Epidemiological study designs

By: Rahel M (BSc, MPH)

1 Debre Tabor, Ethiopia

 Learning objectives
At the end of this session students able to
o Differentiate the various types of study design

o Discuss the difference between descriptive and analytical studies

o Describe the purpose of descriptive studies and analytical studies

o Discuss each of the study designs

o State strength and limitation of studies designs

o Differentiate how to select study design

2
 What is study design
 A study design is a specific plan or protocol for

conducting the study, which allows the investigator to

translate the conceptual hypothesis into an
operational one.

o The basic design strategies in epidemiologic research are

categorized into two according to their focus of

investigation

3
 Study Designs

Descriptive Analytic

Correlational Cross-sectional Observational Interventional

Case-Report/Series Experimental
Comparative cross-sectional

Case-Control Quasi-
Experimental
Cohort
4
•5
 Choosing type of study design depends on:
 what is the research question/ objective
 Time available for study
 Resources available for the study
 Common/rare disease or production problem
 Type of outcome of interest
 Quality of data from various source
Choosing an established design gives you a huge head start in
design, analysis and eliminating biases

6
 Directionality of the study

 Refers to when the exposure variable is observed

relative in time to when the health outcome is observed

 Can be forward, backward, or non-directional

 Affects the researcher's ability to distinguish

antecedent from consequent

7
 Study design

 In a study with forward directionality, the investigator

starts by determining the exposure status for subjects
selected from some population of interest and then
follows these subjects over time to determine whether
or not they develop the health outcome.
-Cohort studies and clinical trials always have
forward directionality

8
 Study design
Forward directionality

Exposure Outcome/Disease
Time

Yes ?
No ?

 cohort studies
 clinical trials

9
 Study design

 In a backward design, the investigator selects subjects on

the basis of whether or not they have the health outcome
of interest, and then obtains information about their
previous exposures.

-Case-control studies always have backward

directionality

10
 Study design

Backward directionality

Exposure Outcome/Disease
Time
? Yes

No
?
 case-control studies

11
 Study design

 In a non-directional design, the investigator

observes both the study factor and the health
outcome simultaneously, so that neither variable
may be uniquely identified as occurring first.

- A cross-sectional study is always non-

directional

12
 Study design
Non- directionality

Time
Exposure ?

Outcome/Disease ?

 cross-sectional studies

13
 Study designs based on

 Timing concerns the question of whether the health

outcome of interest has already occurred before the
study actually began.
 If the health outcome has occurred before the study
is initiated, the timing is retrospective
-case control study is always retrospective

14
 Study design
Timing
Has the health outcome occurred before the study began?

Time
Exposure Health Study
outcome begins

 case-control study
 Retrospective cohort study

15
 Study design

 On the other hand, the health outcome occurs

after the onset of the study, then the timing is
prospective.

-Clinical trials are always prospective

 The timing of a study can have important

implications for the quality of the data

16
 Study design

Timing
Has the health outcome occurred before the study began?

Time
Exposure Study Health
begins outcome

 prospective cohort study

 clinical trials

17
 Descriptive epidemiological study design

18
 Descriptive study designs
o descriptive study designs are used to characterization of the
distribution of health-related states or events

o It focus on the frequency and distribution of disease

o Descriptive study is one of the basic types of epidemiology

describing the frequency and distribution of diseases by time,

place and person

19
 Descriptive study designs
o Descriptive study designs provides the What, Who, When, How

many andWhere of health-related events

The 5W’s of descriptive epidemiology:

o What = health issue (case definition)

o How many=magnitude of the problem

o Who = person

o Where = place

o When = time

20
 Time
o The occurrence of health outcomes (disease) changes over time

o Some of these changes occur regularly, while others are

unpredictable

o The common time change of diseases are cyclic (periodic) and

sporadic occurrences

 It describes the occurrence of the disease in terms of year,

season, day, date of onset , duration etc.

21
 Place
o Describing the occurrence of disease by place

o provides insight into the geographic extent of the problem and its

geographic variation of disorders

o Characterization by place refers not only to place of residence

o but to any geographic location relevant to disease occurrence

22
 Person
o Personal characteristics may affect illness, organization
and analysis of data by “person” may use:
o Inherent characteristics of people (age, sex, race,
ethnicity )
o Biologic characteristics (immune status)
o Acquired characteristics (marital status)
o Activities (occupation, leisure activities, use of
medications/tobacco/drugs)
o Conditions under which they live (socioeconomic
status, access to medical care)
23
 What?

