In epidemiology, researchers are interested in measuring or assessing the relationship of exposure with a disease or an outcome. As a first step, they define the hypothesis based on the research question and then decide which study design will be best suitable to answer that question. How the investigation is conducted by the researcher is directed by the chosen study design. The study designs can be broadly classified as experimental or observational based on the approach used to assess whether exposure and an outcome are associated. In an experimental study design, researchers assign patients to intervention and control/comparison groups in an attempt to isolate the effects of the intervention. Being able to control various aspects of the experimental study design enables the researchers to identify causal links between interventions and outcomes of interest. In several instances, an experimental study design may not be feasible or suitable; in such situations, observational studies are conducted. Observational studies, as the name indicates, involve merely observing the patients in a non-controlled environment without actually interfering or manipulating with other aspects of the study and therefore are non-experimental. The observation can be prospective, retrospective or current depending on the subtype of an observational study.
Case-control studies are used to determine the degree of associations between various risk factors and outcomes. The factors that affect the risk of a disease are called exposures. Case-control studies can help identify beneficial or harmful exposures. In a case-control study, as the name suggests, there are two groups of patients-cases and controls. Cases are patients who have a particular disease, condition, or disability. Controls are those patients that do not have the disease. Typically, researchers identify appropriate representative controls for the cases that they are studying from the general population. Then they retrospectively look in the past for the possible exposures these patients might have had to a risk factor. Selecting the patients for the control group is a very critical component of research based on case-control studies. Due to the retrospective nature of the study design, case-control studies are subject to recall bias. Case-control studies are inexpensive, efficient, and often less time consuming to conduct. This study design is especially suitable for rare diseases that have longer latency periods.
Case-crossover studies are helpful to study triggers within an individual. When the researcher is studying a transient exposure or risk factor, the case-crossover design is useful. This is a relatively new study design where there is a case and a control component both of which come from the same individual. Each case is self-matched by serving as its own control. Determining the period of the control and case components is a critical and difficult aspect of a case-crossover study.
Cohort studies initially classify patients into two groups based on their exposure status. Cohorts are followed over time to see who develops the disease in the exposed and non-exposed groups. Cohort studies can be retrospective or prospective. Incidence can be directly calculated from a cohort study as you begin with exposed and unexposed patients, unlike a case-control study where you start with diseased and non-diseased patients. Relative risk is the measure of effect for a cohort study. Cohort studies are subject to very low recall bias, and multiple outcomes can be studied simultaneously. One of the disadvantages of cohort studies is that they are more prone to selection bias. Studying rare diseases and outcomes that have long follow-up periods can be very expensive and time-consuming using cohort studies.
Cross-sectional studies are observational in nature and give a snapshot of the characteristics of study subjects in a single point of time. Unlike cohort studies, cross-sectional studies do not have a follow-up period and therefore are relatively simple to conduct. As the exposure status and outcome of interest information is collected in a single moment in time often by surveys, cross-sectional study design cannot provide cause-effect relationship and is the weakest of the observational designs. This study design is generally used to assess the prevalence of a disease in a population.
Ecological studies are used when data at an individual level is unavailable or when large-scale comparisons are needed to study the population-level effect of exposures on a disease condition. Therefore, ecological study results are applicable only at the population level. The types of measures in ecological studies are aggregates of individual-level data. These studies, therefore, are subject to a type of confounding called ecological fallacy which occurs when relationships identified at group level data are assumed to be true for individuals. Ecological studies are generally used in public health research.
Randomized Clinical Trials
Randomized clinical trials or randomized control trials (RCT) are considered the gold standard of study design. In an RCT the researcher randomly assigns the subjects to a control group and an experimental group. Randomization in RCT avoids confounding and minimizes selection bias. This enables the researcher to have similar experimental and control groups thereby enabling them to isolate the effect of an intervention. The experimental group gets the exposure/treatment which can be an agent involved in causation, prevention or treatment of a disease. The control group receives no treatment, a placebo treatment or another standard of care treatment depending on the objective of the study. The groups are then followed prospectively to see who develops the outcome of interest. RCT’s are expensive, and researchers using this study design often face issues with the integrity of randomization due to refusals, drops outs, crossovers, and non-compliance.
The key function of an epidemiology study design is to enable the researcher to address the research question with minimal ambiguity logically.
Study design should be well thought of before initiating a research investigation. Choosing an inappropriate study design may undermine overall study validity. Critical thinking about the possible study design issues beforehand will ensure that the research question is adequately addressed.
Study design plays a major role in determining the scientific value of a research study. Understanding the basic study design concepts will aid the clinicians in practicing evidence-based medicine.
Errors in study design are extremely difficult to correct after study completion. Thorough planning is required to avoid weak conclusions or unconvincing results.