Evidence-based dentistry: How research shapes better oral health decisions for patients and clinicians
Cohort studies and longitudinal ecological studies measure one group of individuals (the cohort) over time and record information multiple times to determine risks. These studies offer a long-term look at a person’s health and observe the incidence of a disease. By tracking individuals over time, factors that contribute to the development of caries, periodontal disease, or oral cancer, can be pinpointed. These include socioeconomic status, dietary habits, or early childhood experiences.
Experimental studies
Experimental studies, also known as clinical trials, use independent and dependent variables to provide the highest level of evidence of all the study designs.4 These studies are the highest level of scientific evidence and demonstrate cause and effect.3
Independent variables represent the factor being manipulated by the researchers. The dependent variables depend on changes in the independent variable. If there’s a study to determine whether using different polishing agents to see which one is most effective at removing biofilm, the independent variable being manipulated would be the polishing agent and the dependent variable would be the biofilm. In these studies, the researcher wants to find new effects of treatment in a controlled setting like a lab or clinic.
Key sampling techniques
The purpose of sampling is to allow the researcher to observe the entire population and add more weight to findings.5 Due to limited resources, time, and money, it’s not always feasible to obtain answers from all participants. The samples represent the total population so that the data collected from the samples will be as accurate as that population.
A random sample is formulated so that each person in the population has the same chance of being included.5 An example is a group of patients use a different dentifrice to treat dental hypersensitivity and to follow up with the first 20 who return for an interview about the product. This is not predictable and there is no bias.
Stratified random sampling is dividing a population into subgroups or strata and then randomly selecting from within.6 An example is dividing a large population into women and men and then choosing three from each group. Systematic random sampling involves selecting every Nth member. A simple example is if coworkers are lined up and every fourth person is chosen to create a population.4,6
Conclusion
Valid research is the foundation for understanding why disease occurs and how to prevent and control diseases. Being able to critically think can help clinicians predict which outcomes a patient can expect following procedures, choose the best products, and help patients navigate all the information available on the internet, all while guiding efforts to advance the profession.
link
