Flashcards in Epi Class 5 Deck (27):
Interpretation of the Rate Ratio (RR):
The RR is the number of times greater the risk of disease in the exposed compared to the unexposed.
Relative Rate (RR) says nothing about absolute risk. It is the rate of incident disease in the exposed compared to the rate of incident disease in the unexposed. (It is possible for a disease with a very high RR to still have very low incidence, even in the exposed group.)
RR = a / (a + b)
c / (c + d)
Key measure for Cohort Studies
RR (rate ratio) = IRR (incidence rate ratio)
Key measures for Case-Control Studies
OR (odds ratio)
What do Cohort Studies evaluate?
Cohort Study: Measure the incidence of disease in the exposed versus non-exposed populations.
What do Case-Control Studies evaluate?
Case-Control Study: Compare the exposure history in people with the disease (cases) and people without the disease (controls).
Three types of RR associations
Positive Association= RISKY EFFECT
No Association= NULL ASSOCIATION
Negative Association= PROTECTIVE EFFECT
Attributable Risk (Risk Difference, AR)
the excess rate of disease in the exposed – that is, the incidence of disease
among the exposed that is due to the exposure.
Attributable Risk Formula
AR = Incidence in exposed – Incidence in unexposed
= IE – IU = a - c
(a + b) (c + d)
Definition of Attributable Risk Percent (Etiologic Fraction, AR%)
the proportion of disease among the exposed that is due to the exposure.
Attributable Risk Percent (%) Forumla
IE – IU
AR% = -------------- x 100
RR – 1
AR% = -------------- x 100
Population Attributable Risk (PAR)
the excess rate of disease in the total population that is due to the exposure.
Population Attributable Risk (PAR) Formula
PAR = It – Iu = AR x Pe
To calculate the PAR you must know the prevalence of exposure in the population (Pe). This can be a previously reported Pe from a population survey or the Pe calculated from your cohort study.
Prevalence of Exposure (Pe) Formula
Pe = (a + b) ÷ (a + b + c + d)
Incidence in the total population (It) formula
It = (a + c) ÷ (a + b + c + d)
Incidence in the total population
Incidence in the unexposed population
Population Attributable Risk Percent (PAR%) Formula
PAR% = ------- x 100
Odds Ratio (OR)
Measure of association. Used in Case-Control studies.
Odds Ratio (OR) Formula
OR = a / c ÷ b / d = ad / bc
Odds of exposure among the diseased (cases)
Odds of exposure among the non-diseased (controls)
appropriate measure of association
appropriate measure of incidence
Interpretation of the Odds Ratio (OR)
The OR is the number of times greater the odds of exposure in the disease compared to the not diseased.
POSITIVE association = implies RISKY EFFECT
NO association = NULL association
NEGATIVE association = implies a PROTECTIVE EFFECT
95% Confidence Intervals (CI)
95% CIs are required for determining statistical significance.
If the whole bar is under "1" then exposure is PROTECTIVE
If the whole bar is over "1" then exposure is RISKY
If the bar is on both sides of "1" then exposure has NO RISK ASSOCIATION
Types of matching
Frequency matching (group matching): case group and control group have similar profiles (% women, % by age, etc.)
- no need for special OR: use normal OR
Individual matching (matched-pairs matching): each case is matched to one or more individuals who is very similar (most often used in GENETIC studies where biologically-related family members are linked)
- must use matched-pairs OR
REVIEW TABLE IN NOTES!
For individual matching, a special kind of 2x2 table is used. Look for:
Concordant pairs (a and d): same exposure experience for case and control
Discordant pairs (b and c): different exposure experience for cases and controls
Matched pairs “OR” (not really an odds ratio) = ratio of discordant pairs = b/c