1

Outline
Introduction
Study Design: Experimental Design
Randomized Controlled Trials: Types
Concepts in RCTs
 Randomization
 Allocation concealment
 Inclusion and exclusion Criteria
 Blinding
 Sample size
 Validity of RCTs
Analysis
• Ethical issues in conducting RCTs: Equipoise

• Study Design: Quasi-Experimental Design(Reading

Assignment)
 3

Introduction: Epidemiologic Activities

Describe the health status of populations by enumerating the
occurrence of diseases, obtaining relative frequencies within groups
and discovering important trends(surveillance).

Explain the etiology of diseases by determining factors that “cause”

specific diseases or trends(Causation)

Predict the number of disease occurrences and the distribution of

health status within populations(Prediction).

Control the distributions of disease in the population by prevention

of new occurrences, eradication of existing cases, prolongation of life
for those with disease, or otherwise improving the health status of
afflicted persons (Prevention and Control).
 4

The relationships?

Exposure(X) Outcomes (Y)

 5

Introduction
Study Designs

Descriptive Analytic
Case
Experimental Observational
report/series
Ecological Case-control
Quasi
Experimental Cohort
Cross-sectional

Nested-case
control
 6

Hierarchies of evidence: Precaution!

Misconception! There is a hierarchy of study
designs; randomized trials provide the greatest
validity, followed by cohort studies, with case–
control studies being least reliable (Rothman,
2014)

RCTs:
 Random and Systematic Error
 7

What constitute an experiment?

In epidemiologic studies, experiment implies the

investigators manipulates the exposure assigned
to participants in the study.
 8

Experimental Study
• A study in which a population is selected for a planned trial
of a regimen/intervention/treatment, whose effects are
measured by comparing the outcomes in the experimental
group versus the outcomes in the control group.

• It differ 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.
 9

RANDOMIZATION outcome
Intervention
no outcome
Experimental Design

Study
population
outcome
Control
no outcome
baseline
future

time
Study begins here (baseline point)
 10

Question (so many questions)?

1. Suppose you want to evaluate the impact of Scholarship for girls in
improving their educational outcomes.
2. Effect of behavioral /nutrition education on improving HIV/STI
preventive knowledge among high school students
3. If the doctor wants you to choose which drug is effective?
4. Policy makers wants to know the effectiveness of productive safety
nets or other social protection program?
5. Do you remember palm oil controversy in Ethiopia?
6. Is Deworming good for children ?
7. Are private schools effective?
8. To Gossip or BCC is important to promote immunization use?
9. Effect of Facebook use on academic performance of MPH students?
10. Are Ethiopian so Jealous/envious?
 11

Did coffee berries cause the goat’s dancing?

 12

Types of trial/Experiment

 Anglo-french (meaning to
T ria l choose, sort, select, or try

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

R a nd o m is ed N o t ran d om is ed

B lind ed N o t b lind ed
 13

Randomized controlled trials (RCTs)

• A study design that randomly assigns participants/subjects into an
“experimental/treatment group” or a “control/comparison group”.

 In a simple experiment, one group receive treatment the other does not.

 The decision about which treatment each participant receives(i.e. the allocation to
intervention or control) is made at random( purely by chance rather than decided by the
doctor or participant
• Help to assess the efficacy OR effectiveness of the treatment/interventions

• In RCT, the two (or more) groups of people in a trial are as similar as
possible with respect to extraneous factors that affect the outcome of
interest, except for the treatment they receive and the expected difference
between the control and experimental groups is the outcome variable
being studied.

• RCT is considered as a “Gold Standard” of research design

 14

What does “controlled” trials imply?

• “Controlled” implies predefined:

• Eligibility criteria
• Specified hypotheses
• Primary and secondary endpoints to address hypotheses
• Methods for enrollment and follow up
• Rigorous monitoring
• Analysis plans and stopping rules
 15

Randomized controlled trials (RCTs)

RCTs aim to find out which treatment is best by making a

fair comparison between:

• A new treatment Vs existing treatment

• Two (or more) existing treatments
• A new treatment and no treatment, or a placebo (where
there is no existing treatment)

RCTs do not necessarily require a “no treatment” control –

randomization can just as easily be used to compare
different versions of the same program, or different
programs trying to tackle the same problem.
 16

Types of Randomized controlled trials (RCTs)

 Prophylactic/preventive trials
 Evaluate efficacy of intervention designed to prevent diseases.

