Descriptive Statistics in Epidemiology

Epidemiology’s Foundation: Descriptive Statistics

Descriptive Statistics underpin all epidemiological research. Before complex modeling occurs, data must be summarized, visualized, and contextualized. This process—defining the “who, what, when, and where” of health events—provides the critical framework for identifying patterns, generating hypotheses, and allocating resources. This guide dissects the core statistical measures used to transform raw health data into actionable public health intelligence.

The CDC Principles of Epidemiology emphasize that descriptive statistics characterize disease distribution. Without this foundation, calculating analytical metrics like risk ratios remains impossible.

Epidemiological Data Types

Statistical test selection depends entirely on data classification.

Categorical (Qualitative) Data

Nominal: Unordered categories (e.g., Gender, Blood Type, Yes/No).
Ordinal: Logical order with unequal intervals (e.g., Cancer Stages I-IV, Likert Scales).

Continuous (Quantitative) Data

Interval: Ordered with equal intervals; no true zero (e.g., Celsius Temperature).
Ratio: Ordered with true zero (e.g., Height, Weight, Blood Pressure).

Measures of Central Tendency

These metrics identify the “center” or typical value of a distribution.

• Mean (Average): Sum of values / Count. Sensitive to outliers. Use for normally distributed data.
• Median (Middle): The 50th percentile. Resistant to outliers. Best for skewed data (e.g., incubation periods).
• Mode (Most Frequent): The most common value. Useful for categorical data.

Measures of Dispersion (Spread)

Dispersion describes data variability around the center.

• Range: Difference between maximum and minimum values. Simple but sensitive to extremes.
• Interquartile Range (IQR): Range of the middle 50% (Q3 – Q1). Used with the Median.
• Standard Deviation (SD): Average distance of data points from the Mean. Critical for confidence interval calculation.
• Variance: The square of the Standard Deviation.

Normal vs. Skewed Distributions

Understanding data shape determines the appropriate summary statistic.
Normal Distribution (Bell Curve): Symmetrical. Mean = Median = Mode. Use Mean and SD.
Skewed Distribution: Asymmetrical.
– Positively Skewed (Right): Tail extends right. Mean > Median. Example: Income data.
– Negatively Skewed (Left): Tail extends left. Mean < Median. Example: Age at death in developed countries.
Rule: For skewed data, report Median and IQR.

Measures of Frequency

The core vocabulary of epidemiology.

Ratios, Proportions, and Rates

Ratio: Comparison of two independent values (A/B).
Proportion: Numerator included in denominator (A/(A+B)).
Rate: Proportion with time element (New cases per 1,000 person-years).

Prevalence vs. Incidence

Prevalence: The “snapshot.” Total existing cases at a specific time. Measures burden.
Incidence: The “video.” New cases developing over a period. Measures risk.

Mortality Metrics

Specific rates measure death impact.
Case Fatality Rate (CFR): (Deaths from disease / Diagnosed cases) x 100. Measures virulence.
Crude Mortality Rate: Total deaths / Total population.
Cause-Specific Mortality Rate: Deaths from specific cause / Total population.
Proportionate Mortality Ratio (PMR): Deaths from specific cause / Total deaths.

Standardization

Comparing populations requires adjustment for confounding variables like age.
Crude Rate: The actual observed rate. Misleading if populations differ in age structure.
Adjusted Rate: A hypothetical rate calculated to allow fair comparison between populations with different demographics (e.g., Florida vs. Alaska).

Visualizing Epidemiological Data

Histograms: Continuous data (age distribution).
Bar Charts: Categorical data (disease rates by country).
Box Plots: Visualizing Median, IQR, and outliers.
Scatter Plots: Relationships between two continuous variables (BMI vs. BP).
Epi Curve: Histogram of cases over time. Reveals outbreak type (Point Source, Continuous, Propagated).

FAQs: Descriptive Statistics

Conclusion

Descriptive statistics are the lens through which public health data becomes visible. By mastering central tendency, dispersion, and frequency measures, epidemiologists transform raw numbers into narratives that drive health policy and intervention.