Study designs, limitations, mitigations, and interpretations Flashcards
What are the advantages of case control study design?
Case control studies are great for studying rare exposures and extreme relative risks. They are strong study designs during disease outbreaks and initial explorations of theories. They are also good for studying rare diseases and outcomes. They allow for the exploration of multiple exposures quickly, and work well in resource-limited or constrained settings. They have less potential for loss to follow up compared to cohort studies, and comparatively can be less expensive. They can also be used for diseases with a long induction period.
What are the disadvantages of case control study design, and how can they be mitigated?
liable to systematic error: selection and information biases. Selection bias is especially liable with regards to controls. Regarding information bias, because case control studies are retrospective, they are particularly prone to recall bias. They may also be prone to observer bias if the interview records information with different degrees of accuracy amongst controls rather than cases. They usually require a large sample size. Because they are performed with a specific sample, they cannot be used to estimate disease prevalence or incidence on a population level. Case control studies are also prone to confounding. Confounding occurs when case and control groups differ with respect to factors that might influence occurrence of disease (e.g. smoking). An additional important consideration is whether to include prevalent (people who may have had the disease for some time) or incident (newly diagnosed) cases. The problem with use of incident cases is that it is often necessary to wait for new cases to be diagnosed, whereas if prevalent cases are used, a large number of cases are already available for study. However, it is generally preferrable to use incident cases because if we use prevalent cases only, risk factors identified may be related more to survival with the disease rather than development of the disease (incidence). Finally, case control studies can be prone to reverse causality. This is because it can be difficult to establish the time sequence of exposure and disease when the measurement of exposure is being done after the outcome has been diagnosed. Therefore, it is possible for the exposure to be the result rather than the cause of the outcome. This can be minimized by ensuring that data on the exposure is collect for the time period that is relevant to the research question, and, where possible, use records from before the outcome occured.
What kind of statistical calculations are used by case control studies?
Case control studies use odds ratios, which is also referred to as the “measure of association” - meaning that we are measuring the strength of association between an exposure and an outcome. Odds ratios can be adjusted for confounding using logistic regression.
What does an OR score tell us?
OR > 1 means greater odds of association with the exposure and outcome. OR = 1 means there is no association between exposure and outcome. OR < 1 means there is a lower odds of association between the exposure and outcome. The OR is occasionally referred to as a “cross-product ratio”.
How do you correctly word the interpretation of an OR?
The odds of OUTCOME were OR times higher (or lower) amongst CASES than CONTROLS
How do you calculate an OR?
Odds ratio = odds of exposure in the diseased (cases)/ odds of exposure in the non-diseased (controls) = (a/c)/(b/d) =ad/bc =OR!
What is a case control study?
Case control studies are defined by outcome and not exposure – two or more groups are selected: “cases” with a disease and “controls” with no disease at beginning of the study. These two groups are compared to assess differences in historical exposure to risk factors. They are retrospective, beginning with outcome and working backwards to identify exposure. . Hypothesis definition is more important than in other study designs as it determines the choice of participants.
What is matching, and why is it used in case control studies?
Controls should meet all criteria of cases
except the disease status itself. The source of controls needs to be appropriate if not simply randomly selected from the general population. Controls: can be matched by age, sex or other factors on an individual or group level. Over-matching occurs when matching variables are closely related to exposure variables of interest but not related to disease. (There is no need to
select more than 4 controls for each case and virtually no extra power is added). Controls should be a population that is comparable but for the risk factor under analysis and who would otherwise have been classified among the cases of disease.
What is a cohort study?
A cohort study begins with a group of exposed people and compares them to an unexposed group.
What is confounding, and how can it be controlled for?
Case control studies are also prone to confounding. Confounding occurs when case and control groups differ with respect to factors that might influence occurrence of disease (e.g. smoking). It can be minimized at design stage by using matching, and via logistic regression and stratification at analysis.
What is the difference between a prospective and retrospective study design?
In prospective studies, the exposure is recorded PROSPECTIVELY - i.e. before diagnosis. In retrospective studies, the exposure is recorded RETROSPECTIVELY - i.e. after diagnosis
What is the difference between a case control study and a retrospective cohort study?
Case control is commonly a retrospective study but is not the same as retrospective cohort study.
