Incidence and Prevalence

Incidence counts new cases of a disease, while prevalence counts all existing cases in a population. Incidence measures the occurrence of disease over time among people at risk; prevalence measures the proportion of a population living with the disease at a specified time or during a specified period. They describe different aspects of disease burden and are not interchangeable.

Incidence: measuring new disease occurrence

Incidence concerns the transition from being disease-free to developing disease. Its denominator must represent people who are at risk of developing the condition.

• Cumulative incidence, also called incidence proportion, is the number of new cases during a specified period divided by the number of disease-free people at risk at the beginning of that period.
• Cumulative incidence expresses the risk of developing disease over that period, provided follow-up is adequately accounted for.
• Incidence rate is the number of new cases divided by the total person-time at risk contributed by the population under observation.
• Incidence is useful for studying disease causes and assessing whether prevention reduces the occurrence of new cases.

Prevalence: measuring existing disease

Prevalence includes both newly developed cases and cases that began earlier but remain present. It therefore reflects how widespread a condition is, rather than how quickly new cases arise.

• Point prevalence is the number of people with the disease at a specified point in time divided by the population at that time.
• Period prevalence counts people who had the disease at any time during a specified period, including existing cases and new cases arising during it.
• Prevalence is a proportion, not a measure of the speed at which disease develops.
• Prevalence helps estimate requirements for treatment, medicines, rehabilitation and continuing care, particularly for chronic conditions.

Relationship and interpretation

Incidence measures the inflow of new cases; prevalence reflects the accumulated pool of cases. That pool depends on both new disease occurrence and how long people remain affected.

• For a relatively uncommon disease under stable conditions, prevalence is approximately incidence rate multiplied by average disease duration.
• Longer survival with disease can increase prevalence even when incidence does not rise.
• Recovery or death reduces the pool of existing cases, so low prevalence does not necessarily mean low incidence.
• Neither measure alone captures severity, disability or mortality; these require additional indicators of disease burden.