The time required for an epidemiologic study is dependent on the study design, the study population, and the frequency of both exposure and health condition under investigation plus the time needed for data processing and analysis and manuscript preparation.
In a case control study, much of the study time is at least the time needed to recruit enough cases and controls as determined by power calculation. It is best to use incident cases and if these are identified prospectively, this time frame depends on the incidence of health condition in the source population from which the cases and controls are drawn. Some study settings allow for the rapid identification of sufficient incident cases that were first diagnosed in the past and suitable controls If the condition under investigation is rare, the time required to recruit enough cases prospectively will be longer than if the condition is common. Also, for a rare condition, recruiting cases from a high-risk population will shorten the study time. Some diseases are so rare that the investigator must include prevalent as well as incident cases.
In a prospective cohort study, the study time is partly determined by 1) the incidence of the health condition in both the exposed and unexposed groups, and 2) the difference in incidence between the two groups. In general, cohort study is designed to investigate the relationship between a common exposure and an outcome; it usually takes a long period of time (usually years) to observe a significant difference in incidence between the exposed and unexposed groups. However, if the exposure has a large impact on the incidence of the health outcome or the disease is common, the follow-up time may be shortened.
Retrospective cohort studies, in which a cohort is identified in the past and disease occurrence is ascertained up to the present, can be accomplished relatively rapidly using existing records.The study time of a randomized controlled trial can also range from weeks to years. Of note, if the intervention already shows significant effect on the study group before the planned end time of study, the investigator may choose to terminate the study early.
The lengths of similar randomized controlled trials for the evaluation of the effect of H. pylori eradication on gastric cancer risk may vary. If the primary outcome is gastric cancer incidence, as in the trial conducted by Wong et al (JAMA. 2004 Jan 14;291(2):187-94), several years or decades may be required to observe a meaningful difference in gastric incidence between the intervention and the placebo groups. However, if the primary outcome of interest is progression/regression of gastric premalignant lesion (an intermediate endpoint of cancer), the study time may be shortened to one year, as described by Ley et al (Cancer Epidemiol Biomarkers Prev. 2004 Jan;13(1):4-10).