Cases
Person Time
25

Place 20

15

10

0
1 2 3 4 5 6 7 8 9 10

Who? Where? When?

24
 Descriptive study designs

 Generally descriptive study designs are used

 by health manager to allocate resources and to plan

effectively

 Generate epidemiologic hypothesis

 Inexpensive, less time consuming

25
 While, Analytical study designs
 Analytic studies focus in elucidating the determinants of

disease

 Have comparative groups and apply hypothesis testing

 Data are more likely gathered for specific study.

 Information on outcome of interest is as good as that on

the comparison group.

 Provide information on outcome at individual level.

26
 Categor y of descriptive studies

 If individual as study subjects

o Case report
o Case series
o Cross-sectional survey
 If population as study subjects
o Correlational /ecological studies

27
 Case report
o It is the study of health profile of a single individual using

a careful and detail report by one or more clinicians

o Report is usually documented if there is unusual medical

occurrence, thus it may be first clue for identification of

a new disease occurrences

o It is useful in constructing a natural history of individual

disease

28
 Case series
o Individual case report can be expanded to a case series, which
describes characteristics of a number of patients with the same
diagnosis
o Similar to case report, it is usually made on cases having new or
unusual disease (giving interest to clinicians)
o It is often used to detect the emergence of new disease or an
epidemic occurrences
o E.g. Between Oct 1980 and May 1981, 5 cases of PCP were
reported among young, previously healthy, homosexual
men

29
 Application of case repor t / ser ies

o Useful for the recognition of new disease

o Useful for constructing of the natural history of a

disease

o Enable to know manifestation of new disease

o Use to formulate hypothesis

o detect an epidemic occurrence of a disease

30
 Disadvantage of case report and case series

o Unable to test for statistical association between

exposure and outcome variables
o It is difficult to test for hypothesis because there is no
relevant comparison group
o Rates can not be calculated since the population
corresponding to the source of cases can not be well
defined (no defined denominator)
o Studies are prone to atomistic fallacy (cannot be
inferred to the population)

31
 Ecological studies
o Uses data from entire population to compare disease
frequencies –
o between different groups during the same period of time,
or
o in the same population at different points in time

o The main difference between ecological studies and the

other types of study, ecological studies are carried out at
the population level
o They use aggregate data and do not measure outcomes and
risk factors at individual level(limitation of individual
studies)
32
 Examples
o Circumcision and HIV in Ethiopia
– HIV prevalence of districts in Ethiopia
Vs
– Proportion of male circumcision in the same districts

 Prevalence of depression Vs prevalence of obesity at same

district

o Fluoride content of water and dental caries

– Proportion of people with dental caries in a village
Vs
– Fluoride content of water supply in the village

33
 Characteristic of ecological study
o Measures of association in ecological study is correlation
coefficient (r)

o Correlation coefficient quantifies the extent to which

there is a linear correlation between exposure and
outcome variables (-1< r <1)

Source of data
o It mostly uses secondary data or report from survey data

34
 Fig: Factious data to show correlation between salt
sold and mean diastolic BP (positive r ~ 0.67) –

110
Mean diastolic BP

90

District
70 Linear (District)

50
5 6 7 8 9 10
salt sold (100gms/person/year)

35
 Strength

 Can be done quickly and inexpensively, often using

available data.

 May be best design to study health effects of

environmental exposures,

 questions only sensibly addressed at population (or

community) level

36
 Limitation
1. Inability to link exposure with disease.
- Data on exposure and outcome are not linked at the individual
level;
- Correlation found with aggregate data may not apply to
individuals (this is referred as ecological fallacy)
2. Lack of ability to control for effects of potential confounding
factors
3. It may mask a non-linear relationship between exposure and
disease while it is exist
4. Inability to test hypothesis
5. Roll of chance can not be avoided

