The world health report

Chapter 3

Defining and describing risks to health

Risk assessment and management is a political as well as a scientific process, and public perceptions of risk and risk factors involve values and beliefs, as well as power and trust. For policy-makers who are promoting intervention strategies to lower risks to health, it is obviously important, therefore, to understand the different ways in which the general public and health professionals perceive risks (19). As described in Chapter 2, use of the term "risk" has many different meanings and this often causes difficulties in communication. This report uses the notions of the probability of a subsequent adverse health event, followed by its consequence which is mainly either morbidity or mortality.

While many scientists often assume that risks can be objectively verified, many social scientists argue that risk measures are inherently much more subjective. In addition, other members of the public have yet other notions of risk. How do people define and describe risk factors? How do they estimate risks? Answers to such questions obviously alter people's perceptions. Such information is needed, therefore, to improve communications and to predict public responses to public health interventions, including the introduction of new health technologies and risk factor and disease prevention programmes. Box 3.2 illustrates male perceptions of sexual health risks and the need to use preventive measures against HIV infection and pregnancy.

Box 3.2 Men's sexual behaviour related to risk of HIV infection and pregnancy

A greater understanding of men's perceptions of sexual risk and their risk-taking behaviour is necessary if interventions are to be more successful in improving the reproductive health of both men and women. In a question-naire survey of reproductive risk behaviours in the capital cities of Argentina, Bolivia, Cuba and Peru, young adult males (aged 20--29 years) were asked whether they would take measures to prevent HIV infection and pregnancy during sexual intercourse with different categories of female partners. Samples of 750--850 men were selected randomly in each city. The percentages who reported having taken preventive measures -- usually the use of condoms -- to reduce the risk of HIV transmission or pregnancy are shown below.

The findings were very similar in all four cities, though the men clearly perceived the risks as being different with different partners. Preventive measures against HIV infection were believed to be highly necessary for sexual intercourse with prostitutes, strangers and lovers, but considerably less so with married partners. However, just over half the young men said they would use such measures when having intercourse with a virgin or a fiance. The need for measures to prevent pregnancy was perceived, however, to be higher than that for HIV infection. To avoid pregnancy, such measures were commonly used with all sexual partners and even with about half the spouses.

Source: (20)

A complicated question is how the mortality outcome associated with a particular risk factor should be expressed. Even choosing or framing the end-point as death is surprisingly complex and can make large differences in the way risk is both perceived and evaluated. The following is a well-known example from occupational health, which shows how the choice of risk measure can make a technology appear less or more risky to health (21). Between 1950 and 1970, coal mining in the USA became much less risky if the measure of risk was taken to be accident deaths per million tons of coal produced, but it became more risky if risk was described in terms of accident deaths per 1000 miners employed. Which measure is more appropriate for decision-making? From a national perspective, and given the need to produce coal, deaths of miners per million tons of coal produced appears to be the more appropriate measure of risk. However, from the point of view of individual miners and their trade unions the death rate per thousand miners employed is obviously far more relevant. Since both measures for framing the risks in this industry are relevant, both should be considered in any risk management decision-making process.

Each way of summarizing deaths embodies its own set of inherent and subjective values (7). For example, an estimate based on reduction in life expectancy treats deaths of young people as more important than deaths of older people, who have less life expectancy to lose. However, counting all fatalities together treats all deaths of the young and old as equivalent. This approach also treats equally deaths immediately after mishaps and deaths that follow painful and lengthy debilitating diseases. Such choices all involve subjective value judgements. For instance, using "number of deaths" may not distinguish deaths of people who engage in an activity by choice and benefit from it directly, from those of people who are exposed to a hazard involuntarily and who get no direct benefits. Each approach may be justifiable but uses value judgements about which deaths are considered to be the most undesirable. To overcome such problems, information should be framed in a variety of different ways so that such complexities are revealed to decision-makers.