Chapter 14- Deriving inferences from epidemiologic studies Flashcards
Which 2 types of studies depend on nonrandomized observations?
Case-control and cohort studies
Natural experiments
When researchers take advantage of groups who have been exposed for non-study purposes, like occupational cohorts in specific industries. Then, the exposed group is compared to an unexposed group
Sequence of studies in human populations (5)
- Clinical observations
- Available data
- Case-control studies
- Cohort studies
- Randomized trials
Clinical observations
When researchers observe an association between variables- like when a surgeon observes that almost all of the lung cancer patients they operate on are smokers
Available data
Analyzing routinely available data could provide more information about the research question. Then, new studies (case control and cohort) can be carried out
Case-control study
If looking at whether smoking causes lung cancer, the researcher could compare the smoking histories of their patients with lung cancer with those of a group of patients without lung cancer. If the case control study suggests that a certain exposure might be associated with disease, a cohort study might be done next
Cohort study
Comparing smokers and nonsmokers and determining the rate of lung cancer in each group or comparing workers exposed to a toxin to workers without the exposure
2 step process in carrying out studies
- We determine whether there is an association or correlation between an exposure or characteristic and the risk of a disease- studies of group or individual characteristics can be done
- If an association is demonstrated, we determine whether the observed association is likely to be a causal one
Real or spurious associations
Poor sample selection could result in a spurious or false association between the variables
Confounding variable
A third factor linked to two variables that appear to be related. If a confounding variable is discovered, the relationship isn’t causal
Direct causation
A factor directly causes a disease without any intermediate step
Indirect causation
A factor causes a disease but only through intermediate steps. In human biology, intermediate steps are almost always present in any causal process
Types of causal relationships (4)
- Necessary and sufficient
- Necessary but not sufficient
- Sufficient but not necessary
- Neither sufficient nor necessary
Necessary and sufficient
Without a specific factor, the disease never develops (the factor is necessary), and in the presence of that factor, the disease always develops (the factor is sufficient). This type of relationship rarely occurs. Developing an infectious disease after exposure would represent a necessary and sufficient relationship, although it is more complicated since many other factors influence susceptibility to infection
Necessary but not sufficient
This means that multiple factors are required, often in a specific temporal sequence. Carcinogenesis is an example- it is a multistage process involving a promoter acting after an initiator has acted. An initiator or a promoter acting alone will not produce a type of cancer
Necessary
Without a specific factor, the disease never develops
Sufficient
In the presence of a specific factor, the disease always develops
Sufficient but not necessary
In this relationship, the factor alone can produce the disease, but so can other factors acting alone. For example, radiation exposure or benzene exposure can cause leukemia. However, cancer does not develop in everyone who has experienced radiation or benzene exposure
Neither sufficient nor necessary
This model is complex and most accurately represents the causal relationships that operate in most chronic diseases. One example is the risk factor clusters for the development of CHD, which generally don’t overlap. Individuals may develop CHD if they are exposed to smoking, diabetes, and low HDL, or a combination of hypercholesterolemia, hypertension, and lack of physical activity, or many other combinations of these factors. Each of these CHD risk factors is neither sufficient nor necessary
Analytic epidemiology
In analytic epidemiology, we aim to identify whether exposure to a determinant/factor is a cause of a health-related event or outcome
Koch’s postulates (3)
- The organism is always found with the disease
- The organism is not found with any other disease
- The organism, when isolated from one who has the disease and cultured through several generations, produces the disease in animals
The postulates only apply to infectious diseases, but indicate whether a pathogen and a disease have a causal relationship
Guidelines for judging whether an observed association is causal (9)
- Temporal relationship
- Strength of the association
- Dose-response relationship
- Replication of the findings
- Biologic plausibility
- Consideration of alternate explanations
- Cessation of exposure
- Consistency with other knowledge
- Specificity of the association
Temporal relationship
If a factor is believed to be the cause of a disease, exposure to the factor must occur before disease develops. It’s typically easier to establish a temporal relationship in a prospective cohort. In other studies, exposure information may have be located from past records. The temporal relationship is also important in regard to the length of the interval between exposure and disease- it could take years for the disease to develop
Strength of the association
Can be measured by the relative risk or odds ratio. The stronger the association, the more likely it is that the relationship is causal
Dose-response relationship
As the dose of exposure increases, the risk of disease also increases. If a dose-response relationship is present, it is strong evidence for a causal relationship. However, it this relationship is absent, it doesn’t rule out a causal relationship. Some diseases have a threshold of exposure that is necessary for the development of disease
Replication of the findings
If a relationship is causal, we would expect to find it consistently in different studies and different populations. An observed association should also be seen consistently within subgroups of the population and in different populations
Biologic plausibility
Refers to coherence with the current body of biologic knowledge. Epidemiologic observations may precede biological knowledge, but interpreting the meaning of the findings may be difficult if this is the case
Consideration of alternate explanations
In judging whether a reported association is causal, the extent to which the investigators have taken other possible explanations into account and the extent to which they have ruled out such explanations are important considerations
Cessation of exposure
If a factor is a cause of disease, we would expect the risk of the disease to decline when exposure to the factor is reduced or eliminated. However, in certain cases the disease process may be irreversible
Consistency with other knowledge
If a relationship is causal, we would expect the findings to be consistent with other data. Absence of this consistency would not completely rule out this hypothesis, however
Specificity of the association
An association is specific when a certain exposure is associated with only one disease- this is the weakest of all the guidelines. Smoking cigarettes causes many diseases- absence of specificity doesn’t negate a causal relationship. Researchers should assess the total pattern of evidence observed before they come to a conclusion
What if outcomes were like “pies”?
Each “pie” represents a conceptual scheme for how disease occurs. Each slice of the pie represents a component cause- there can be multiple. Necessary causes would be those that appear in every pie for the disease- it is required for the disease to develop. Sufficient causes are those that can bring about disease (rarely a single factor)