• Example: Vaccine trials, Supplement and Nutrition or Health

education trials

 Treatment trials
 Evaluate efficacy of curative drug or interventions or a drug
interventions designed to manage or mitigate the sign or symptoms of
the diseases.

• Less individual-level self-selection

 17

Types(RCTs)…

• Individual level randomized trials

• Eligible individuals are randomized (conventional medical RCTs
and many behavioral RCTs)
• Self-selection of persons volunteering for enrollment

• Cluster randomized trials

• Instead of randomizing individuals, randomization can be done at
cluster levels, such as villages, or schools, or health clinics or
communities, hospitals, or other aggregates of people (e.g.,
workplace, bars, brothels).
• Clusters (are randomized, and all consenting persons enrolled
Less individual-level self-selection
 18

Why cluster-randomized trials?

• „It address contamination: where treated individuals mix

and chat and potentially “share” treatment with individuals in
the control group.

• Nature of the intervention (e.g., mass media campaign,

population-level interventions)
• Acceptability and reduced stigma (everyone gets the same
treatment within a cluster)
• Can reduce participant self-selection → maximize
generalizability
• Can get data on many nested population subgroups
• Allows assessment of population-level effects (Population
Attributable Fraction)
 19

Cluster randomized controlled trials…

• When randomizing at the cluster level, the unit of

randomization is the unit at which we will randomly roll out
the program; i.e. the cluster (example, a school, village,
hospital ward). The unit of analysis, defined as the unit at
which we will collect data and compare outcomes.

• This distinction of unit of analysis is important when we

calculate the sample size. Because sample size is affected
by the intra-cluster correlation (ICC), which refers to how
similar or dissimilar individuals within a cluster are.

• The ICC will determine how many individuals per cluster,

and how many clusters, you will need to sample.
 20

Trial designs

• „Most trials have two arms (intervention vs control),

• Multiple interventions can also be compared to a single
control arm

• Equivalency trials: head-to-head comparison of two or

more treatments, without a control group (e.g.,
contraceptive trials)
• Factorial designs e.g., intervention A, intervention B,
intervention A+B vs control
 Other types of RCTs:
The Stepped Wedge Design

• One individual/cluster
receives the intervention in
each time period
5

Participants/Clusters
4

• Order of intervention 2

determined at random
1
1 2 3 4 5 6

Time periods

Shaded cells represent intervention periods

• All individuals/clusters get Blank cells represent control periods
Each cell represents a data collection point

the intervention by the end

of the process

• Data collected in each time

period
21
 22

Cross over design

 23

Randomization
 24

What is Randomization?
• Is a method based on chance alone by which eligible participants are
assigned to one of study interventions. Therefore, which treatment
condition/possible assignment is assigned is as a result of equal
chance, and known chance of being selected and which treatment
condition will be assigned is unpredictable.

• Randomization is the best way of ensuring that the results of trials are
not biased by the way participants in each group are selected.

• Everyone has equal probability of receiving the program or not.

• To ensure that randomization has “succeeded”, study groups must be

equivalent in terms of baseline indicators and contextual variables
that might be important.
 25

Why to randomize?
• Avoid selection bias
• Randomization is intended to limit the occurrence of conscious and unconscious bias in
the conduct and interpretation of a trial arising from the influence that the knowledge of
the impending treatment assignment may have on the recruitment and allocation of
subjects.

• Balance intervention groups with known (e.g. age, sex, severity of the diseases) and
unknown baseline characteristics that may influence the outcome. Randomization gives
each participant a known (usually equal) chance of being assigned to any of the groups.

• Allow attribute difference in outcomes to difference in efficacy of the treatment under

study (causality)

• Control for confounding factors

• To approximate a controlled experiment.

• Statistically efficient. Basis for statistical inference‰ i.e. it fulfills assumptions

underlying tests for statistical inference.
 26

Randomization ratio/allocation
1: 1
1:1:1
1:2
1.1.2
 27

Forms of Randomization

• Simple Randomization
• Permuted/Block Randomization
• Stratified Randomization
• Cluster Randomization
• Dynamic (adaptive) random allocation
 28

Simple randomization
• Each participant is randomly assigned to a treatment with known
probability, regardless of the treatment assignment of other
participants.
• How?
• Tossing a Coin?
• Sequence of Random Numbers
• Computer generated sequence

• Example: toss a coin

• H new intervention
• T Placebo control

Limitations
• Significant deviation from equal assignment may happen by chance
• Significant imbalance in important baseline characteristics may happen by
chance
 29

Simple Randomisation: Problems

• Simple randomisation can suffer from ‘chance bias’.

