Define Randomized Clinical Trial Study
Study in which participants are randomly (i.e. by chance) assigned to one or two or more treatment arms of clinical trial. Randomization removes investigator bias guarantees statistical test will have valid significance levels. Comparison groups can have no intervention, observation, placebo, or usual care. Blinding/Masking removes further bias.
Pros: best level of evidence to connect exposure/outcome, most controlled
Cons: expensive, length of time to conduct
Define Cohort Study
Start with exposure and look forwad in time from point of exposure to determine outcomes of interest then compares incidence of outcome between exposed & unexposed groups.
Pros: approximates RCT design (exposure before outcome), evaluate multiple outcomes, provide actual measure of risk of outcome of interest, incidence and relative risk are extractable
Cons: Potential for loss to follow-up, requires large sample size, not inteded for study of rare outcomes, takes a long time (not efficient for long-developing outcomes), expensive
Define Case Control Study
Start with the outcome and look backwards in time to determine common exposure for sample.
Pros: study of rare diseases or disease with long latency, requires relatively few subjects, less time than cohort study, allows evaluation of multiple exposures as potential cause of disease, well-suited for outbreak investigations
Cons: recall bias (relies on recall of subjects for information on past exposure), selection of appropriate control group, yields odds ratio which is only estimate of relative risk, cannot calculate incidence rates
Define Cross-sectional Study
AKA prevalence study. Looks at sample population at single point in time then determines who was exposed & unexposed and who had the disease/remained disease free.
Pros: good measure of disease prevalence, info on clinical setting expecations, used in evaluation screening and diagnostic tests, planning health services, quick, easy, and inexpensive to carry out (only contacting sample once)
Cons: Outcome and exposure status measured at same time, temporal relationships between exposure and disease difficult to assess (cannot determine causality), limited to study of prevalence (no incidence or duration)
Define Ecological Study
Compares disease rates between populations given some sort of exposure. Data is based on the group/population not the individual.
Pros: Clues to etiological hypotheses, help set research priorities, study large populations at low cost, address questions that might be difficult using other epidemiologic approaches
Cons: No data on individual study subjects, "ecologic fallacy" - innappropriate inference from ecologic data, when true relationship from group data does not mean a true relationship at individual level
Define Case report/series Study
Observational, detailed description of clinical presentation of a patient(s). The report provides descriptive statistics only wrt person, place, time.
Pros: great detail on clinical features of a disease or symptoms and is the best way to quickly report a new observation
Cons: usually small, highly selected patient group, no comparison group, not hypothesis based
Define, calculate and interpret different ways to express absolute risk
Probability of an event in a population under study. Used interchangeably with incidence.
Ratio of # new cases over given period of time/# people in the group
Define, calculate and interpret different ways to express risk difference
AKA attributable risk. Difference between risk of disease in exposed and unexposed groups. It is the incidence due to exposure.
AR = Incidence(exp) - Incidence(unexp)
Define, calculate and interpret different ways to express relative risk
The likelihood of exposed person getting the disease relative to a non-exposed person.
Ratio of Incidence rate(exp) to Incidence rate(unexp)
0 < RR < 1 -- Protective effect to exposed group
RR = 1 -- Equal risk to exposed/unexposed groups
RR > 1 -- Exposed group is at greater risk (risk factor) than unexposed
Define, calculate and interpret different ways to express population-attributable risk
Measure excess incidence of disease in a community that is associated with a risk factor.
Pop-AR = AR x Prevalence of exposure
Define, calculate and interpret different ways to express odds ratio
The odds that a case is exposed divided by the odds that a control is exposed. It is an estimate of relative risk.
To calc take cross product of 2x2 table
OR = ad/bc
Define Prevalence ratio
Disease Prevalence(exp)/Disease Prevalence(unexp)
Compare and contrast study designs
Rank research study types based on quality: