The need for health data in integrated assessments

In order to carry out integrated assessment of environment and health risks, homogeneous health data are required across the study area. These data are used primarily to provide background rates which feed into the assessment of health impacts, though they can also be important as a basis for testing the plausibility of the estimated effects (e.g. to ensure that they are in line with observed rates of morbidity or mortality) and for the purpose of piloting (e.g. to provide a basis for approximating health effects).   

In all these cases data are needed for each of the assessment scenarios.  The data also need to be consistent in all key respects, i.e.:

• in terms of classification of diseases;
• spatially (e.g. between regions or countries);
• in terms of population structure (male/female, age group);
• temporally (i.e. in terms of reporting periods). 

Types of health data and their sources

The two main types of health data are used in impact assessments: on mortality (number of deaths) and morbidity (incidence or prevalence of diseases).  These are generally available from routone datasets (i.e. data gathered for administrative purposes).

Mortality is usually recorded in registers at local and national levels. Thanks to the International Classification of Diseases (ICD) the cause of death is clearly known and consistently identified (within the inevitable constraints of individual diagnosis).  Most European countries are currently using the ICD-10 standard (the most recent standard) which facilitates comparisons. Depending on the geographical level mortality data are available per year or by period.

Morbidity is based on hospital admissions, in-patient care and medication, and as such is more relevant when assessing the incidence or prevalence of a disease. As with mortality data, morbidity records are available for different periods of time. Morbidity data are derived from a number of sources, including routine reporting by hospitals and GPs, as well as specialist registers maintained to monitor specific diseases (e.g. cancers, congenital abnormalities, communicable diseases).  Medication data may also be obtained as sales data from pharmacies or pharmaceutical companies.  Compared to mortality data, inconsistencies tend to be much greater both because of uncertainties in diagnosis and differences in the reporting systems and their efficiency.  

Many other, more ad hoc, sources of health data are also available, including data gathered for research (or more generally for public health) purposes through longitudinal surveys and household surveys.