Epidemiological or Observational Research
Observational, or epidemiological, evidence is differentiated from other forms of research in that no treatment is given to the study participants. In this kind of study, observations are made between group characteristics and risk for disease; however, because so many unidentified confounding factors are at play in a given population, and because no treatment is being compared to non-treatment, this type of research cannot establish causality between a treatment and disease outcome. Instead, observational research aims at establishing associations, or correlations describing how two health variables tend to coexist. For example, epidemiological studies have provided insight into the correlation of high cancer rates in areas that have done extensive mining or agriculture. Epidemiological studies cannot prove, but may lend support toward the long-term effectiveness, or lack thereof, for different treatment and healthcare strategies, such as specific drugs, antibiotics or vaccinations.
Observational or epidemiological research include: prospective cohort studies, restrospective studies and cross-sectional studies.
- An example of a prospective study is one that would assess blood levels of vitamin D in a given population and then track the incidence of cancer over the next 10 years, potentially establishing an association between vitamin D and cancer.
- An example of a restrospective study is one that would examine a population of patients who already have a given disease and then look back to identify risk factors or exposures that preceded onset of the disease.
- Cross-sectional studies identify the prevalence of a given condition among a group of individuals at one point in time, for example the prevalence of iron deficiency among pregnant women in Canada.
- Can show trends in health care
- Useful in identifying associations that may be further investigated
- Useful for identifying potentially harmful or dangerous associations, for instance between environmental toxins and disease
- Inability to control for unidentified factors which may be affecting the association under investigation.