Study Design: Observational Studies

Abstract

Observational study designs are those where the investigator/researcher just observes and does not carry out any intervention(s)/actions to alter the outcome. The three most common types of observational studies are cross-sectional, case control and cohort (or longitudinal). In cross-sectional studies, both the exposure/risk factor(s) and the outcome(s) are determined at a single time point. They can provide information on prevalence of a condition and snapshot of probable associations that can be used to generate hypothesis. Case-control studies are where subjects are selected based on presence/absence of outcome and the risk factors are determined during the study after enrolment of study subjects. The association between exposure and outcome is reported as odds ratio. These studies; however, have high risk of bias, which must be taken care of during study design. Cohort studies are prospective in nature, where subjects are selected based on presence/absence of exposure, and the outcome(s) is determined at the end of study. These studies can provide incidence of disease/outcome and the association between exposure and outcome is reported as relative risk. They are useful to ascertain causality. High dropouts of study participants and confounding can be problems encountered in these studies.