When calculating disease prevalence, should I express it as a percentage (multiply by 100) or per 1,000 (multiply by 1,000), and how do I choose the appropriate multiplier?

Expressing Disease Prevalence: Choosing the Right Multiplier

Express prevalence using the multiplier that makes the resulting number fall between 1 and 100 for optimal interpretability—use per 100 (percentage) for common diseases and per 1,000 or higher for rare conditions. 1

The Core Principle

The choice of multiplier (×100, ×1,000, ×10,000, or ×100,000) is driven by practical communication rather than statistical rules:

• Calculate the raw proportion first: Divide the number of cases by the total population 1
• Select the multiplier that yields interpretable numbers: The goal is to produce values that are easy to understand and compare 1, 2

When to Use Each Multiplier

Use Per 100 (Percentage) When:

• The disease is relatively common (prevalence ≥1%): For example, expressing SARS-CoV-2 prevalence as 0.5%, 1%, or 2% is standard practice 1
• The resulting number is ≥1: A prevalence of 0.5% is more intuitive than "5 per 1,000" 1
• Communicating with general audiences: Percentages are universally understood 1

Use Per 1,000 When:

• The disease is uncommon (prevalence 0.1% to <1%): This prevents awkward decimals like "0.32%" 1
• Presenting absolute risk data: The AGA guideline consistently presents outcomes as "per 1,000 people" when discussing test performance 1
• The numbers remain manageable: For example, "3 per 1,000" is clearer than "0.3%" 1

Use Per 10,000 or 100,000 When:

• The disease is rare: Ménière's disease prevalence ranges from 3.5 to 513 per 100,000, making this the appropriate denominator 1
• Epidemiologic surveillance: Public health reporting often uses per 100,000 for standardization across populations 1
• The condition affects <0.1% of the population: Using percentages would require multiple decimal places 1

Practical Algorithm for Decision-Making

Step 1: Calculate the raw prevalence (cases ÷ population) 1

Step 2: Apply this decision tree:

• If raw prevalence ≥0.01 (≥1%) → multiply by 100 and express as percentage 1
• If raw prevalence is 0.001 to <0.01 (0.1% to <1%) → multiply by 1,000 1
• If raw prevalence is 0.0001 to <0.001 (0.01% to <0.1%) → multiply by 10,000 1
• If raw prevalence <0.0001 (<0.01%) → multiply by 100,000 1

Step 3: Verify the final number is between 1 and 100 for optimal readability 1

Common Pitfalls to Avoid

• Mixing denominators within the same report: Choose one multiplier and use it consistently for all comparisons 1
• Using percentages for very rare diseases: Expressing prevalence as "0.0082%" is less clear than "8.2 per 100,000" 1
• Ignoring your audience: Clinical contexts may favor percentages, while epidemiologic reports typically use per 100,000 1
• Forgetting that the multiplier affects interpretation of confidence intervals: When prevalence is 0.5% with a test specificity of 94%, expressing results "out of 1,000 people" makes the 60 false positives more tangible than saying "6%" 1

Context-Specific Considerations

For Diagnostic Test Performance:

• Low-prevalence settings (<0.5%): Express as percentage but present absolute numbers per 1,000 to illustrate false positive burden 1, 2
• The CDC approach: When discussing positive predictive value in populations with 2% prevalence, results are shown both as percentages (44% PPV) and as absolute numbers (16 true positives out of 36 positive tests per 1,000 screened) 2

For Population Health Reporting:

• Standardize to per 100,000 for cross-population comparisons, as demonstrated in Ménière's disease epidemiology where estimates ranged from 3.5 to 513 per 100,000 across different geographic regions 1