37
Cross-sectional study design

38
 Cross-sectional studies
o It is also called prevalence study (survey study)

o It is the major type of descriptive study designs

o Survey is conducted in a population, to find prevalence

of a disease and exposure at a point in time

o Exposure and disease status are assessed simultaneously

among individuals at a point in time

39
Cross-sectional ….
factor present

No Disease
Study factor absent
population
factor present
Disease
factor absent

time

Study only exists at this point in time

 Uses of Cross-Sectional Studies
o It helps administrators in assessing the health status and
health care needs of a population
o Hypothesis generation
o Used to assess prevalence of acute and chronic diseases,
disabilities and utilization of health care resources

o The purpose is for effective health care intervention

planning, priority setting, resource allocation and
administration

41
 Cross -sectional studies...
o Cross-sectional survey could provide information about the
frequency of health conditions by providing a ‘snapshot’ at a
specified time

o In this study, measure of association is made using odds ratio

(OR) i.e. Prevalence ratio

o Prevalence ratio of exposure among diseased to non-diseased or

prevalence ratio of disease among exposed to non-exposed groups

42
 Analysis of cross sectional
 Measure of association is odds ratio

Odds of disease among exposed = odds of exposure among diseased

Odds of disease among non-exposed odds of exposure among non-diseased
OR = ad/bc

43
 Example of cross-sectional study undertaken in Ethiopia

 Census – house hold level

 Ethiopian demographic and health survey(EDHS) –

house hold level

 National immunization survey

44
 Comparative cross-sectional study design

45
 comparative Cross -sectional studies

o Cross-sectional study design can be considered as

analytic study, if it assesses presence of statistical
association between exposure and outcome

o For factors that remain unaltered over-time such as sex,

race, blood group etc...

o It can provide a good evidence in disease causation i.e.

test hypothesis

46
 Comparative cross sectional

• It is a research study being conducted at a point in time

at 2 or more geographical locations or for 2 or more
groups of people etc for comparison purpose

• Example comparing depression among pregnant and

non pregnant women

• Comparing obesity among urban and rural

populations

47
 Advantages cross-sectional study

 Good design for hypothesis generation

 Can estimate overall and specific disease prevalence

 Can study entire populations or a representative sample

 Can estimate exposure proportions in the population –

e.g. proportion of cigarette smokers in Debre tabor town, latrine

utilization proportion

 Can study multiple exposures or multiple outcomes or diseases

 Relatively easy, quick and inexpensive

48
 Limitation of cross-sectional studies

o Since exposure and disease status is assessed at a single point in

time, temporal relationship between exposure and disease can
not be clearly determined(Egg or chicken dilemma”)

o Not good for rare diseases or rare exposures

o The prominent limitation is “Recall of previous exposure may

be difficult
o Do not provide incidence data

49
Analytical study design

50
 Introduction to analytic studies
Application:
 To search for cause - effect relationship and mechanism
o Why?
o How?

 It focuses on determinants of disease by testing hypothesis

regarding exposure and outcome of interest
o Proof versus sufficient evidence?

 To quantify the association between exposure and outcome

of interest

51
 Analytic studies...

Basic features:
 Key feature of analytic epidemiology is comparison group
 Appropriate comparison group needed:
o Exposed versus non-exposed
o Case versus control
o Experimental versus non-experimental
 It is the use of comparison group that allows testing of
epidemiologic hypotheses
 Provide information on outcome at individual level

52
 Two types of analytic studies:

o Observational studies

o Interventional studies

53
 Observational studies
o An investigator observes the natural course of an event

o An investigator measures but does not intervene

Interventional studies
o An investigator assigns study subjects to exposed and non-
exposed and follows to measure for disease occurrence

o An investigator manipulates the intervention

54
 Observational analytical study designs

Two basic observational analytic studies:

o Case-control studies

o Cohort studies

55
Case-Control Studies

56
 Definition of case-control studies
 A case control study is one in which persons with a condition

(“cases”) and suitable comparison subjects (“controls”) are

identified, and then the two groups are compared with respect to
prior exposure to risk factors

 Subjects are sampled by their outcome status

57
 case-control studies
o Compares one group among whom a problem is present

with another group where the problem is absent in order to

find out factors contributing to the problem

o Problem examples- malnutrition, lung cancer, contracting

cholera, neonatal death

58
 case-control studies

Case-control study examines the association

between disease and potential risk factors by
taking:

o Case group or series of patients who have a disease of

interest
o Control (comparison) group of individuals without the
disease are selected for investigation and
o then proportions with the exposure of interest in each
group are compared

59
 case-control studies..

o The investigator looks back in time to measure exposure of

the study subjects to the risk factors

o The exposure to the risk factors is then compared among

cases and controls

o To determine if the exposure to the risk factors could

account for the health condition of the cases

60
 • the starting point is selection of cases and controls
then exposure history of cases and controls is assessed
the direction of inquiry is always backwards in time

