Epidemiological Study Design

Study Notes

Study design encompasses both non-intervention (observational) and intervention (experimental) studies.
Non-intervention studies involve describing or analyzing research objects without intervention.
Intervention studies involve manipulating objects or situations through preventive or therapeutic measures to determine the outcome.

Studies can be categorized as observational (non-experimental, population or individual-based) or experimental (interventional).
Observational studies, which can be further classified into descriptive and analytic, evaluate relationships without intervention.
Experimental studies investigate interventions and their impact. These can be further subdivided.

Descriptive analytic, studies can include health surveys, ecological studies.
Cross-sectional is a study design that involves collecting data at one point in time. Prevalence studies analyze the proportion of a population with a certain condition at a specific time.
Analytic, studies evaluate relationships between variables; this includes case reports, case series, case-control studies, and cohort studies.
Experimental studies include randomized controlled trials (e.g., clinical trials), quasi-experimental studies (such as field trials and community trials), and non-randomized controlled trials.

Randomization is critical for minimizing bias; this includes parallel design, cross-over design, split-mouth design, pretest and post-test design, external control.
Control groups are necessary. Controls can include positive controls, negative controls, and placebos.
Blinding helps to reduce bias. Blinding concepts include open-label, single-blinded, double-blinded, and triple-blinded.
Sample size should be adequate to ensure that any observed differences between groups are due to the intervention.
Follow-up period must be sufficient to detect new diseases or lesion extensions arising during the trial. This often is dependent on the specific condition being studied. For example, caries studies might take years while a short-term study on plaque inhibiting agents would only need days or short time periods.
Inter and Intra-examiner reliability tests ensure consistent measurements across different observers or the same observer over time.

The hierarchy ranks types of evidence from strongest to weakest.
Meta-analyses and systematic reviews are considered strongest.
Randomized controlled trials (RCTs) follow next, then cohort studies, case-control studies, and so on.
Case reports and opinions are lowest in the hierarchy.

Databases: are online resources for medical and healthcare research; popular databases include PubMed, Cochrane Library, PubMed Central, ScienceDirect, Scopus, Web of Science, EBSCO, ResearchGate, and Google Scholar.
Publishers: medical and healthcare publications come from organizations or individuals who create and publish the information; popular publishers include Springer, Elsevier, Thieme.