Epidemiology/Introduction

SESSION OBJECTIVE

At the end of this session you should know the uses of absolute and relative measures of effect in the world of health planning.

Relative and absolute comparisons measure different aspects of the association between an exposure and a health outcome.

Relative comparisons of the measure of occurrence in exposed versus the unexposed group show the strength of the association and provide some idea of the aetiological role of the exposure for that health outcome.

Example 1:

The incidence rate for lung cancer in those exposed to tobacco smoking is 12 cases per 100 000 population per year. That for lung cancer in non-smokers is 1 case per 100 000. The incidence rate ratio of 12 is very high. Most associations with suspected aetiologic agents are between 1 and 3, and often less than 2. This relative risk would hold true irrespective of the size of the proportion of the population that smokes.

The statistical properties of relative risks are easier to work with and these measures of effect are preferred by epidemiologists and biostatisticians in the main.

Absolute comparisons measure the social and healthcare (often referred to as the Public Health) impact of the exposure.

Example 2:

If the proportion of smokers is very low, as is the case for instance in the USA (around 15%), then the social and health care implications of the effect of tobacco in lung cancer would be very different compared to a Southern European country where as many as 60 or 70 % of the population are smokers. One way of measuring these different implications involves making an absolute comparison viz. the incidence rate difference is 11 per 100 000 smokers per year. Depending upon the numbers of smokers in the population one can work out the number of cancers prevented and the likely impact of a successful smoke cessation programme on the health services. 92% of cases could be prevented in smokers.

Example 3:

A more frequently cited example is the relationship between smoking and ischaemic heart disease. The relative risk is only 1.2 implying a weak but nevertheless significant effect of smoking on heart disease. To be precise it elevates the measure of occurrence by 20%. However, because so many people smoke in society and because ischaemic heart disease is so common even a 20% rise in incidence or prevalence is of major significance. Hence the absolute comparison is often used for public health purposes to obtain numbers of say heart attacks prevented in order to relate this to health services planning objectives.