Prevalence vs. Incidence Rate

Measuring Disease Occurrence and Causal Effects 53
already succumbed to an infectious disease may no longer be susceptible to it for

some period of time. If a substantial proportion of a population is affected by the
outbreak, the number of susceptible people will decline gradually as the epidemic
progresses and the attack rate increases. This change in the susceptible population
leads to a more rapid decline over time in the number of new cases compared
with the incidence rate in the susceptible population. The incidence rate declines
more slowly than the number of new cases because in the incidence rate, the
declining number of new cases is divided by a dwindling amount of susceptible
person-time.
A propagated epidemic is one in which the causal agent is transmitted through
a population. Influenza epidemics are propagated by person-to-person transmis-
sion of the virus. The epidemic of lung cancer during the 20th century was a
propagated epidemic attributable to the spread of tobacco smoking through many
cultures and societies. The curve for a propagated epidemic tends to show a more
gradual initial rise and a more symmetric shape than for a point-source epidemic
because the causes spread gradually through the population. Transmission of
infectious disease within a population is discussed further in Chapter 6, which
also presents the Reed-Frost model, a simple model that describes transmission
of an infectious disease in a closed population.
Although we may think of point-source epidemics as occurring over a short
time span, they are not always briefer than propagated epidemics. The epidemic
of cancer attributable to exposure to the atomic bombs detonated in Hiroshima
and Nagasaki was a point-source epidemic that began a few years after the explo-
sions and continues into the present. Another possible point-source epidemic
that occurred over decades was an apparent outbreak of multiple sclerosis in the
Faroe Islands that followed the occupation of those islands by British troops dur-
ing the Second World War4 (although this interpretation of the data has been
questioned5). Propagated epidemics can occur over extremely short time spans.
An example is epidemic hysteria, a disease often propagated from person to
person in minutes. An example of an epidemic curve for a hysteria outbreak is
depicted in Figure 4–5. In this epidemic, 210 elementary school children devel-
oped symptoms of headache, abdominal pain, and nausea. These symptoms were
attributed by the investigators to hysteric anxiety.6
Prevalence Proportion
Incidence proportion and incidence rate are measures that assess the frequency of
disease onsets. The numerator of either measure is the frequency of events that
are defined as the occurrence of disease. In contrast, prevalence proportion, often
referred to simply as prevalence, does not measure disease onset. Instead, it is a
measure of disease status.
The simplest way of considering disease status is to consider disease as being
either present or absent. The prevalence proportion is the proportion of people
in a population who have disease. Consider a population of size N, and suppose
that P individuals in the population have disease at a given time. The prevalence
proportion is P/N. For example, suppose that among 10,000 women residents of
a town on July 1, 2001, 1200 have hypertension. The prevalence proportion of
hypertension among women in that town on that date is 1200/10,000 0.12,
54 EPI DEM IOLOGY
Figure 4–5 Epidemic curve for an outbreak of hysteria among elementary school
children on November 6, 1985.
or 12%. This prevalence applies only to a single point in time, July 1, 2001.
Prevalence can change with time as the factors that affect prevalence change.
What factors affect prevalence? Clearly, disease occurrence affects prevalence.
The greater the incidence of disease, the more people there are who have it.
Prevalence is also related to the length of time that a person has disease. The
longer the duration of disease, the higher the prevalence. Diseases with short
duration may have a low prevalence even if the incidence rate is high. One rea-
son is that if the disease is benign, there may be a rapid recovery. For example,
the prevalence of upper respiratory infection may be low despite a high incidence,
because after a brief period, most people recover from the infection and are no
longer in the disease state. Duration may also be short for a grave disease that
leads to rapid death. The prevalence of aortic hemorrhage would be low even
with a high incidence because it usually leads to death within minutes. The low
prevalence means that, at any given moment, only an extremely small propor-
tion of people are suffering from an aortic hemorrhage. Some diseases have a
short duration because either recovery or death ensues promptly; appendicitis is
an example. Other diseases have a long duration because, although a person can-
not recover from them, they are compatible with a long survival time (although
survival is often shorter than it would be without the disease). Diabetes, Crohn’s
disease, multiple sclerosis, parkinsonism, and glaucoma are examples.
Because prevalence reflects both incidence rate and disease duration, it is not
as useful as incidence alone for studying the causes of disease. It is extremely
Measuring Disease Occurrence and Causal Effects 55
useful, however, for measuring the disease burden on a population, especially if

