Analytic Epidemiological Designs

Analytic Epidemiological

Designs

Basic Question in Analytic Epidemiology

Exposure Disease

Cohort Study

Cohort is an ancient Roman military unit of 300 – 600 men.

A group of soldiers

marching forward in

battle

In epidemiology, cohort is a group of

people who share a common characteristic

or experience within a defined period (e.g.,

are born, are exposed to a drug or a

vaccine, etc.).

Thus a group of people who were born

on a day or in a particular period, say 1980,

form a birth cohort.

A study design that follows over time one or more

populations (called cohorts) to determine which

patient characteristics (risk factors) are

associated with the development of a disease or

outcome.

Key Point:

Presence or absence of risk factor

is determined before outcome

occurs.

General consideration while selection of cohorts

Both the cohorts are free of the disease.

Both the groups should equally susceptible to disease

Both the groups should be comparable

Diagnostic and eligibility criteria for the disease should be defined well in advance.

timeStudy begins here

Studypopulation

free ofdisease

Factorpresent

Factorabsent

disease

no disease

disease

no disease

presentfuture

ANALYSIS

• Calculation of incidence rates among exposed and non exposed groups

• Estimation of risk

Incidence rates of outcome

N

dc

ba

Yes No

Disease Status

Yes

No

Exposure Status

a+b

c+d

b+d a+c

Total

Exposed

Non Exposed

Relative Risk• General interpretation of relative risk (RR)

• If RR > 1 Positive association between disease and risk factor = 1 No association

< 1 Negative association • The “reference group” is in the denominator• Reference group generally chosen as the “unexposed” group• RR estimates the magnitude (strength ) of association between

exposure & disease

Risk estimation

AR: Attributable RiskExcess risk attributed to exposure

AR%: Attributable Risk percentExpress the gain or the benefit if

the exposure removed

Incidence

I. among

exposed

A R% A R

Base-line risk

I. among non

exposed

Smoking Lung cancer Total

YES NO

YES 70 6930 7000

NO 3 2997 3000

73 9927 10000

Find out RR and AR for above data

• Incidence of lung cancer among smokers70/7000 = 10 per 1000

• Incidence of lung cancer among non-smokers3/3000 = 1 per thousand

RR = 10 / 1 = 10(lung cancer is 10 times more common among smokers than non smokers)

AR = 10 – 1 / 10 X 100= 90 %

(90% of the cases of lung cancer among smokers are attributed to their habit of smoking)

Timeframe of Studies• Prospective Study - Outcomes have not yet occurred as

study begins. looks forward, looks to the future, examines future events, follows a condition, concern or disease into the future

time

Study begins here

Timeframe of Studies• Retrospective Study - Outcomes have already occurred as

the study begins. “to look back”, looks back in time to study events that have already occurred

time

Study begins here

Strengths

• We can find out incidence rate and risk• More than one disease related to single

exposure • can establish cause - effect• good when exposure is rare• minimizes selection and information bias

Weaknesses

• losses to follow-up• often requires large sample• Ineffective for rare diseases• long time to complete• Expensive• Ethical issues

...… several famous large cohort

studies continue to provide important

information..…

Framingham Heart Study

Case Control study

Why case-control study?• In a cohort study, you need a large number of the

subjects to obtain a sufficient number of case, especially if you are interested in a rare disease.

• Gastric cancer incidence in Japanese male: 128.5 / 100,000 person year

• A case-control study is more efficient in terms of study operation, time, and cost.

Case Control

Definition….• The case-control study is an analytic epidemiologic research design in

which the study population consists of 2 groups who either have (cases) or do not have a particular health problem or outcome (controls).

• The investigator looks back in time to measure exposure of the study subjects. The exposure is then compared among cases and controls to determine if the exposure could account for the health condition of the cases.

Case Control

Case-Control Studies

Cases: DiseaseControls: No disease

Case Control

Case-control study - Sequence of determining exposure and outcome status

• Step1: Determine and select cases of your research interest

• Step2: Selection of appropriate controls

• Step3: Determine exposure status in both cases and controls

Case Control

Design of Case Control Studycases with the disease

Should have clear case definition i.e. clear criteria for defining the disease of interest

May be taken from clinics, hospitals, disease registries

Preferred newly diagnosed

Case Control

Design of Case Control StudyAppropriate controls without the disease

Should comes from the same study base or population as cases

Can come from geographical sample, medical inistitution, neighbors, friends,….

Can have multiple control groupsMay be matched

Case Control

１） a population-based case-control studyBoth cases and controls are recruited from the

population.

２） a case-control study nested in a cohortBoth case and controls are members of the cohort.

３） a hospital-based case-control studyBoth case and controls are patients who are

hospitalized or outpatients.Controls with diseases associated with the exposure

of interest should be avoided.

