MARIEL LOPEZ & MARITZA RENEAU
Foreign Languages
Warm-Up
 Identification and Classification of Outcome
 Medical condition
 Psychological or social problem
 Positive
 Identification of Exposure
 Higher probability
 Protective effect
Warm-up
Outcome
Medical condition
Psychological or social
problem
Risk Factor-Possible effect
Higher probability
Protective effect
Positive
Warm-up
Medical condition
Lung cancer
Psychological or social
problem
Teen pregnancy
Risk Factor-Possible effect
Higher probability
Smoking
Parents with low level of
education-
Protective effect
eat breakfast
Positive
Good academic
performance
Warm-up
 Enduring Epidemiological Understanding: Making
group comparison and identifying association
 General model
 Specific model : Smoking and lung cancer
Warm-up
Association of interest
Exposure
Disease
Warm-up
Association of interest
Smoking
What do you think is the best method to demonstrate a
causal relation? Choose the best answer
a. Experimental study
b. Observational study.
Lung cancer
Warm-up
Association of interest
Smoking
What do you think is the best method to demonstrate a
causal relation? Choose the best answer
a. Observational study. Choose the best answer
a. Case-control
b. Cohort
c. Cross-sectional
Lung cancer
Warm-up
 Cohort study- handout
 Design
 Advantages and disadvantages
Warm-up
Association of interest
Smoking
Lung cancer
Can you think of some examples of other exposures or
lifestyle choices that might be the real culprits in causing lung
cancer?
Enduring Epidemiological
Understanding
 Explaining Association and Judging Causation
LESSON OBJECTIVES
 To Understand Confounding
 To Calculate and Interpret Relative Risk
 To use Stratification in order to Identify Confounding
Variables
In what phase of the
study can stratification
be used?
a. Design
b. Analysis
Introduction- Confounding Variable
Bedsores and Mortality
Association of interest
Bedsores
Mortality
Medical
Severity
CV
Can you think of some examples of other exposures or
lifestyle choices that might be the real culprits in causing
Mortality?
Bedsores and Mortality Study
 Objective: The association between bedsores and
death among elderly hip fracture patients.
 Sample: 9,400 patients aged 60 and over, admitted
with hip fracture to one of 20 study hospitals.
 Methods: Medical charts were reviewed by research
nurses in order to identify exposure and outcome.
Analysis – Bedsores and Mortality
RR- Unadjusted
Died
Did not die
Total
Bedsores
79
745
824
No bedsores
286
8,290
8,576
Total
365
9,035
9,400
# of people with bedsore who died
# of people with a bedsore who did not die
Total # of people with a bedsore
# of people without a bedsore who died
# of people without a bedsore who did not die
Total # of people without a bedsore
Proportion of people with a bedsore who died
Proportion of people without a bedsore who died
Analysis – Bedsores and Mortality
RR- Unadjusted
Died
Did not die
Total
Bedsores
79
745
824
No bedsores
286
8,290
8,576
Total
365
9,035
9,400
# of people with bedsore who died
79
# of people with a bedsore who did not die
745
Total # of people with a bedsore
824
# of people without a bedsore who died
286
# of people without a bedsore who did not die
8,290
Total # of people without a bedsore
8,576
Proportion of people with a bedsore who died
79/824=9.6%
Proportion of people without a bedsore who
died
286/8,576=3.3%
RR=.096/.033=2.9
Introduction- Confounding Variable
Bedsores and Mortality
Association of interest
Bedsores
Mortality
Medical
Severity
CV
Can you think of some examples of other exposures or
lifestyle choices that might be the real culprits in causing
Mortality?
Analysis – Bedsores and Mortality
Adjusted by Medical Severity (PCV)
RR
U=.096/.033=2.9
High Medical Severity Group – 5 or more
diseases when admitted to hospital
Died
Did not die
Total
Bedsores
55
51
106
No bedsores
5
5
10
60
56
116
Total
RR=55/106= 1.04
5/10
Low Medical Severity Group- <5
Died
Did not die
Total
Bedsores
24
694
718
No
bedsores
281
8,285
8,566
Total
305
8,979
9,284
RR=24/718= 1.02
281/8,566
Bedsores and Mortality
PCV Medical Severity
 Is Medical Severity a confounding variable?
 According to the stratification analysis….
 According to the definition

CV


Outcome
We would expect that the people with HMS
would have a higher probability of death that
people with LMS
CV
RF
 We would expect that people with HMS would have a
higher probability of bedsores that people with LMS.
Analysis – Bedsores and Mortality
Adjusted by Medical Severity
(PCV)
MS
Mortality
High Medical Severity Group – 5 or more
diseases when admitted to hospital
Died
Did not die
Total
Bedsores
55
51
106
No bedsores
5
5
10
60
56
116
Total
Proportion of HMS
who died= 60/116=
51.7%
Low Medical Severity Group- <5
Died
Did not die
Total
Bedsores
24
694
718
No
bedsores
281
8,285
8,566
Total
305
8,979
9,284
Proportion of HMS
who died= 305/9,284=
3.3%
Analysis – Bedsores and Mortality
Adjusted by Medical Severity
(PCV)
MS
Bedsores
High Medical Severity Group – 5 or more
diseases when admitted to hospital
Died
Did not die
Total
Bedsores
55
51
106
No bedsores
5
5
10
60
56
116
Total
Proportion of people
with bedsores among
those with HMS
106/116= 91.4%
Low Medical Severity Group- <5
Died
Did not die
Total
Bedsores
24
694
718
No
bedsores
281
8,285
8,566
Total
305
8,979
9,284
Proportion of
people with
bedsores among
those with LMS
718/9,284= 7.7%
Conclusion
 The fact that the adjusted RR was different from the
unadjusted RR is evidence that there is confounding.
 Another symptom of confounding was identified by
showing that there was an association both between
bedsores and MS and dying and MS.
 There was no association between bedsores and
mortality.
More…..
 In our example, there is confounding by MS but does
that mean that the association between bedsores and
dying is not real?
 If your answer is no, why do you say so?
More…..
 In our example, there is confounding by MS but does
that mean that the association between bedsores and
dying is not real?
Answer: No. Patients with bedsores really do have a
higher risk of dying but it is not because they have
bedsores.
Bedsores are guilty by association!
Activity
 Student handout
Descargar

CONFOUNDINGS IN EPIDEMIOLOGY