ESTIMATION OF RISK, IS THERE AN ASSOCIATIONPSM.pptx

Estimating Risk : Is There an
Association

LEARNING OBJECTIVE
• To explore the concept of absolute risk.
• To introduce and compare relative risk and odds ratio as
measures of association between an exposure and a
disease.
• To calculate and interpret a relative risk in a cohort study.
• To calculate and interpret an odds ratio in a cohort study and
in a case-control study and to describe when the odds ratio is
a good estimate of the relative risk.
• To calculate and interpret an odds ratio in a matched-pairs
case-control study.

BASIC STUDY DESIGNS IN EPIDEMIOLOGIC
INVESTIGATIONS

ABSOLUTE RISK
• The incidence of a disease in a population is termed as the
absolute risk.
• Indicates the magnitude of the risk.
• Does not indicate whether the exposure is associated with an
increased risk of the disease.
• Does not stipulate any explicit comparison ; but implicit
comparison.

HOW TO DETERMINE ASSOCIATION BETWEEN
DISEASE AND EXPOSURE
• Eg:-Results of an investigation of a foodborne disease
outbreak.
• Suspect foods were identified ;
• For each food, the attack rate of the disease for those who
ate the food(exposed)
• The attack rate for those who didn’t eat the food(unexposed)

• To determine whether an
excess risk is associated with
each of the food items?
 One approach is----
• Calculate the ratio of the attack
rate in ‘exposed’ to the attack
rate in ‘unexposed’, for each
food.
• As shown in the column C of
the following table-

• Alternate approach---
• As shown in column D
• Subtract the risk in the
‘unexposed’ from the risk
in ‘exposed’.
• The difference represents
the excess absolute risk in
those who were exposed.
Thus an excess risk can be
calculated as :
 The ratio of risks(or of
incidence rates) :
Disease risk in exposed/
Disease risk in unexposed
 The difference in the
risks(or in the incidence
rates) :
(Disease risk in exposed)
– (Disease risk in
unexposed)

DIFFERENCE IN BOTH THE WAYS OF
CALCULATION
 Does the method that we choose to
calculate excess risk make any
difference?
 Another example-Two communities, A
and B.
 Incidence if a disease in exposed in A
is 40% ; In unexposed-10%
 Ratio of the rates= 4.0
 Difference in incidence
rates=30%.
 Similarly in B- Ratio -> 90/60=1.5

RELATIVE RISK
• Both case control and cohort studies designed to determine if
there is an association between exposure and disease.
• As per Cohort study, question would be-
• Relative risk = Risk in exposed/Risk in unexposed
• It can also be defined as
• Probability of an event occurring in exposed people
compared with the probability of the event in unexposed
people ; as the ratio of these two probabilities.

INTERPRETING THE RELATIVE RISK
• If the Relative risk = 1
• No evidence for any increased risk in exposed
individuals .
• If Relative risk >1
• Evidence of positive association , which may
be causal.
• If Relative risk <1
• Evidence of negative association ; protective
effect.
• People who are given an effective vaccine

CALCULATION OF RELATIVE RISK IN COHORT
STUDIES

RELATIVE RISK : EXAMPLE
• This Fig shows the data based on merging 2,282 middle aged men followed for 10yrs
in the Framingham study
• 1,838 middle aged men followed for 8years in Albany, New York.
• Data relate smoking, cholesterol level, BP to risk of Myocardial Infarction and Death
from CHD.
• The value of 1 is assigned to the LOWEST of the risk ; other risks are calculated
relative to this figure.
• Left pic- shows risk in non smoker with low cholesterol level(set to be as 1)
• Both smoking and elevated cholesterol levels contribute to the risk of MI and death
from CHD.
• Right pic- Comparable analysis of BP and smoking

ODDS RATIO(RELATIVE ODDS)
• In a Case-Control study , Incidence of disease in exposed or
in unexposed is not known because
• We start with diseased people(cases) and non diseased
people(control).
• Relative risk cannot be calculated directly; but Odds Ratio.
• OR can be obtained from either of the study(cohort/case
control study)

 Proportion of the cases exposed and the proportion of controls who were exposed.

Odds = Probability that Epi Beauty will win the race/ Probability that Epi
Beauty will lose the race
 P=Probability that Epi will win the race.
1-p= Probability that Epi Beauty will lose the race.
 Odds =p/1-p
 Probability of winning=60%
 Odds of winning=60%/40%= 1.5

ODDS RATIO IN COHORT AND CASE CONTROL
STUDIES

INTERPRETING ODDS RATIO
 When is the ODDS ratio a good estimate of the Relative Risk

CALCULATION OF ODDS RATIO IN AN
UNMATCHED CASE CONTROL STUDY

CALCULATION IN MATCHED PAIRS CASE
CONTROL STUDY

ESTIMATION OF RISK, IS THERE AN ASSOCIATIONPSM.pptx

ESTIMATION OF RISK, IS THERE AN ASSOCIATIONPSM.pptx

Related slideshows

More Related Content

ESTIMATION OF RISK, IS THERE AN ASSOCIATIONPSM.pptx