Absolute Risk vs Relative Risk in Drug Side Effects: How to Interpret Numbers

Imagine you’re told a new medication cuts your risk of a heart attack in half. That sounds impressive-until you find out your original risk was only 2%. Cutting it in half means going from 2% to 1%. That’s not a miracle drug. It’s a small benefit. But most ads don’t tell you that. They just say "50% reduction." And that’s where things get misleading.

What’s the Difference Between Absolute and Relative Risk?

Absolute risk is the real, actual chance something will happen to you. If 2 out of every 100 people in your group have a heart attack in a year, your absolute risk is 2%. Simple. Straightforward. It doesn’t change based on comparisons.

Relative risk is a comparison. It tells you how much more or less likely something is in one group versus another. If a drug lowers heart attack risk from 2% to 1%, the relative risk reduction is 50%. That’s because 1% is half of 2%. But here’s the catch: the absolute risk only dropped by 1 percentage point. That’s the real impact.

Think of it like this: Absolute risk tells you how big the puddle is. Relative risk tells you how much smaller it got after you dipped a towel in it. You need to know both to understand what actually happened.

Why Do Drug Ads Use Relative Risk?

Pharmaceutical companies know that 50% sounds better than 1%. And it’s not just ads-it’s in clinical trial reports, doctor presentations, and even patient leaflets. A drug that reduces the chance of a rare side effect from 1 in 100,000 to 1 in 10,000,000 sounds amazing: a 99.9% relative risk reduction. But the absolute drop? Just 0.099%. That’s less than one extra person helped per 1,000 people taking the drug.

This isn’t an accident. A 2021 study found that 78% of direct-to-consumer drug ads in the U.S. highlight relative risk reductions without mentioning absolute numbers. Why? Because bigger numbers sell. A 90% reduction sounds life-changing. A 0.1% reduction sounds negligible. But both can describe the same drug.

How to Calculate Absolute and Relative Risk

Let’s break it down with numbers.

Say 10 out of 100 people on a placebo (no drug) develop nausea. That’s a 10% absolute risk in the control group.

Now, 15 out of 100 people on the new drug get nauseous. That’s 15% absolute risk in the treatment group.

Absolute risk increase = 15% - 10% = 5 percentage points. So, for every 100 people taking the drug, 5 more get nauseous than if they hadn’t taken it.

Relative risk = 15% ÷ 10% = 1.5. That means you’re 1.5 times more likely to get nauseous on the drug than off it.

Relative risk increase = (1.5 - 1) × 100 = 50%. So, the risk went up by 50% relative to the placebo group.

The same math works for benefits. If a drug reduces stroke risk from 4% to 3%:

- Absolute risk reduction = 1 percentage point

- Relative risk reduction = 25% (because 3% is 75% of 4%, so you’ve cut 25% of the original risk)

The drug didn’t prevent 25% of strokes in everyone. It prevented one stroke per 100 people treated.

The Number Needed to Treat (NNT) and Harm (NNH)

This is where things get practical. NNT tells you how many people need to take a drug for one person to benefit. NNH tells you how many need to take it for one person to be harmed.

If a drug reduces heart attacks from 2% to 1%, the absolute risk reduction is 1%. So the NNT is 1 ÷ 0.01 = 100. That means 100 people need to take the drug for one heart attack to be prevented.

If a drug causes nausea in 15% of users versus 10% on placebo, the absolute risk increase is 5%. So the NNH is 1 ÷ 0.05 = 20. That means 20 people need to take the drug for one to get nauseous.

NNT and NNH turn abstract percentages into real-world numbers. If you’re told you need to take a pill every day for a year to prevent one heart attack, but 1 in 20 people will get sick from it-you can decide if that trade-off is worth it.

Real Examples That Mislead

In 2013, after the Fukushima nuclear disaster, headlines said cancer risk had "increased by 70%" in children. That sounds terrifying. But the actual absolute risk went from 0.75% to 1.25%. That’s a 0.5 percentage point increase. In a population of 10,000 kids, that’s 5 extra cases-not 70 extra cases.

Another common example: statins. Many patients refuse them because they heard it "cuts heart attack risk in half." But if their baseline risk was 1%, cutting it in half means going to 0.5%. That’s one fewer heart attack per 200 people over 10 years. For someone with low risk, the benefit is tiny. For someone with high risk-say, 10%-it’s a much bigger deal.

A 2022 Reddit thread had a primary care doctor say: "I had a patient refuse statins because they read online it cuts heart attack risk in half. They didn’t realize their risk was only 2% to begin with. They thought they’d go from 50% to 25%-not 2% to 1%."

What Experts Say

Dr. Steve Woloshin and Dr. Lisa Schwartz from Dartmouth have spent years studying how medical numbers are presented. Their research shows that only 8% of patients can accurately interpret relative risk on its own. But when absolute risk is shown with visual aids-like a grid of 100 people, where 2 are colored red for risk-the number jumps to 62% understanding.

They recommend always starting with absolute risk: "Your chance of having a stroke this year is 4%. This drug lowers it to 3%. That means for every 100 people like you, one stroke will be prevented. But 5 out of 100 will have a side effect like muscle pain." The European Medicines Agency now requires both absolute and relative risk in patient leaflets. The U.S. FDA has started pushing for the same, but enforcement is still weak. Many ads still hide the small absolute benefit behind big relative numbers.

How to Protect Yourself

You don’t need to be a statistician to understand your risks. Here’s what to ask:

• "What’s my risk without the drug?"
• "What’s my risk with the drug?"
• "How many people need to take this for one person to benefit?"
• "How many people will have a side effect?"
• "Is this benefit or risk being shown as a percentage or as a real number?"

If someone says "this reduces risk by 50%," ask: "50% of what?" Then get the baseline number.

Visual tools help. Look for risk ladders or pictograms-simple icons showing 100 people, with some shaded to show who’s affected. These are used in top medical journals and are now required in many European patient leaflets.

Why This Matters for Your Health

Misunderstanding risk leads to bad decisions. People skip life-saving drugs because they think the side effects are worse than the benefit. Others take drugs with tiny benefits and big side effects because they think the "50% reduction" means they’re protected.

A 2019 study found that 60% of doctors couldn’t correctly convert a relative risk reduction to absolute terms. If your doctor doesn’t get it, how can you?

The goal isn’t to scare you. It’s to give you real information. If a drug lowers your risk of a rare but serious side effect by 90%, but your original risk was 0.01%, that’s still just 0.001% lower. You’re not saving yourself-you’re saving one person in 10,000.

On the flip side, if a drug reduces your risk of death from 10% to 5%, that’s a 50% relative reduction-and a 5 percentage point absolute drop. That’s a major win. You need the numbers to know which is which.

What’s Changing

More medical schools now teach risk interpretation. Harvard added a required course in 2022 after finding 68% of graduating students couldn’t interpret basic stats. The Cochrane Collaboration has created standard templates for reporting risk that 37 journals now use.

By 2025, 90% of clinical trial registries will require both absolute and relative risk data to be submitted. That’s up from 65% in 2020. It’s slow, but it’s moving.

The problem isn’t the math. It’s the presentation. When you see "reduces risk by X%," always ask: "Reduced from what?" That one question gives you control over your health decisions.

Don’t let percentages trick you. Numbers don’t lie-but how they’re shown can.