Types of case-control studies

Case Control

Case

-Con

trol D

esig

nStudy

population

Cases(disease)

Controls(no disease)

factor present

factor absent

factor present

factor absentpresent

past

time

Study begins here Case Control

31

Disease

+ -

Exp + a b

Exp - c d

d 1 d 0

Case-control Study – DesignSelect subjects on the basis of disease status

ORa /cb /d

adbc

Case Control

Interpretation of (OR) odds ratio> 1 means the exposure is a risk factor.

= 1 means the exposure is not associated with the disease.

< 1 means the exposure is protective

Case Control

Lung cancer Controlscases

N=100N=100Smokers (NOT recently started)

↓ ↓ 70 40 

An example of unmatched case-control study

Cases Controlssmoker 70 40

Non-smoker 30 60

Odds ratio=Case Control

Advantages

1. Simple, not time consuming (quick) and inexpensive.2. Suitable for rare diseases. 3. Can examine multiple exposures for a single disease.4. Support, but not provide causal association.5. Suitable for diseases of long latency period6. Dose response relationship can be assessed7. Small sample size8. No ethical problem

Case Control

Disadvantages1. Recall bias 2. Selection bias3. Different diagnostic tools so “case groups” may be not

homogenous.4. The chosen cases are selective survivors (the history of

died cases may be different) thus the cases does not represent a universe of cases.

5. The time sequence between the exposure and the disease is not clear.

6. Control of confounding factors7. Not suitable for rare exposure8. Only one outcome

Case Control

36

Controlling extraneous variables (confounding)• Exposure of interest may be confounded by a factor that is associated

with the exposure and the disease i.e., is an independent risk factor for the disease

A B

CCase Control

37

How to control for confounding• At the design phase

• Randomization• Restriction• Matching

• At the analysis phase• Age-adjustment• Stratification• Multivariable adjustment (logistic regression modeling, Cox

regression modeling)

Case Control

MatchingSelection of controls to match specific characteristics of cases

a) Frequency matchingSelect controls to get same distribution of variable as cases (e.g. age group)

b) Individual matchingSelect a specific control per case by matching variable (e.g. date of birth)

Case Control

Intervention or Experimental studies

Therapeutic• Study population

• Patients with disease

• Objectives• Cure patients• Diminish symptoms• Prevent recurrence of disease/risk of

death

Preventive• Study Population

• Population at risk

• Objectives• Reduce the risk of developing disease

Clinical trials are the most well known experimental design Such designs are differentiated from observational designs by the fact that there is manipulation of the study factor (exposure), and randomization (random allocation) of subjects to treatment (exposure) groups.

RCT

Why Performed ?

1. Provide stronger evidence of the effect (outcome) compared to observational designs, with maximum confidence and assurance

2. Yield more valid results, as variation is minimized and bias controlled

3. Determine whether experimental treatments are safe and effective under “controlled environments” (as opposed to “natural settings” in observational designs), especiallywhen the margin of expected benefit is doubtful / narrow (10 - 30%)

RCT

Experimental Studies• A study in which a population is selected for a planned trial of a regimen,

whose effects (consequences of some treatment on some outcome) are measured by comparing the outcome of the regimen between 2 groups.

• The subjects in the study who actually receive the treatment of interest are called the treatment group.

• The subjects in the study who receive no treatment or a different treatment are called the comparison group.

RCT

Expe

rimen

tal

Desig

n

timeStudy begins here (baseline point)

Studypopulation

Intervention

Control

outcome

no outcome

outcome

no outcome

baselinefuture

RANDOMIZATION

RCT

Types of trials

B lind ed N o t b lind ed

R a nd o m ised N o t ran d om ised

C o n tro lled N o t co n tro lled

T ria l

RCT

Design - conductDifferent phases

• Enrollment (selection of study population)

• Allocation of study regimes

• Follow-up• Maintainence and assessment of adherence• High and uniform rates of ascertainment

• Analysis and interpretation

RCT

Population hierarchy for intervention study

Reference population

Experimental populationExclusion criteriaInformed consent

ExcludedRefused

Study population

Intervention group Control group

Outcome

Losses to follow-up Losses to follow-up

Random allocation

RCT

RCT Advantages (I)

• the “gold standard” of research designs. They thus provide the most convincing evidence of relationship between exposure and effect. Example:

• trials of hormone replacement therapy in menopausal women found no protection for heart disease, contradicting findings of prior observational studies

RCT

RCT Advantages (II)

• Best evidence study design• No inclusion bias (using blinding)• Controlling for possible confounders• Comparable Groups (using randomization)

RCT

Disadvantages

• Large trials (may affect statistical power)• Long term follow-up (possible losses)• Compliance• Expensive• Public health perspective ?• Possible ethical questions

RCT

• Blinding

• Hiding information about the allocated study regimes from key participants in a trialDepending on

outcome of interestEthics, feasibility, compromise

By Using Placebo which is Inert medication, i.e No effect, intended to give the patient the perception they are receiving treatment

Types:1. Single – blind :Observer or subject are kept ignorant about allocated study regime2. Double blind :Both observer and the subject are kept ignorant about allocated study

regime

RCT

THANK YOU