Direction of inquiry
Exposed
Cases
Non-exposed
Population
Exposed Controls
Non-exposed

61
Time
 Key activities in the design and conduct of case-control
studies

 Selection of cases and controls

 Sample size determination

 Exposure assessment

 Analysis of data

62
 Steps in case control study
1. Define cases and controls
2. Identify group of cases
3. Identify group of controls
4. Ask (review) both groups for previous history of exposure
to risk factors under study
5. Measure frequency of exposure to risk factors occurrence
in both groups
6. Compare frequency of exposure to risk factors between
cases and controls
7. Conclude that previous history of exposure to risk factors
contributed for the cases more than controls or not

63
 Application of case control study design

o It is good to do for rare diseases

o It may be possible to explore a wide range of potential
exposures to risk factors for a single disease

64
 What are cases
o Establish a clear operational definition or use standard
definition of disease (outcome) of interest in order to have a
clear understanding of exposure-disease association
o It is the outcome of interest under-study
o It can be:
– A disease
E.g. HIV status, malaria case
– A behavior
E.g. Alcohol drinking habit, cigarette smoking
– Occurrence of an event
E.g. migration

65
 Selection of cases
o Define ‘disease’ and how it will be measured
o Selecting the source population of cases
(homogeneous cases)
o Sources of cases are commonly:

– All persons with the disease in a population during a

specific time of period
– All persons with the disease seen at a particular
facility (e.g. a hospital) in a specific time period

66
 Source of cases
Hospital-based:
o easy and in-expensive to conduct
o it is prone for selection bias
Population-based:
o avoids selection bias
o allows the description of a disease in the entire population
and the direct computation of rates of disease in exposed and
non-exposed groups

67
 What are controls

o It is the comparison group (referent)

o It should be free of the disease (outcome of interest

under study)

o It should be as similar as the cases in all aspects

except for the disease of interest under study
68
 What are controls
o Controls must have the same opportunity of getting
exposure to risk factors as cases and should be subjected to
the same inclusion and exclusion criteria

o Practically, no one control group is optimal for all situations

comparable with cases

o It needs scientific, economic and practical considerations

69
 Selection of controls
o Controls should be selected from the same study base (target
population) as cases

o Should be selected independently of their exposure status to

the primary risk factors in question

o If they had developed illness, they should be excluded or should

be considered as cases

o Comparable information should be obtained from controls as it

is from cases

70
 Source of controls
1. Population-based controls

2. Hospital-based (health institution) controls

3. Friends/relatives
4. Neighborhood
5. Dead control
6. Hospital visitors
7. Accident victim’s

71
 Ratio of controls to cases
o A single control group is optimal in most of the times

o However, ratio of controls to cases may vary from 1:1 to

4:1

o Decision based on power calculation

o More than one control groups may be recommended when

the single control is not appropriate or has a specific
deficiency

72
 Assessment of exposure status

 the method used to measure exposure must be identical in

both cases and controls to avoid the risk of bias
 As much as possible the data collectors should be blinded
to outcome status of participants in order to minimize bias
 that could result from intensively looking for history of exposure
among cases

 Whenever possible it is valuable to record the time of

exposure and outcome

73
 Analysis
 Presence and strength of association in a sample

 OR: odds of exposure in cases divided by odds of

exposure in controls

 Statistical significance of association

 p-value, CIs

 Statistical analysis – chi-square, logistic regression

74
 Strength of case control studies
o To investigate rare disease (less than 10%)

o Can examine multiple etiologic exposures for single

outcome

o Suitable for the evaluation of diseases with long latency

period

o Quick with time and in-expensive

o Relatively efficient with small sample size comparing with

cohort studies

75
 Limitation of case-control studies
o In-efficient for rare exposures

o No calculation of rates and risks possible

o Prone to selection and information bias

o Affected by inaccuracy and incompleteness of records

o Affected by recall bias

o Difficulty in selecting controls

o OR exaggerates risk

76
 Cohort study design

77
Cohort

 Cohort is a group of persons with common characteristics,

usually an exposure or involvement in a defined population
who are followed or traced over a period of time.
 It can be
 a community cohort of specific age and sex,
 an occupational cohort ( miners, military personnel, a
diagnosed ) or
 a birth cohort (school entrants, marriage cohort,
immigration cohort or
 treatment cohort ( cases treated with radiotherapy,
surgery, hormonal treatment).
 Cohort study
Cohort study

 Cohort study begins with a cohort initially free of disease

(outcome of interest) .

 classified according to a given exposure and then followed

(traced) over time to see how their exposures affect their

outcome of interest.

