(M) Validity of Epidemiologic Studies Flashcards
try to provide accurate answers to questions
Epidemiological studies
T or
Estimates ≠ Real Prevalence or Real Risk → Error
T
involves generating and testing hypothesis about factors that cause or prevent disease
Epidemiologic research
eliminate alternative explanations for his/her findings
major objective of every investigator who tests an etiologic hypothesis
one in which the observed association is not due to various sources of error (systematic and random errors)
valid study
from the poor design and/or conduct of the study
* Noncomparability of Groups
* Measurement Flaws
bias from systematic error
Underestimate or overestimate the true measure of association
bias from incorrect estimate of the measure of association
Results from procedures used to select subjects and factors that influence participation in the study
Selection Bias
Selection Bias
Groups being compared should be as similar as possible with respect to all other factors that may be related to the disease except the determinant under investigation
Principle in the Selection of Study Groups
Types of Selection Bias
- Sampling Bias
- Allocation Bias
- Responder Bias
Systematically excluding or over-representing certain groups
Sampling Bias
E.g. A study to estimate the prevalence of smoking in a population, choosing a city center as location for study
Sampling Bias
Systematic differences in the way in which subjects are recruited into different groups for a study
Allocation Bias
E.g. A study may fail to do random sampling
First 20 patients who arrived at the clinic are allocated to a new treatment
Next 20 patients are allocated to an existing treatment
However, the patients who arrived early may be fitter or wealthier, OR alternatively the doctor may have asked to see the most seriously ill patients first
Allocation Bias
Missed responders or non-responders
Responder Bias
E.g. A study may send questionnaires to members of the control group. If these subjects are from a different social class to the cases, there may be differences in the proportion of responses that are received.
Responder Bias
AKA. measurement error, misclassification bias, observation bias
Information Bias
A flaw is measuring exposure or outcome variables that resultes in incorrect information between comparison group
Information Bias
Systematic differences in data collection, measurement or classification
Information Bias
Types of information bias
- Recall Bias
- Social acceptability
- Recording Bias
- Interview Bias
- Follow-up Bias
- Misclassification Bias
People suffering from a disease may have spent more time thinking of possible links between their past behavior and their disease than non-sufferers
Recall Bias
Cases may report more exposure to possible hazards
Common in case-control studies
Recall Bias
Some subjects may exaggerate or understate their responses, or deny that they engage in embarrassing or undesirable activities
Ex. cheating – they may deny this because it is not socially accepted
Social Acceptability Bias
Medical records may contain more information on patients who are “cases”
Recording Bias
Interviewers may phrase questions differently for different subjects, or write down their own interpretations of what subjects have said
Question of phrasing is different from each group
Interviewer Bias
In studies that follow up at intervals, people from certain groups may tend to be lost to followup, or a disproportionate number of exposed subjects may be lost to follow-up compared with non-exposed subjects
Follow-up Bias
Patients may be systematically misclassified as either having disease or exposure
Misclassification Bias
E.g older people of lower social class may be less likely to express dissatisfaction with a health-related service
Some groups may give different responses
T or F
Investigators may look more closely at exposed patients, to try to find the presence of a disease, or they may be more attentive to certain types of subjects.
True
The mixing of effects between the exposure, The disease, and a third variable
Confounding
what is the third variable
Confounder
When present, the association between exposure and disease is distorted
The “third variable problem”
confounder
Have an effect on the independent variable, and have big effect on disease
The “third variable problem”
confounder
Occurs when a separate factor (or factors) influences the risk of developing a disease, other than the risk factor being studied.
Confounding (confounder)
the factor has to be related to the exposure, and it also has to be an independent risk factor for the disease being studied.
To be a confounder
common causes of confounding.
Age and Sex
wow naol
Confounding =
Spuriousness
(not genuine or authentic )
if u see this card
go over the example of confounding
T or F
Majority of study is PERFECTLY VALID
F - no study dumbass
fators that contributes to study why there is no perfectly valid
- Residual confounding
- Unpredictable nature of chance
- Complexity of bias
Eliminate Effect of Confounding in Studies
- Randomization
- Matching
- Stratified analysis
ensuring that samples are randomly selected
Randomization
In Case-Control study, controls are matched to cases at the start of the study according to particular characteristics which are known to be present in cases (e.g. age, sex, ethnic group, smoking, etc)
Matching
dividing subjects into groups at the analysis stage (e.g. by sex, age group, smoker/non-smoker) and analyzing on this basis.
Stratified Analysis
if u see this card
yey tapos kana
talon sa jeep
