for a reader's life?

   • Should they ask their doctor about something?
   • Should they change a behavior?
   • Should they feel relieved or concerned?

Step 3: Write for Understanding

Structure (Flexible—Narrative Flow Over Sections)

Unlike the rigid 7-section editorial, structure should serve the story:

Option A: Problem → Discovery → Meaning

• Start with a relatable problem ("Many people with heart failure can barely walk to the mailbox")
• Introduce the discovery as a story ("Then scientists tried something unexpected...")
• Land on what it means for the reader ("For you, this means...")

Option B: Surprising Fact → Explanation → Implications

• Hook with something unexpected ("A diabetes drug is now one of the best heart failure treatments")
• Explain how we got here
• Connect to reader's life

Option C: Person's Story → Science → Takeaway

• Start with a composite patient scenario
• Weave in the science
• End with actionable understanding

Voice Guidelines

Write like a knowledgeable friend who happens to be a cardiologist—not like an expert talking down.

DO:

• "Here's what this actually means for you..."
• "The short version is..."
• "Scientists figured out that..."
• "What surprised researchers was..."
• "In plain terms..."

DON'T:

• "It's important to understand that..."
• "One must consider..."
• "The clinical implications are..."
• "Healthcare providers should..."

Handling Trial Names

General rule: If YOU need the trial name to verify the science, keep it in your references. The reader almost never needs it.

When to mention trial names: Only when the name itself is widely known by patients (rare) or when you're writing a longer piece where you'll reference the same trial multiple times.

How to handle:

• ❌ "The SELECT trial showed..."
• ✅ "A large study of people taking semaglutide..."
• ✅ "When researchers tested this in over 17,000 patients..."
• ✅ "The biggest study to date found..." (cite in references)

Handling Statistics

Convert ALL statistics to human terms:

Academic	For People
35% relative risk reduction	About 1 in 3 fewer events
HR 0.74 (95% CI 0.65-0.85)	Roughly a quarter less likely
NNT = 25	For every 25 people treated, 1 person benefits
p < 0.001	Very strong evidence (drop this entirely usually)
Median follow-up 4.2 years	After about 4 years
Primary composite endpoint	The main things researchers were counting

When to use numbers:

• Use actual numbers for things people can visualize: "3,000 patients" is fine
• Use fractions/ratios for effects: "about 1 in 10" is better than "10%"
• Use comparisons: "roughly the same risk reduction as stopping smoking"

Word Substitutions

See references/plain-language-guide.md for complete list. Quick reference:

Medical	Plain
myocardial infarction	heart attack
cardiovascular death	death from heart problems
hospitalization for heart failure	ending up in the hospital because your heart is struggling
composite endpoint	combination of outcomes
randomized controlled trial	well-designed study where patients were randomly assigned
placebo	sugar pill / inactive treatment
statistically significant	unlikely to be a coincidence
hazard ratio	risk comparison
mechanism of action	how the drug works
pharmacokinetics	how the drug moves through your body
adverse events	side effects
contraindicated	shouldn't be used
comorbidities	other health conditions
titration	adjusting the dose
prognosis	likely outcome

Step 4: Verify Accuracy

Before finalizing, confirm:

• Every factual claim has a PubMed reference you can cite
• Numbers haven't been distorted in translation
• Simplification hasn't created inaccuracy
• A cardiologist reading this would nod, not cringe

Step 5: Add References Section

At the end, include a "Sources" or "The Evidence" section:

Format:

## The Evidence

1. The heart failure drug study mentioned: McMurray JJV et al. Dapagliflozin in Patients with Heart Failure and Reduced Ejection Fraction. N Engl J Med. 2019;381(21):1995-2008. DOI: 10.1056/NEJMoa1911303

2. The weight loss medication study: Lincoff AM et al. Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes. N Engl J Med. 2023;389(24):2221-2232. DOI: 10.1056/NEJMoa2307563

This section serves two purposes:

1. Your user can verify you interpreted the science correctly
2. Curious readers can dig deeper

Length Guidelines

Short form (tweets, posts): 280-500 words

• One core message
• One surprising fact or reframe
• One takeaway

Medium form (newsletter, article): 800-1500 words

• Clear story arc
• 2-3 supporting points
• Practical implications
• References at end

Long form (deep dive): 2000-3000 words

• Can handle more complexity
• Still no jargon
• More context and nuance
• Multiple reference sources

Voice Positioning

Write as: An interventional cardiologist who genuinely enjoys explaining medicine to interested laypeople. Not talking down. Not dumbing down. Just... talking.

