Changes

=='''Causation'''==

=='''Causation'''==

The ultimate aim of most analytic studies is to demonstrate evidence of a causative association between a factor of interest and a disease. Definitively establishing a causal connection between an exposure and disease is very difficult - indeed, it is considered to be conceptually impossible by philosophers<ref>Rothman, K.J. 'Causation and Causal Inference in Epidemiology', American Journal of Public Health, 2005, Vol 95, No. S1 [http://ajph.aphapublications.org/cgi/reprint/95/S1/S144]</ref>. , Additionally, the complex interplay between different host, agent,and environmental factors, as well as many diseases being multifactorial in nature makes the identification and interpretation of possible 'causative' factors difficult. In order to assist the investigation of causation, risk factors for disease ('''component causes''') may be classified as '''necessary''' and '''sufficient''' causes. A necessary cause is one which is required for disease to occur - i.e. it is present in every case of disease (this is commonly the case with transmissible diseases, where a pathogenic agent is often identified as the necessary cause). A sufficient cause is a combination of component causes which would result in disease.

The ultimate aim of most analytic studies is to demonstrate evidence of a causative association between a factor of interest and a disease. Definitively establishing a causal connection between an exposure and disease is very difficult - indeed, it is considered to be conceptually impossible by philosophers<ref>Rothman, K.J. 'Causation and Causal Inference in Epidemiology', American Journal of Public Health, 2005, Vol 95, No. S1 [http://ajph.aphapublications.org/cgi/reprint/95/S1/S144]</ref>. , Additionally, the complex interplay between different host, agent,and environmental factors, as well as many diseases being multifactorial in nature makes the identification and interpretation of possible 'causative' factors difficult. In order to assist the investigation of causation, risk factors for disease ('''component causes''') may be classified as '''necessary''' and '''sufficient''' causes. A necessary cause is one which is required for disease to occur - i.e. it is present in every case of disease (this is commonly the case with transmissible diseases, where a pathogenic agent is often identified as the necessary cause). A sufficient cause is a combination of component causes which would result in disease, even if the individual components alone will not, and may or may not include necessary causes.

=='''Analytic studies'''==

=='''Analytic studies'''==

Analytic studies may compare exposure to these factors in animals with disease to those without, or may compare the occurrence of disease amongst 'exposed' animals to 'unexposed' animals. Although there are considerable difficulties in establishing causation of disease, the results of these studies can provide useful evidence regarding possible risk factors for disease. Although it is important to carefully consider the source population when undertaking these studies, the results may be able to be extrapolated out to a wider population than for descriptive studies. One example is that of the association between smoking and lung cancer - although the prevalence of lung cancer in the UK is likely to be very different from that in Mongolia, the association between smoking and development of lung cancer in both of these countries is likely to be similar.  

Analytic studies investigate possible associations between risk factors and disease by either comparing the risk factor exposure status in animals with disease to those without, or by comparing the occurrence of disease amongst 'exposed' animals to 'unexposed' animals. Although there are considerable difficulties in establishing causation of disease, the results of these studies can provide useful evidence regarding possible risk factors for disease. Although it is important to carefully consider the source population when undertaking these studies, the results may be able to be extrapolated out to a wider population than for descriptive studies. One example is that of the association between smoking and lung cancer - although a descriptive study of the prevalence of swine influenza amongst pigs in the UK is unlikely to be of much use in estimating the prevalence amongst pigs in the USA, an analytic study investigating risk factors for infection may (or may not) be similar. When considering whether or not results can be extrapolated out to a larger 'target' population, host, agent and environmental factors (such as breeds of pigs, strains of virus, farming systems and wildlife exposure) should be considered.

==References==

==References==