Week one - study types Flashcards
memorise the different study types in epidemiology (30 cards)
Define the difference between descriptive studies and Causal studies:
In descriptive studies are only used for measuring either the exposure or the outcome. Eg what proportion of people are exposed. Or what proportion of people have the outcome. They are not trying to find out what causes the result.
Causal studies aim to determine causes the outcome. The require a group of people where some are exposed and some people not exposed. Some people who are exposed get the outcome and some don’t and visa versa.
What study designs are usually used for descriptive studies?
Usually a cross-sectional study or a survey however could be a descriptive cohort study (an example of a cohort study being used for a descriptive study = following a group of primary school students over a period of time to find out how many of them get an injury – this is descriptive as it is not trying to figure out WHY they got the injury but simply looking at the incidence).
What study designs can be used for studies that are testing an intervention?
Randomised control trials, cohort studies, case-control studies, case-series studies, ecological studies, before and after studies.
What study designs can be used for studies testing Aetiology (the harmfulness of something)?
Randomised control trials, cohort studies, case-control studies, case-series studies, ecological studies, before and after studies.
Which study designs will help to answer a question of ‘which test is better’ using one test against a reference standard?
Cross- sectional analysis
Which study designs can be used to compare two diagnostic tests?
RCT, cohort studies, case-control studies.
Which study designs help determine the frequency of occurrence of a disease?
Cross-sectional studies or cohort studies.
What is a randomised control trial and what are its main strengths and weaknesses?
Randomised control trials (RCTS) start with a group of people. The group of people are allocated BY CHANCE into either an exposed group or a non exposed group. These people are followed for a period of time to see which people are more likely to develop the outcome. Strengths = randomisation balances the known and unknown factors or variables amongst each group and therefore rules out confounding bias. Weaknesses = expensive, requires good follow up and equality of groups is highly dependant on group size.
Difference between a cohort study and an RCT
Cohort studies follow the exact same design as an RCT except that the participants are not allocated to groups by randomisation and could be allocated either by participant choice (eg lifestyle choice) or sometimes by investigators choice. Most commonly used in public health as a lot of things are impossible to randomise (eg diet and level of activity). Strengths = next best evidence to an RCT. Generally provide the best information concerning the courses of disease and the most direct and and intuitive measurements at of the risk of developing the disease. Weaknesses = not randomised cannot rule out confounding bias. Usually long and expensive. Can only usually be used to study conditions that are fairly common. Groups in cohort studys are intrinsically less similar that in an RCT.
What is a case-control study and what are the positives and negative?
In case-control studies participants are selected based on whether they have an outcome or don’t and then they are studied to find out whether or not they were exposed to work out the relative risk. For example a group of people with lung cancer would be selected and then a group of people with similar characteristics without lung cancer would be selected. Both groups would be followed back in time to find out what exposures they had and how much more likely the were to get the outcome if they were exposed vs if they were not exposed (relative risk). Positives = better for rarer diseases. Generally quicker to undertake and cheaper to perform. Disadvantage = it is crucial to have a well defined study base.
What is the difference between a well defined study base and a poorly defined study base in case-control studies?
If a study base is well defined you would be able to take anybody in the history of the world and be able to easily tell whether they belong in the study base. has a defined person, time and place.
Define a crossover trial and its positives and negatives.
A crossover trial is a trial where the outcomes among a group of people exposed to the factor of interest are compared with the outcomes among the same group of people when they were not exposed. The participants serve as their own controls.
Positives = it eliminates much of the variability that is inherent when we compare different groups of people and that can never be fully eliminated using randomisation. Can produce statistically valid and valid results using fewer participants
Negatives = rarely possible in real life and an individual can not be exposed and unexposed at the same time, therefore there is the potential that time moving on could have led to changes within the individual and impacting on outcomes. Only good for assessment assess factors that have a rapid effect.
What is the difference between a clinical trial and a preventative trial?
Clinical trials compare two (or more) forms of treatment for patients with a disease where as preventative trials aim to intervene to reduce an individuals risk of developing a disease.
What is a single- blind study and a double-blind study?
Single blind studies are studies where the patient is unaware of their allocation into either exposed or un exposed. Double blind studies are where both the participant and the trial investigators are unaware of whether the participant is in the intervention or placebo group.
Describe a N-of-1 trial.
It’s a variant of a crossover trial that only involves a single patient. Often both participant and treating doctor are blinded.
Positives = comparing the same person on and off treatment and control at different times and it comes close to achieving true exchangeability.
Negatives = not generalisable to the same extent as RCTs however is good to guide individual clinical decisions.
What is a cluster RCT?
Same designs as an RCT however groups or clusters of people are randomised together.
Define a community trial
A community trial is a cluster trial implemented at a community level. For example one community is randomly allocated to having fluoride in their water system and another is not.
Give an example of a pre-post study and what are its main flaws
A pre-post study often looks at historical controls where measure health outcomes following an intervention in a community are compared to the measured outcomes before that intervention. For example measuring the mortality rates of road accidents before and after the introduction of seatbelt wearing legislation. Main flaws = it assumes that the only or most important thing that has changed in the intervention and this may not be the case.
What are other names for a cohort study?
A cohort study is sometimes referred to as a prospective or a longitudinal study.
What is the difference between an observational study and an interventional study and in what circumstances are observational studies more appropriate?
Interventional study intervene by adding an exposure to someone and testing the outcome. Observational studies do not involve intervening in any way and instead the researcher takes an observational role and measures rate of occurrence, compares patterns of exposure and disease to identify particular exposures or risk factors to that disease. Often you can not deliberately expose someone to something that is considered harmful because it would be inethical. Hence the widespread use of observational studies in epidemiology.
Give an examples of a way to avoid recall bias when undertaking a cohort study.
Collect exposure data prior to the outcome/ at the commencement of the study. In a cohort study that requires long term follow up recontacting participants every few years to collect updated exposure data.
What is a historical or retrospective cohort study?
A cohort study which studies the records of past exposure of a group who can then be traced to determine there current health.
Briefly describe a case-cohort study.
In a case-cohort study a group of participants are selected from the original cohort at the baseline (ie. Prior to development of desease). Detailed exposure information is then compared to ALL the people in the original cohort study that did develop the disease. This reduces effort and cost while maintaining the major advantage of a cohort study in that exposure date were originally retrieved prior to development of disease.
Briefly describe a nested case-control study.
It is essentially a case-control study that is ‘nested’ within an existing cohort study. The cases are selected by ALL cohort members who developed the disease of interest and the ‘controls’ are matched at random from cohort members of the same age and gender who were disease free at the time the case was diagnosed. Just like case-controls it is assumed the exposure experience of the non-cases in the cohort is similar. Nested case controls are simpler to analyse than case-cohort studies but require a different control group for each case group.