those who have disease require specific medical attention. For example, the preva-
lent number of people in a population with end-stage renal disease predicts the
need in that population for dialysis facilities.
In a steady state, which is the situation in which incidence rates and disease
duration are stable over time, the prevalence proportion, P, has the following
relation to the incidence rate:
P
= ID [4–2]
1− P
–
In Equation 4–2, I is the incidence rate and D is the average duration of disease.
The quantity P/(1 P) is known as the prevalence odds. In general, when a pro-
portion, such as prevalence proportion, is divided by 1 minus the proportion, the
resulting ratio is referred to as the odds for that proportion. If a horse is a 3-to-1
favorite at a racetrack, it means that the horse is thought to have a probability of
winning of 0.75. The odds of the horse winning is 0.75/(1 0.75) 3, usually
described as 3 to 1. Similarly, if a prevalence proportion is 0.75, the prevalence
odds would be 3, and a prevalence of 0.20 would correspond to a prevalence
odds of 0.20/(1 0.20) 0.25. For small prevalences, the value of the prevalence
proportion and that of the prevalence odds are close because the denominator of
the odds expression is close to 1. For small prevalences (eg, 0.1), we can rewrite
Equation 4–2 as follows:
P IID [4–3]
Equation 4–3 indicates that, given a steady state and a low prevalence, prev-
alence is approximately equal to the product of the incidence rate and the
mean duration of disease. Note that this relation does not hold for age-specific
–
prevalences. In that case, D corresponds to the duration of time spent within that
age category rather than the total duration of time with disease.
As we did earlier for risk and incidence rate, we should check this equation to
make certain that the dimensionality and ranges of both sides of the equation are
satisfied. For dimensionality, the right-hand sides of Equations 4–2 and 4–3 involve
the product of a time measure, disease duration, and an incidence rate, which has
units of reciprocal of time. The product is dimensionless, a pure number. Prevalence
proportion, like risk or incidence proportion, is also dimensionless, which satisfies
the dimensionality requirement for the two equations, 4–2 and 4–3. The range of
incidence rate and that of mean duration of illness is [0,∞], because there is no
upper limit to an incidence rate or the duration of disease. Therefore Equation 4–3
does not satisfy the range requirement, because the prevalence proportion on the
left side of the equation, like any proportion, has a range of [0,1]. For this rea-
son, Equation 4–3 is applicable only for small values of prevalence. The measure
of prevalence odds in Equation 4–2, however, has a range of [0,∞], and it is appli-
cable for all values, rather than just for small values of the prevalence proportion.
We can rewrite Equation 4–2 to solve for the prevalence proportion as follows:
ID
P= [4–4]
1 + ID
56 EPI DEM IOLOGY
Prevalence measures the disease burden in a population. This type of epide-

miologic application relates more to administrative areas of public health than
to causal research. Nevertheless, there are research areas in which prevalence
measures are used more commonly than incidence measures, even to investigate
causes. Examples are birth defects and birth-related phenomena such as birth
weight or preterm birth. We use a prevalence measure when describing the
occurrence of congenital malformations among liveborn infants in terms of the
proportion of these infants who have a malformation. For example, the propor-
tion of infants who are born alive with a defect of the ventricular septum of the
heart is a prevalence. It measures the status of liveborn infants with respect to
the presence or absence of a ventricular septal defect. Measuring the incidence
rate or incidence proportion of ventricular septal defects would require ascer-
tainment of a population of embryos who were at risk for developing the defect
and measurement of the defect’s occurrence among these embryos. Such data
are usually not obtainable, because many pregnancies end before the pregnancy
is detected, and the population of embryos is not readily identified. Even when
a woman knows she is pregnant, if the pregnancy ends early, information about
the pregnancy may never come to the attention of researchers. For these reasons,
incidence measures for birth defects are uncommon. Prevalence at birth is eas-
ier to assess and often is used as a substitute for incidence measures. Although
prevalence measures are easier to obtain, they have a drawback when used for
causal research: Factors that increase prevalence may do so not by increasing
the occurrence of the condition but by increasing the duration of the condition.
For example, a factor associated with the prevalence of ventricular septal defect
at birth could be a cause of ventricular septal defect, but it could also be a fac-
tor that does not cause the defect but instead enables embryos that develop the
defect to survive until birth. On the other hand, there may be practical inter-
est in understanding the factors that are related to being born alive with the
defect.
Prevalence is sometimes used in research to measure diseases that have insidi-
ous onset, such as diabetes or multiple sclerosis. These are conditions for which
it may be difficult to define onset, and it therefore may be necessary in some set-
tings to describe the condition in terms of prevalence rather than incidence.
Prevalence of Characteristics
Because prevalence measures status, it is often used to describe the status of

characteristics or conditions other than disease in a population. For exam-
ple, the proportion of a population that engages in cigarette smoking often
is described as the prevalence of smoking. The proportion of a population
exposed to a given agent is often referred to as the exposure prevalence.
Prevalence can be used to describe the proportion of people in a popu-
lation who have brown eyes, type O blood, or an active driver’s license.
Because epidemiology relates many individual and population characteris-
tics to disease occurrence, it often employs prevalence measures to describe
the frequency of these characteristics.

Prevalence vs. Incidence Rate

Uploaded by

Copyright:

Available Formats

Prevalence vs. Incidence Rate

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Prevalence vs. Incidence Rate

Uploaded by

Copyright:

Available Formats

Measuring Disease Occurrence and Causal Effects 53

already succumbed to an infectious disease may no longer be susceptible to it for

useful, however, for measuring the disease burden on a population, especially if

Prevalence measures the disease burden in a population. This type of epide-

Because prevalence measures status, it is often used to describe the status of

You might also like