Subjects are sampled by their exposure status

79
 Design of cohort study

80
 Types of cohort study….

 Cohort studies can be either

– prospective

– retrospective

81
 Prospective cohort study

 The investigator collects information on the exposure status

of the cohort members at the time the study begins, and
identifies new cases of disease (or deaths) from that time
forward
 The exposures may have occurred at the beginning of study
 But the outcome has certainly not yet occurred
 After the selection of the cohort, participants must be
followed over time to assess incidence of disease

82
 Prospective cohort study…

 Exposure status determined at present.

 Study groups followed up and disease outcome will be
ascertained in the future.

83
 Prospective cohort study…

84
 Prospective cohort study……..

Prospective cohort study can be;

85
 Basic Elements of Cohort Study;

 Disease free population at entry of the cohort

 Selection by exposure status rather than outcome

 Follow up is needed to determine the incidence of the

outcome .
 Compares incidence rates among exposed against non

exposed groups.

Analyze data: compare incidence rates in the different

exposure groups (exposed and non exposed )

-Measure of association in cohort studies: relative risk

86
 Exposure of interest

It can be;
Exposure in usual sense
E.g
. ingestion of contaminated food
. droplets from someone with active pulmonary
tuberculosis
Behaviors
E.g. sharing needles, drinking alcohol, fatty food eating
habit etc.
Treatment
E.g. intervention - education program

87
 2 by 2 table and measure of association in cohort study

88
 Example for cohort study
 Researchers conducted a 20 year prospective cohort
study to assess a possible association between eating fish
and heart disease. They found that, of the 200 subjects
who ate fish, 15 developed heart disease while of the
400 subjects who did not eat fish, 20 developed heart
disease.
a) Draw 2-by-2 table?
b) Calculate RR?

89
 Steps in cohort studies

1. Define the population at risk

2. Determine exposure status to a factor of interest
3. Make sure that study subjects are free of the disease of
interest at time of enrolment
4. Follow exposed and non-exposed forward in time to
ascertain whether they develop the outcome of interest
5. Compare the outcomes in the exposed and non-exposed
groups
6. Conclude whether the exposure to risk factor contributes
for the outcome of interest

90
 Data collection and follow up
 Data on the exposure of interest to the study hypotheses
Exposure is any risk factor that can associated with the occurrence of
that disease
 Data on the outcome of interest to the study hypotheses
- Development of the disease of interest specific to the exposure
o Characteristics of the cohort that might confound the association
under study
o Other socio demographic variables like age, sex, income----
 Follow up
The follow-up is the most critical and demanding part of a cohort study
o Lost to follow-up should be kept to an absolute minimum (< 10-15%)
91
Possible outcomes in prospective cohort studies

Possible outcomes in prospective cohort study

 No disease
 Disease present
 Computing risk
 Lost to follow up
 Sources of exposure information

o Pre-existing records(medical & employment

records)

o Information supplied by the study subjects

o Direct physical examination or screening tests

o Other sources like measurement of air or

water(environmental monitoring)

93
 Sources of outcome information

 Depend on the specific resources available as well as the

particular disease under investigation
o Surveillance data
o Death certificates
o Medical records
o Directly from the study subjects
o Periodic direct medical examinations of the study subjects

94
 Ascertainment of outcome of interest

The aim of good outcome ascertainment is to ensure

that the process of finding cases, whether deaths, illness
episodes, or people with a characteristic, is as complete
as possible using:
o Operational definition
o Standard definition
o Eligibility criteria
o Case definition
o Case finding
o Case ascertainment

95
 Ascertainment of outcome of interest…

o Define the outcomes of interest (set diagnostic criteria)

 each multiple outcome must be defined

o End-points must be ascertained in a similar manner for

both the exposed and non-exposed groups

o Any outcome measurement should be done equally both

to the exposed and non-exposed groups.

