Introduction to Study

Designs
B. Analytic
Studies
CMHS-SNM, Huye campus

Introduction to Study Design –

Analytical Studies
 Epidemiologic
• Approach
Define the population of interest
• Select the sample to be studied
• Enumerate the disease, or health-
related state in the population
• Define the exposure and outcome
variables
• Determine association between a
factor or exposure and disease or
outcome of interest
– Study characteristics
Introduction to Study Designof groups Studies
– Analytical and
individuals in those affected and
 Epidemiologic Approach
• If an(2)
association is found, establish if
there is a cause and effect
relationship between the exposure
and outcome using analytical studies
• Criteria that are used to establish
cause-effect relationships include:
– Biologic plausibility (biology, medicine)
– Dose-response relationship
(pharmacology, biology)
– Specificity of exposure (clinical
medicine)
– Strength of association,
Introduction to Study Design –replication and
Analytical Studies
 The Epidemiologic
Study
• Enables assessment of the effect of an
exposure variable of interest on the
outcome
?
Exposure of Outcome of
interest interest
(e.g. contraceptive (Breast cancer)
use)

• Exposure – characteristic,
independent variable
• Outcome – effect, dependent variable
Introduction to Study Design –
Analytical Studies
Exposure - Outcome
Model
Simple Model

Exposure
Outcome

Complicated Model
Confounder(s)

Exposure
Introduction to Study Design – Analytical Studies
Outcome
 Taxonomy of Study
Designs
I. Observation Studies
A.Descriptive Studies
– Case reports
– Case Series
– Ecological studies
– Cross-sectional studies
B.Analytical Studies
– Case-Control Studies (retrospective
studies)
– Cohort Studies (prospective or
longitudinal studies)

II.Experimental Studies
A.Randomized Controlled Trials (Clinical
Trials)Introduction to Study Design – Analytical Studies
 Descriptive vs
• Key isAnalytical
whether one can determine that the
exposure came before the outcome
• In a descriptive study one cannot
determine that the exposure came before
the outcome ie descriptive studies are
subject to the egg and chicken affair
• In an analytical study one can
determine that the exposure came
before the outcome and also establish
a cause and effect
Introduction to Studyrelationship
Design – Analytical Studies
Descriptive vs
Analytical (2)
• Descriptive epidemiology:
– Who, What, When, and
Where

• Analytic epidemiology:
– Why and How

Introduction to Study Design – Analytical Studies

 Analytical
Epidemiology
• The epidemiologist seeks to answer
the following questions:
1. Do persons with the disease have the
exposure more
frequently than those without the
disease?
2. Do persons with the exposure develop the
disease more frequently than those who
do not have the exposure?
• In short, through analytical
epidemiology one undertakes
Introduction to Study Design – Analytical Studies
to
test hypotheses regarding disease
 Design of Case-Control
Study
Exposed Non-Exposed Exposed Non-
Exposed

Disease Non-Disease
“Contro
“Case
ls”
s”

Introduction to Study Design – Analytical Studies

 Design of Case Control
StudiesFIRST,
SELECT

CASES CONTROLS
(With (Without
Disease) Disease)
Exposed a b
Measu
re
past Not d
c
expos Exposed
ure
Total a+c b+d
s
Proportions a B
b+d
Introduction to Study Design – Analytical Studies
Exposed a+c
 Characteristics of Case-
Control Study
• Sampling is with regard to disease
outcome or effect
• History of exposure or characteristic is
determined to have occurred prior to the
time of study
• Establish a temporal relationship between
exposure and disease outcome
• A comparison or control group is
essential to establish
Introduction a –cause
to Study Design Analyticaland
Studies
 Sources of Cases
• All cases diagnosed in the community
or general population (in hospitals,
facilities or clinics)
• All cases diagnosed in a sample of the
general popn
• All cases diagnosed in all hospitals in
the community
• All cases diagnosed in a single
hospital
Introduction to Study Design – Analytical Studies
 Issues in Selection of
Cases
• Need clear cut definitions and diagnostic
criteria for cases
• Cases selected from a single hospital may
reflect the referral pattern and not the
risk factors for disease
• If hospitalized cases are to be used, it is
better to use several hospitals
• Use of incident cases versus prevalent
cases
Introduction to Study Design – Analytical Studies
 Sources of Controls
• Sample of the general population in a
community
• Non-cases in a sample of the general
population, or sub-sample of a sample of
general population
• Sample of patients in all hospitals in the
community, who do not have the disease
or related diseases
• Sample of patients in the same hospital
as the Introduction
cases to Study Design – Analytical Studies
 Issues in the
Selection of
• Matching
Controls
a) Group or frequency matching
b) Individual matching