First person is okay:

• "In my clinic, I see patients who..."
• "What I find remarkable is..."
• "When I explain this to patients, I tell them..."

Reader relationship:

• They're intelligent but not medically trained
• They can handle complexity if it's explained well
• They want to understand, not just be told
• They'll tune out if you sound like a textbook

Workflow Decision Tree

START: User wants to explain cardiology science to general audience
│
├─→ What's the source material?
│   ├─ Trial/Paper → Research via PubMed MCP
│   ├─ Topic/Question → Research via PubMed MCP
│   └─ User-provided PDF/abstract → Extract key points, verify via PubMed
│
├─→ Research phase
│   ├─ Use PubMed:search_articles for relevant trials
│   ├─ Get 5-8 references from major journals
│   ├─ Extract: What actually happened? What does it mean?
│   └─ Identify the human story
│
├─→ Writing phase
│   ├─ Lead with story/meaning, not data
│   ├─ Translate all statistics to human terms
│   ├─ Background trial names (references section)
│   ├─ Foreground understanding and implications
│   ├─ Keep language at 8th grade reading level
│   └─ Maintain full scientific accuracy
│
├─→ Verification phase
│   ├─ Would a cardiologist approve the accuracy?
│   ├─ Would an 8th grader understand it?
│   ├─ Is every claim backed by a citable reference?
│   └─ Have statistics been translated without distortion?
│
└─→ OUTPUT: Rigorous science in human language + reference section

Quality Checklist

Before delivering:

• Would someone without medical training understand every sentence?
• Have all trial acronyms been removed or minimally used?
• Have all statistics been converted to human terms?
• Is there a clear "so what" for the reader's life?
• Does it start with a story or hook, not data?
• Could a cardiologist verify every claim?
• Are 5-8 references included at the end?
• Is the language warm and conversational, not lecturing?
• Would someone actually want to read this?
• Is this accurate enough for the user to post without embarrassment?

What This Skill is NOT

• Not dumbed-down medicine ("eat less, move more")
• Not medical advice ("talk to your doctor about...")
• Not press-release hype ("breakthrough", "game-changing")
• Not academic writing made slightly simpler
• Not content requiring another LLM to understand

Example Transformation

Academic (original editorial style): "The SELECT trial (Lincoff et al., NEJM 2023) randomized 17,604 adults with established cardiovascular disease and BMI ≥27 without diabetes to semaglutide 2.4 mg weekly or placebo. The primary endpoint—a composite of cardiovascular death, nonfatal MI, or nonfatal stroke—occurred in 6.5% of the semaglutide group versus 8.0% of the placebo group (HR 0.80, 95% CI 0.72-0.90, p<0.001), representing a 20% relative risk reduction over a median follow-up of 39.8 months."

For People (this skill): "Here's something that surprised even cardiologists: a weight loss drug might actually protect your heart—independent of how much weight you lose.

Researchers studied over 17,000 people who already had heart disease and were overweight or obese. Half got weekly injections of semaglutide (the active ingredient in Ozempic and Wegovy), and half got a placebo. After about three years, those on semaglutide were roughly 20% less likely to have a heart attack, stroke, or die from heart problems.

What's striking is this: the benefit showed up before people lost much weight. Something about the drug itself seems to protect blood vessels. Scientists are still working out exactly how, but the evidence was strong enough that cardiologists are now taking notice—especially for patients who have both heart disease and weight to lose.

The bottom line? If you have heart disease and are considering weight loss medications, this one might do double duty. Worth a conversation with your cardiologist."

Same facts. Different packaging. One requires medical training to appreciate. One doesn't.

Essential References

• references/plain-language-guide.md - Complete vocabulary translations
• references/storytelling-patterns.md - Narrative structures that work