There are two types of studies, observational studies and clinical trials.
Observational Study
These can be divided in three groups:
- Cross sectional study
- Case control study
- Cohort study
Cross Sectional (Prevalence) Study
A cross sectional study is also called a prevalence study. It is performed on the population that has survived. Therefore, it is prone to selection bias. Cross sectional studies are often performed in epidemiology and can be helpful in trying to establish causes for disease. For example the relation between smoking and lung cancer can be established with a cross sectional study.
Case Control Study
A case control study is usually a retrospective study. Patients are selected by past exposure of a risk factor and compared to a group that has not been exposed to this risk factor. Again, being retrospective, this type of study is prone to selection bias. Advantages of case control studies are that they are cheap, quick and easy to perform. For example, suppose we suspect that smoking adversely affects fracture healing. From our records, we could identify a group of patients with tibial fractures and determine whether fracture healing is prolonged in patients who smoke. It can be difficult selecting an appropriate control group that is similar in all but the risk factor.
An alternative case control study is where we try to match a control to each patient in the study group. The control will be similar in all (age, sex etc) but the past expose to the risk factor. This is called a case matched study. In this type of study, the control group is more comparable to the study group.
Cohort Study
A cohort study is usually prospective in design. A group of patients, who have been exposed to a risk factor, are followed in time and the outcome is observed. Clearly, compared to a case control study, a cohort study is more difficult to do and more time consuming. It also involves considerable administration. Another problem with cohort studies is that there is bias due to patients lost to follow up. An advantage is that the design is prospective and clearly the weight of evidence is greater as compared to a case control study. For example, a cohort study could be performed on patients who had a hip replacement, were the outcome is followed over time.