Prevalence

Prevalence is the proportion of a defined population that has a particular disease, condition, or characteristic at a specific point in time (point prevalence) or during a specified period (period prevalence).^[1][2] It is a fundamental epidemiological measure that quantifies the burden of a condition in a population and is expressed as a proportion, a percentage, or a number per population unit (such as per 100,000).

Prevalence data are indispensable for public health planning, healthcare resource allocation, and health policy development. Knowing how many people currently have a condition determines the need for treatment facilities, medications, and healthcare personnel. In diagnostic research, prevalence directly affects positive and negative predictive values, meaning the same diagnostic test performs differently in populations with different disease frequencies.^[1] Health economists use prevalence estimates in disease burden calculations and cost effectiveness models. Epidemiologists track changes in prevalence over time to evaluate the impact of public health interventions, identify emerging health threats, and monitor trends in chronic diseases.

Point prevalence is calculated by dividing the number of individuals who have the condition at a specific moment by the total population at that same moment.^[1][2] Period prevalence extends this concept over a defined time window (such as one year) and includes both individuals who had the condition at the start of the period and those who developed it during the period. The denominator is typically the mid period population or the average population over the interval.

Prevalence differs conceptually from incidence, which measures only new cases arising during a specified period. In a steady state, and when prevalence is low, the two are linked by the approximate formula: prevalence equals incidence rate multiplied by average disease duration.^[1] This means that conditions with long duration (such as diabetes or osteoarthritis) can have high prevalence even with moderate incidence, while conditions with short duration (such as influenza) have a point prevalence much lower than their incidence over a season.

Accurate prevalence estimates require careful attention to case definition, sampling methodology, and ascertainment methods.^[3] Population based cross sectional surveys provide the most representative estimates but are expensive and logistically complex. Administrative databases and disease registries offer alternative sources but may underestimate prevalence if not all cases seek care or are recorded. Screening studies detect previously undiagnosed cases and may reveal higher true prevalence than estimates based on diagnosed cases alone. When prevalence is estimated from a sample, confidence intervals should be reported to reflect sampling uncertainty, and the sampling design should be described in sufficient detail for readers to assess representativeness.

Use prevalence when you need to describe the current burden of a disease or condition in a population, plan healthcare resources, or calculate predictive values for diagnostic tests. Point prevalence is appropriate for conditions that are present at a single moment, while period prevalence is more suitable for conditions that fluctuate or have episodic presentations. Use incidence rather than prevalence when the research question concerns the rate at which new cases arise, such as when evaluating risk factors or assessing the impact of preventive interventions.

A national health survey of 150,000 adults conducted over three months finds that 9,750 respondents meet criteria for a current major depressive episode at the time of their interview, yielding a point prevalence of 6.5%. Including these current cases, 18,000 respondents meet criteria for a major depressive episode at some point in the past 12 months, giving a 12 month period prevalence of 12.0%. Public health planners use these figures to estimate the need for mental health services and to track whether depression rates are increasing compared to previous survey cycles.