96
 Analysis of cohort studies

o The primary objective of the analysis of cohort study data is to

compare disease occurrence in the exposed and unexposed
groups
o It is a direct measurement of a risk to develop the outcome of
interest
o Calculation and comparison of rates of the incidence of the
outcome for exposed and non-exposed subjects using rate ratio
(relative rate) , risk ratio or relative risk (RR) as measure of
association
o Control of confounding

RR= incidence of a disease among exposed (a/(a+b))

incidence of a disease among non-exposed (c/(c+d))

97
 Analysis …

Data analyzed in terms of

 Incidence rate of outcome among exposed and non exposed
 Estimation of risk,
Risk Ratio = (CIe) / (CIu)---- closed cohort study
Rate Ratio = (IRe) / (IRIu)-----dynamic(person time year)
we can also measure absolute risk
AR,ARP,PAR and PARP

98
Strength for cohort
o Particularly efficient when exposure is rare

o Can examine multiple effects of a single exposure

o Used to calculate direct measurement of incidence (risk).

o Allows for accurate measurement of exposure variables

o Can elucidate temporal relationship between exposure and

outcome of interest.
 Eliminates recall bias and adjust for confounding variable

99
 Limitation

 Expensive

 Time consuming

 Cannot be applied for diseases with long incubation

period

 Relatively poor for studying rare diseases or outcomes

 Selection bias

100
 Interventional study designs

101
 Definition of interventional studies
• Assess/evaluate performance of an
intervention/treatment
– by allocating intervention to one or more groups and
comparing those who get and don’t get the intervention

• Intervention – action/measure undertaken to avert

adverse health outcomes
– therapeutic – drugs,
– Preventive–vaccines, health education,…

102
 Interventional studies
• The design starts with identification of study population
who are grouped into intervention and comparison
group

• Then, the investigator allocates the intervention to be

evaluated to the intervention group but not to
comparison/control group

• Finally the rate of outcome of interest is compared

between intervention and control group

103
 RANDOMIZATION outcome
Intervention
no outcome
Study
population
outcome
Control
no outcome

baseline
future

time
104 Study begins here (baseline point)
 Key Features of Experimental Design
 Investigator manipulates the condition under

study
 Always prospective

105
 Design Similarity with Observational Studies

• Individuals are enrolled based on their exposure like

cohort studies
• Usually grouped into exposure and control group
• Then followed-up for development of outcomes

106
 Design Difference with Observational Studies
 The investigator determines who is exposed i.e.
allocates an intervention
 unlike in observational studies where the investigator
simply observes the occurrence of events
 The design is controlled by the investigator
 The method of allocation of intervention is mostly is
through randomization

107
 Other Differences with Observational Studies

• Interventional studies are the strongest design for

assessing cause-effect relationship
– Due to design characteristics which minimize confounding
and bias
• Ethics and cost
– Cost – costly due to huge expenses
– Interventions, follow-up, investigations, skilled manpower

108
 Ethical concerns
 The intervention should be beneficial
 The intervention should not be harmful
– Precludes the direct evaluation of harmful agents

• All groups should be treated equally

– There must be sufficient uncertainty about the
intervention’s effectiveness
– At the same time there must be sufficient evidence about
its effectiveness

109
 Stages in interventional studies

• Due to ethical concerns most interventional

studies are conducted in phases
• Phases
– I – starting from experimentation with animals and
in laboratory
– II – then with healthy individuals
– III – then with volunteer patients
– IV – finally trial is conducted

110
 Types of interventional study designs
• Experimental interventional based on population
• Clinical trials (therapeutic trials) usually performed in
clinical setting and the subjects are patients

• Field trials (preventive trials) used in testing medicine for

preventive purpose and the subjects are healthy people

• Community trials (preventive or therapeutic trials where

unit of randomization is group of individuals

111
 Clinical Trials
• Assess efficacy of therapeutic interventions
– Therapeutic interventions include drugs, surgery,
physiotherapy
– Patients are the study subjects

• Clinical trials could aim at cure, prevention of

disability or preventing recurrence
• E.g. trial of anti-malarial, anti-diabetics agents

112
 Field Trials
• Assess effectiveness of preventive interventions
– Aim at preventing occurrence of disease
– Include vaccines, health education, sanitary interventions,…
– Deal with individuals who have not developed the disease
– E.g. trial of vaccine against cholera

• Similarity with clinical trials

– Unit of randomization is an individual not group unlike
community randomized trials

113
 Community Trial:
– Unit of the study is the community
– Communities as study subjects
– Health promotion (preventive interventions) are
used as an exposure
– Occurrence of new disease is the outcome
E.g. fluoridation of water to prevent dental caries

114
 Based on design
 Uncontrolled trial - no control group.

control will be past experience (history).

