Case Control Studies

What are case-control studies?

Case control studies are widely used in epidemiology. Their retrospective design (looking backward in time, as opposed to prospective studies which look forward) enable researchers to study factors associated with diseases which, because of their rarity, would be difficult to study prospectively, such as in cohort studies or RCTs. In the case of SIDS the costs of prospectively studying enough babies to be able to obtain a useful sample effectively prohibits such studies (simplistically, if the SIDS rate was 1 per 3000, you would have to follow 150,000 babies to obtain a still relatively small sample of 50 SIDS cases). Instead the case-control design enables researchers to compare two groups: the ‘cases’, who already have a disease, with ‘controls’, who do not have that disease. By working backwards from the incidence of the disease (or not) researchers can investigate risk factors which may be associated with it.

A well-used example is the association between smoking and lung cancer: The association can be seen by comparing ‘cases’ (who have lung cancer) with matched ‘controls’ (who do not have lung cancer). Within each group there will be non-smokers and smokers. By comparing the incidence of smoking in the ‘cases’ group with that in the ‘control’ group researchers can begin to identify whether an association exists, in which ‘direction’ that association lies, and - to a degree - its magnitude.

Much work investigating SIDS risk factors has utilised the case-control methodology. This has enabled researchers to compare ‘cases’ (i.e. babies who have died from SIDS) with ‘controls’ (age-, and sometimes socioeconomically-, matched babies who have not died from SIDS). By looking backward at factors to which babies were or were not exposed (e.g. sleeping on back vs on front, formula-feeding, smoking), along with individual characteristics (e.g. sex, age, health-status), researchers can identify associations between exposure and incidence.

In the next sections we look at the statistics associated with case control studies, and how they are interpreted.