• Selective re-call bias

– Cases more likely to recall exposure than
the controls

• One control per case versus multiple

controls
• Controls of the same type versus
Introduction
multiple to Study of
controls Design – Analytical types
different Studies
 Advantage of Case-
Control
• Well suitedStudy
to the study of rare
diseases or those with long latency
• Relatively quick to mount and conduct
• Relatively inexpensive
• Requires comparatively few subjects
• Existing records can occasionally be
used
• No risk to subjects
• Allows Introduction
study of multiple potential
to Study Design –
Analytical Studies
 Disadvantage of Case-
•
Control Study
Relies on recall/records for information on past
exposures
• Validation of information is difficult or impossible
• Control of extraneous variables may be
incomplete
• Selection of appropriate controls may be difficult
• Rates of disease in exposed and unexposed
individuals cannot
be determined
• Method relatively unfamiliar to medical
community and difficult
Introduction to explain
to Study Design –
Analytical Studies
 Analytical
• The Epidemiology
epidemiologist seeks to answer the
following questions:
1. Do persons with the disease have the
exposure more
frequently than those without the disease?
2. Do persons with the exposure develop the
disease more frequently than those who do
not have the exposure?
• In short, through analytical
epidemiology one undertakes to test
hypotheses regarding disease
aetiology
Introduction to Study Design – Analytical Studies
 Design of a Cohort
Study

IDENTIF Exposed Non

Y:
Exposed

Develop Do not Develop Do not

Disease Develop
Disea Disease Develop
Disea
se se

Introduction to Study Design – Analytical Studies

 Selection of Study
Populations
IDENTIFY Defined
population
:
Then,
by history NON -
and/or RANDOMIZED
tests
separate
into:
and Non-
Exposed an
d exposed Prospectiv
e
Study

Develop Do not Develop Do not

Disease develop Disease develop
Disease Disease
Introduction to Study Design –
Analytical Studies
 Prospective (Cohort)
CONCURREN Studies NON-
Defined CONCURRENT
T
Populatio 2007
2017
n

NON -
RANDOMIZED
201 Exposed Non-exposed 200
7 7

202 Disease No Disease No 201

7 Disease 7
Disease
TIME-FRAMES FOR A HYPOTHETICAL CONCURRENT AND
A NON- CONCURRENT PROSPECTIVE STUDY
CONDUCTED IN 2017
Introduction to Study Design – Analytical Studies
 Characteristics of Cohort
• Study
A cohort is a group that share a similar
experience
• Start the study with regard to exposure
status in a disease free at-risk eligible
population
• The 2 exposure groups are followed
forward in time to ascertain the
occurrence of new cases of disease
• A comparison group is essential to
establish a cause and effect relationship
Introduction to Study Design – Analytical Studies
 Advantages of cohort
• studies
Can measure disease incidence
• Can study the natural history
• Provides strong evidence of casual
association between Exposure and
Disease (time order is known)
• Provides information on time between
Exposure and Disease
• Multiple diseases can be examined
• Good choice if exposure is rare (assemble
special exposure cohort)
• Generally less
Introduction susceptible
to Study to bias
Design – Analytical Studies vs Case
 Disadvantages of cohort
•
studies
Takes time, need large samples,
expensive
• Complicated to implement and conduct
• Not useful for rare diseases/outcomes
• Problems of selection bias
• At start = assembling the cohort
• During study = loss to follow-up
• With prolonged time period:
• loss-to-follow up
• exposures change
(misclassification)
Introduction to Study Design – Analytical Studies
• Confounding
 Nested Case-Control
•
Studies
Case control studies conducted within
an ongoing cohort study i.e nested
within a cohort study
• Cohort Study:
– Population is identified, classified into
exposed and non-
exposed and then followed up over time
– Detailed questionnaire and biological samples
collected at baseline
– During follow-up a proportion develop disease

Introduction to Study Design – Analytical Studies

 Nested Case-Control
• Studies
Nested Case Control Study
– Identify all cases and a random sample of
controls
– If the controls are a sample of those who do
not develop disease at the time each case
occurs
• INCIDENCE DENSITY SAMPLING OR RISK
SET SAMPLING
– If the controls are a sample of the total
cohort at baseline
• CASE COHORT STUDY
Introduction to Study Design – Analytical Studies
 THANK YOU!

Introduction to Study Design – Analytical Studies