 Non-randomized controlled - there is control group

but allocation into either group is not randomized.

 Randomized controlled - there is control group and

allocation into either group is randomized.

115
 Based on trial objective:
Phase I: Clinical Pharmacology and Toxicity

o This trial provides metabolic and pharmacologic profiles of the

drug
o Trial on small subjects to test a new drug with small dosage to
determine the toxic effect
o The metabolism of the drug may also be investigated at this stage
o The objective is to determine a safe drug dose for further studies
of therapeutic efficacy
o Design is dose-escalation to establish a maximum tolerated dose
for a new drug
o Subjects: 10-20 normal volunteers or patients with disease

116
 Phase II: Clinical Investigation of ‘Treatment
Effect’

– Trial on small group to determine the therapeutic

effect

– The purpose is to assess the efficacy of the drug or

device, to determine the appropriate dosage , and

– to investigate its safety and dose response

relationship
117
 Clinical Investigation for Treatment
Effect…

Objective:

o To get preliminary information on effectiveness and

safety of the drug

– Design: Often single arm (no control group)

– Subjects: <100 patients with disease

118
 Phase III: Full-Scale Evaluation of the Treatment
o Study on large population to test effectiveness (RCT)
o It is most rigorous and extensive type of scientific clinical
investigation of a new treatment
o The drug’s indications of use, recommended doses, and side effects
o This one is classical phase and the purpose is to assess the
effectiveness
o The drug manufacturer can request to market the drug for the
indication covered by the trial
o Confirmatory trial
o 100 -1000 study subjects enrolled

119
 Phase IV: Post Market Surveillance

o The purpose of this trial is to re-assess the effectiveness, safety,

acceptability and continued use of the drugs
o Post marketing surveillance may be conducted to determine
long- term safety and efficacy of the drug

– Objective: To get more information on long-term side effects

– Design: no control group

– Subjects: patients with disease using the treatment

120
 Conduct of individual randomized controlled trials

• Selection of study population

• Sample size determination
• Allocation of treatment regimen
• Follow-up
• Ascertainment of outcomes

121
 The gold standard features of RCT
Randomization
Blinding
placebo
 Randomization
 Randomization-allocation to treatment groups
by chance mechanism
 Each individual has equal chance of receiving the
alternative treatments
 Purpose- promotes comparability between
groups

122
 Blinding(masking
 A blinded study (Masked study) is a study in which
observer(s) and/or subjects are kept ignorant of the group to
which the subjects are assigned
• Blinding-keeping someone unaware of which treatment has
been given
 A clinical trial should, ideally, have a double-blind design to
avoid potential problems of bias during data collection and
assessment.
 Helps to reduce or eliminate expectation bias on the part of
the subject and observer bias on the part of the researcher

123
 Types of Blinded Studies
1. Single Blind
 The study subjects doesn't know to which group they are assigned and used
to avoid placebo effect (psychological component of offering treatment)
2. Double Blind
 Neither the study subjects nor the data collector know the group to which
the subject has been assigned
 to avoid bias in selection
3. Triple blind
 The study subjects, the data collector and the individual who is doing the
analysis are ignorant of the group to which subjects belong
 to prevent misdiagnosis by the assessor-especially if outcome assessment is
subjective e.g. - clinical diagnosis)

124
 placebo
 A placebo is a biologically inactive substance given to the control group so
that they think they are being treated equally as treatment group

 A biologically inert intervention which is used to elicit any non-

specific psychological effects of the test intervention

 The placebo should be similar to the drug being tested in respect to

appearance (size, texture, odor, color and taste)

 May not be ethical if an intervention of known effectiveness is

already in general use

125
 Quasi-experimental
• are trials in which randomization and/or control
group is missing

• No randomization - two comparison groups in which

one serves as control but no random allocation of
subjects to comparison group

• No control group - Before and after designs

(pretest-posttest) studies

126
 Features of quality interventional
studies(strength)
Randomization minimizes selection bias and
confounding factors
Blinding minimizes observer bias
Placebo could blind patients
The best type (gold standard)
Reduce variation by extraneous factors other than the
factors under study
• Strongest for assessing cause-effect relationship

127
 Limitations of interventional studies:

o Ethical issues in question

o Feasibility issues

o Cost implications

o Loss to follow up and drop outs issue

128
Hierarchy of Epidemiologic Study Design

129
 !!!

130