The difference between the two is : in case control studies the investigator allocates individuals into either cases or controls by their disease status whereas in a retrospective cohort study, the investigator splits study individuals into cases or controls based on their exposure status.
case controls can provide odds ratio, which is the odds of disease among exposed individuals divided by the odds of disease among unexposed. Whereas cohort studies provides relative risk, which is a degree of risk/probability of disease occurrence due to exposure compared to non-exposure.
Can case control studies ever be prospective?
Case control studies could be prospective if they are nested within a cohort study, because they are being pursued with the prospective cohort.
What is a cross-sectional study?
In cross-sectional studies, the frequency of a particular exposure(s) or outcome(s) are measured in a defined population at a particular point in time
What are the advantages of cross-sectional study design?
- Cross sectional study designs are advantageous for guiding public health decision making
- The design allows for the investigation of multiple possible exposures
- The design is particularly useful for informing the health needs of a population
- The design is particuarly useful in informing the planning and allocation of health resources
- They are well suited to resource-limited contexts
What are the disadvantages of a cross-sectional study design?
- Not the strongest design to provide strong evidence about the causes of health-related outcomes
- Associations identified may be difficult to interpret
- Not suitable for studying rare or highly fatal diseases, or diseases with a short duration
- Not good for describing or comparing the frequency of rare exposures
- less value for investigating aetiological relationships than prospective studies collecting data on incident cases
- The direction of causality is not always clear
- There is limited ability to draw valid conclusions about possible causality
What is random sampling?
Random sampling is a part of the sampling technique in which each sample has an equal probability of being chosen. A sample chosen randomly is meant to be an unbiased representation of the total population.
What is a relative risk?
It is a ratio measure of association. A risk ratio (RR), also called relative risk, compares the risk of a health event (disease, injury, risk factor, or death) among one group with the risk among another group. It does so by dividing the risk (incidence proportion, attack rate) in group 1 by the risk (incidence proportion, attack rate) in group 2. The two groups are typically differentiated by such demographic factors as sex (e.g., males versus females) or by exposure to a suspected risk factor (e.g., did or did not eat potato salad). Often, the group of primary interest is labeled the exposed group, and the comparison group is labeled the unexposed group.
How do you calculate a relative risk/risk ratio?
= risk of disease (incidence proportion, attack rate) in primary interest group ( normally the exposed) /Risk of disease (incidence proportion, attack rate) in comparison group (normally the unexposed)
= (a/a+b)/(c/c+d)
In summary, it is the ratio between the risks of exposure and non-exposure
How does setting up a random sampling frame work?
- Random sampling is when potential participants are randomly selected from a defined population, which relates to the study question at hand
- A random sampling frame must use a process where all potetnial participants have an equal chance of being selected
- The potential participants which meet the defined population criteria must first be listed in the program
- This can involve a computer program which randomly allocates each potential participant with a number
- The program can then be run to randomly select the needed number of participants for the sample size that has been identified, using the randomly allocated numbers
What is an ecological study?
Ecological studies are used to understand the relationship between outcome and exposure at a population level, where ‘population’ represents a group of individuals with a shared characteristic such as geography, ethnicity, socio-economic status of employment. What differentiates ecological studies from other studies is that the unit analysis being studied is the group, therefore inferences cannot be made about individual study participants. On the other hand, details of outcome and exposure can be generalized to the population being studied. Examples of such studies include investigating associations between units of grouped data, such as electoral wards, regions, or even whole countries.
When are ecological studies used?
Correlating population disease rates with factors of interest, such as healthcare use.
Demonstrating changes in mortality over time (time series)
Comparing the prevalence of a disease between different regions at a single point in time (geographical studies)
What are the advantages of an ecological study?
Ecological studies are particularly useful for generating hypotheses since they can use existing data sets and rapidly test the hypothesis. The advantages of the ecological studies include the large number of people that can be included in the study and the large number of risk-modifying factors that can be examined. They can also be generalized well to populations.
What are the disadvantages of an ecological study?
The term “ecological fallacy” means that risk-associations apparent between different groups of people may not accurately reflect the true association between individuals within those groups. Ecological studies should include as many known risk-modifying factors for any outcome as possible, adding others if warranted.
