Epidemiology Flashcards
Equation for Relative Risk with chart
[A/(A+B)] / [C/(C+D)]
What does an RR of 1 signify
Same risk in both groups
no association between exposure and outcome
What does an RR > 1 signify?
exposure is a risk factor for developing the condition
What does an RR= 0 signify?
exposure is protective against the condition
What is the equation for absolute risk reduction?
ARR = [A/(A+B)] - [C /(C+D)]
What is the significance of ARR = 0
no association
What is the significance of ARR > 0
positive causal relationship between exposure and disease
What is the significance of ARR < 0
exposure is protective
What is the equation for NNT?
NNT = 1/ARR
What is the equation for relative risk reduction
RRR = 1-RR
Equation for Odd’s Ratio
OR = (A x D) / (B x C)
What type of study can use an odd’s rattio
All
What type of study CANNOT calculate a relative risk?
case-control study
What is the significance of OR = 1
No association between exposure and disease
What is the significance of OR < 1
exposure protects against disease
What is the significance of OR > 1
exposure increases chance of disease
Calculation for sensitivity
A / (A+C)
Calculation for specificity
D / (B+D)
Equation for positive predictive value
A / (A+B)
Equation for Negative predictive value
D / (C+D)
What is the positive predictive value
probability of having the disease if you have a positive test
What is the negative predictive value
Probability of not having a disease if you have a negative test
Equation for positive likelihood ratio
= sensitivity / (1-specificity)
Equation for negative likelihood ratio
= (1-sensitivity) / specificity
What are the types of selection bias?
sampling bias
recruitment bias
volunteer bias
indication bias
survival bias
What are the types of observation bias?
recall bias
Hawthorne effect
Evaluation bias
Loss to follow up
Classification errors
Types of descriptive studies
Correlation studies
Case report/series
cross-sectional studies
cohort studies
Types of analytic studies
Case-control
case-crossover
cohort
randomized control trial
Disadvantage of correlation study
unable to determine cause/effect
high confounding factors
Disadvantages of case report/series
no control group
less generalizable
can’t determine cause/effect association
Advantage of cross-sectional studies
evaluates participants individually therefore more nuance re: independent variables
Disadvantages of cross-sectional studies
no evaluation of temporality of risk factors
cannot evaluate incidence
Advantages of case-control studies
**rare diseases
exposure occurred a long time ago
evaluate multiple variables/exposures
good for hypothesis testing
identify risk factors related to a condition
Disadvantages of case-control studies
can’t define incidence
selection bias
recall bias
can’t use relative risk
What is a case-crossover study?
determine if a transitory exposure increases risk of having a disease in that time of exposure (eg. increased risk of heart attack during intense activity)
Advantages of case-crossover studies
reduce interindividual variability because each person is their own control
Disadvantage of case-crossover studies
recall bias
How is a cohort study done?
Participants chosen based on whether or not they’ve been exposed to a chosen variable then compare incidence of new pathology in the different groups
Advantages of cohort studies
better temporal relationship understanding
can evaluate rare exposures
can study risk factors that would otherwise be unethical
can estimate incidence based on exposure
less selection bias
Disadvantages of cohort studies
longer term and more costly
risk of losing patients to follow-up
What is a sampling bias?
participants not representative of general populaion
What is recruitment bias
recruited participants have certain characteristics related to the disease
What is indication bias
exposure studied is part of diagnostic criteria for a disease
What is the hawthorne effect
people act differently when observed
What is evaluation bias
evaluator is preferentially looking for certain exposures
What are the features of a confounding factor?
factor that is associated to the exposure and associated to the issue but the association is independent of the exposure
