Intro
Detailed descriptions of the level and distribution of diseases and injuries and their causes are important inputs into strategies for improving population health. A substantial body of work has focused on quantifying causes of mortality and, more recently, the burden of disease (Murray and Lopez 1997; Preston, 1976; see also chapter 3 in this volume). Data on disease or injury outcomes alone, such as death or hospitalization, tend to focus on the need for curative or palliative services. Reliable and comparable analyses of risks to health are critical for preventing disease and injury. Investigators have frequently analyzed morbidity and mortality due to risk factors in the context of methodological traditions of individual risk factors and for selected populations (Kunzli and others 2000; Leigh and others 1999; McGinnis and Foege 1993; Peto and others 1992; Single and others 1999; Smith 2000; Smith, Corvalan, and Kjellstrom 1999; Willet 2002). As a result, most estimates have been affected by the following shortcomings, which limit comparability:
-
The causal attribution of morbidity and mortality due to risk factors has been estimated relative to arbitrary exposure levels without standardizing baseline exposure across risk factors. For example, the implicit baseline for the burden of injuries attributable to occupational factors has been "no work," because estimates have been based on occupational registries intended to register all injuries, regardless of whether they are avoidable (Leigh and others 1999).
-
The intermediate stages and interactions in the causal process have not been considered when calculating the disease burden attributable to risk factors. As a result, attributable burden could only be calculated for those risk factor and disease combinations for which epidemiological studies had been conducted.
-
The outcomes of analyses have been morbidity or mortality from specific diseases without conversion to a standard unit, making comparisons among different diseases and/or risk factors difficult.
To permit the assessment of risk factors in a unified framework while acknowledging characteristics specific to individual risk factors, the Comparative Risk Assessment (CRA) project initiated a systematic evaluation of the changes in population health that would result from modifying the population distribution of exposure to a risk factor or to a group of risk factors (Murray and others 2003; Murray and Lopez 1999; Ezzati and others 2004). In particular, the CRA framework
-
compares the burden of disease due to the observed distribution of exposure in a population with the burden from an alternative distribution consistently defined across risk factors;
-
considers multiple stages in the causal network of multiple risk factors and disease outcomes to allow inferences about combinations of risk factors for which epidemiological studies have not been conducted, including the joint effects of changes in multiple risk factors;
-
converts the burden of disease and injury into a summary measure of population health that permits comparing fatal and nonfatal outcomes while also taking severity and duration into account (the summary measure used in this chapter is the disability-adjusted life year [DALY], whose definition and calculation are described in chapter 3).
Therefore, even though CRA is similar to other risk assessment exercises in the sense that it applies knowledge about the hazardous effects of risk factors from epidemiological research to data on exposure in the broader population, it creates conceptual and methodological consistency in measuring the impacts of various risk factors on population health. Furthermore, we have attempted to use consistent and comparable criteria for evaluating the scientific evidence on prevalence, causality, and magnitude of hazardous effects across risk factors. As a result, the unified framework for describing population exposure to risk factors and their consequences for population health is an important step in linking the growing interest in the causal determinants of health across a variety of disciplines from natural, physical, and medical sciences to the social sciences and humanities.
We note that risk assessment as defined here is distinct from intervention analysis, whose purpose is to estimate the benefits of a given intervention or group of interventions in a specific population at a particular time. Rather, risk assessment aims at mapping alternative population health scenarios that arise from changes in the distribution of exposure to risk factors, irrespective of whether exposure change is achievable using existing interventions. The alternative visions of population health in turn contribute to identifying those risk factors for which effective or cost-effective interventions should be implemented or new interventions should be developed.
