 # Measures of frequency of disease

In observational studies, where the occurrence of a condition is of interest, we must initially define the study population. We often refer to the study population as the population at risk for developing the condition or the disease. In observational studies, the study population is defined through our research question and the choice of case definition.

Once we have defined what constitutes the case and the study population, we can quantify the occurrence of the outcome of interest with reference to the size of the population at risk using the following 2 common types of measures of occurrence.

• 1.

Prevalence is the number of existing cases (prevalent cases) divided by the number of persons in the study population at a defined point in time:

Prevalence = number of cases in the study population at a specified time number of persons in the study population at a specified time

Prevalence can be interpreted as the probability that a certain person is a case at the defined time point; like all probabilities, it takes values from 0 to 1, or 0% to 100%. Prevalence is a measure of occurrence with no units; however, the point in time should be specified. The term “prevalence rate” is often incorrectly used instead of “prevalence.” Prevalence is a proportion and not a rate; therefore, those terms should not be confused. An example of prevalence is the proportion of people aged 20 to 30 years with confirmed periodontal disease in the study population at a defined time point. Prevalence is like a snapshot in time indicating how many have the condition and how many do not have the condition regardless of when the condition developed. For the prevalence of periodontal cases, we are not distinguishing old from new cases. Prevalence can be encountered either as point prevalence (as described above) or as period prevalence, which answers the question of whether someone had the condition during a specified period of weeks, months, years, and so on.

• 2.

Incidence, in contrast to prevalence, quantifies the number of new cases (or incident cases) that develop in a population at risk during a specified time interval. Incidence is usually calculated by using incidence risk or incidence rate.

Incidence risk or cumulative incidence indicates the proportion of noncases in a population that subsequently becomes cases within a specified time interval:

Cumulative incidence = number of new cases at a specified time period number of noncases at the beginning of that time period