Epidemiological Measurements, Measures of Effect

Epidemiological measurements

Basically, there are three important epidemiological measures of disease:

1. Disease frequency measures: These are characterized by the occurrence of disease, disability or death in human population and therefore are fundamental to descriptive and etiologic investigations. The frequency of occurrence of outcome under consideration can be measured in terms of prevalence and incidence.
2. Measures of disease association: These measures assess the strength of the statistical relationship between a given factor and disease (outcome or end point of the interest).
3. Impact assessment or measure of potential impact: These reflect the expected contribution of a study factor to the frequency of a disease in a particular population.These measures are useful for predicting the efficacy or effectiveness of the therapeutic maneuvers and intervention strategies within a specific population e.g. vaccine. Essentially, potential impact measures are a combination of the frequency and the association measures.

Rates, Ratios, Proportions and Indices

The general methods or tools used for describing disease pattern are measurements of mortality and morbidity. The morbidity data analysis and interpretation is quite difficult than mortality data because there may be episodes of illnesses in a given period than the number of deaths during a given time interval. The number of deaths or illnesses in different places, populations or periods of time cannot describe the pattern of disease. Therefore, morbidity rates, ratios, proportions or indices are required for the proper and acceptable descriptions of disease patterns.

Rate

A rate measures the probability of occurrence of some particular event. A true rate is an instantaneous change in one quantity per unit change in another quantity, where the latter is usually a time. In epidemiologic study, it is not possible to measure the true instantaneous rate disease in a population. A rate is of the form:

Rate = a / a + b * k

where a=number of times an event has occurred in a specific interval of time, a + b= number of persons exposed to risk of the event during the same interval ( a is a portion of a + b), and k= some round number (100 or 1000 etc.) or base, depending upon the relative magnitude of a and a + b.

Characteristics of rate

1. It has aunit of measurement (with dimension) i.e. time.
2. It has no finite bound. Theoretically, a rate can approach infinity.

Ratio

A ratio is the expression of the relationship between a numerator and denominator, which may involve either an interval in time or may be theinstantaneous change in time. In ratio, numerator is not included in the denominator thereby distinguishing it from proportion. The ratio is in the form of:

Ratio = / y

Where, x = number of events or items counted and not necessarily a portion of y, y = number of events or items counted

There are two kinds of ratio:

1. One that has dimension (unit of measurement) e.g. number of hospital beds per 100,000 persons in population, number of infant deaths in a population during one year per 1000 live birth.
2. One that is dimensionless (has no unit of measurement) e.g. by dividing one proportion or rate to another.

Proportion

A proportion is an expression in which the numerator is always included in the denominator and the base is equal to 100. Therefore, a proportion is always expressed as a percentage i.e. the decimal representation of the population multiplied by 100.

Characteristics of proportion

1. It is dimensionless (no unit of measurement). Since numerator and denominator have the same dimension, any dimensional contents are cancelled out.
2. The value ranges between 0 and 1 (0.0 ≤ p ≥ 0.1)

Index

An index is the best available approximation to a true rate. This usually occurs when we unable to count directly the number at risk (denominator) and then forced to use something else which can count to give us an approximation of the number at risk.

Thus, for example, in order to calculate the maternal mortality ratio, we should, strictly speaking, divide the number of deaths, assigned to deliveries and complications of pregnancy, childbirth and the peuperium by the total number of pregnancies occurring in a specified time interval. Since, we are unable to count the total number of pregnancies; we replace this by the number of registered live births in order to obtain an estimate of the number of pregnancies. The result of maternal mortality rate is therefore not a rate in the strict sense, but it is nevertheless a useful index, which gives an impression of the risk of dying associated with pregnancy.

Uses of Ratios, Proportions and Rates

Population characteristics such as age, sex, race, exposures and other variables are explained using rate, ratio, and proportion. Rate, ratio and proportion are used to describe morbidity, mortality and natality.

Measures of Effect

An effect is the difference in disease occurrence between two groups of people who differ with respect to causal characteristics, often referred to as exposure. Simply, the effect is the result of a cause (exposure or factor). It is the end-point of a causal mechanism that characterizes the outcome, a certain creates. An effect may be an absolute effect or relative effect.

Exposure is defined as the characteristics that may have a potential role in a causal mechanism. Biological trait (age, sex, BMI), behavior (health seeking, risk behavior), treatment (experimental, standard therapy), procedure (laboratory method) are some examples of common exposure.

Absolute effect measures

Absolute effect measures give the difference between two attributes. Attributable risk, population attributable risk, and risk difference are widely used measures.

Attributable risk

Attributable risk is defined as the proportion of disease among exposed individuals that can be attributed to the exposure. It is measured by subtracting the risk of disease among unexposed from the risk among the exposed individuals. The attributable risk assumes that causes other than the one under investigation have had equal effects on the exposed and unexposed groups.

Attributable risk = incidence in exposed group-incidence in the unexposed group / incidence in the exposed group

AR = I_e- I_u / I_e

It is assumed that the incidence rate of the outcome among the exposed would be reduced if the exposure were eliminated.

Population attributable risk

With a given outcome, exposure factor and population, the attributable risk among the population is the proportion by which the incidence rate of the outcome in the entire population would be reduced if exposure were eliminated.

Population

attributable risk = incidence in the total population-incidence in the unexposed group / incidence in the total popualtion

AR (Pop) = I_p-I_u/ I_p

Relative effect measures

Relative effect measure is the ratio between two attributes. Risk ratio, relative risk ratio, incidence rate ratio (IRR), prevalence ratio are important relative effect measures.

References

Atlas, RM and R Bartha. Microbial Ecology:Fundamentals and Applications. The Benjamin Cummins Publication co. Inc., 1998.

Gordis, L. Epidemiology. third edition. 2004.

Maier, RM, IL Pepper and CP Gerba. Environmental Microbiology. Academic press Elsevier Publication, 2006.

park, K. Park's Text Book of social and prevention Medicine. 18th edition. 2008.