• Chance bias is when randomisation, by chance, results in

groups which are not balanced in important co-variates.

• Less importantly can result in groups that are not evenly

balanced.
 30

Permuted Block Randomization

• When every block of x participants are randomly and
equally assigned to treatment A and B

• In a two group design, Blocks having equal numbers of As

and Bs (A = intervention and B = control, for example) are
used, with the order of treatments within the block being
randomly permuted

• Example take block size of 4 (6 possibilities).

• Used for small studies to maintain reasonably good

balance among groups
 31

Stratified Block Randomization

• Stratification variables may influence the outcomes (e.g. gender,

severity of the diseases, site)

• A set of permuted blocks is generated for each combination of

strata/factors

• Stratified block randomization can further restrict chance imbalances

to ensure the treatment groups are as alike as possible for selected
prognostic variables or other patient factors.

• Stratification simply means having separate block randomisation

schemes for each combination of characteristics (‘stratum’)

• e.g. in a study where you expect treatment effect to differ with age and sex you
may have four strata: male over 65, male under 65, female over 65 and female
under 65
 32

Cluster randomization
• A cluster randomization trial is one in which intact social
units, or cluster of individuals rather than individual
themselves, are randomized into different interventions
group.
• E.g. household, neighborhood, schools, hospital

• Usually done when individual randomization is impractical,

to prevents contaminations or due to convenience and
economical

• Don’t forget to match treatment/intervention group

 33

Dynamic (adaptive)
random allocation
• Dynamic randomization methods allocate patients to treatment
group by checking the allocation of similar patients already
randomized, and allocating the next treatment group "live" to best
balance the treatment groups across all stratification variables.

• Covariate adaptive allocation(minimization)

• Changing the probabilities based on observed prior sequences

by covariates

• Response-adaptive allocation
• Changing the probabilities based on observed prior observed outcomes
 34

Inappropriate randomization
methods
• Assigning patients alternately to treatment group is not random
assignment

• Assigning the first half of the population to one group is not

random assignment

• Assignments by methods based on patient characteristics such as

date of birth, order of entry into the clinic or day of clinic
attendance, are not reliably random

• Examples of non-random allocation:

• Even or odd
• Days of the week
• Morning or afternoon patients
• First patient the day
 35

Allocation Concealment
„ It is very important that those responsible for recruiting people
into a trial are unaware of the group to which a participant will be
allocated, should that subject agree to be in the study.

This avoids both conscious and unconscious selection of patients

into the study. Blinding is done to minimize participant or observer
bias ‰

It is usually possible to identify a staff member not involved with the
trial who can keep the randomization list or envelopes.

They should be instructed to keep the list private, and to only reveal
a treatment allocation after receiving information demonstrating that
the patient is eligible and has consented to the trial.
 36

Randomization
Allocation Concealment
• The sequence of allocation to different groups cannot be
seen by subjects or providers

• Examples:
• Sealed, opaque envelopes
• Central voice-response system
• Online systems
 37

Blinding

• open label: no blinding

• Single blind: patient (usually; occasionally may be assessor) blinded

to treatment. Single blinding (observer but not participant knows the arm of
randomization, e.g., cluster-level trials) ‰

• Double blind: patient and assessors (who often are also the health
care providers and data collectors) blinded to treatment. Double blinding
(neither observer or participant know the arm of randomization)

• Complete blind: everyone involved in the study blinded to treatment.

No one knows the treatment group allocation
• Provider
• Subject
• Outcomes assessor
 38

Issues leading to Blinding

• Most investigators have firm views about which of a

range of alternative treatments is more effective and
often, which is more appropriate for particular groups of
patients. As a result, there is a strong temptation by
investigators to channel particular groups of patients to
particular treatments (channeling effect )

• There is also a risk of the investigators subconsciously

losing their objectivity in their assessments of treatment
effects simply because of their clear preference for
particular treatments

• There is a risk of having other forms of bias, which can

be satisfactorily controlled by proper blinding
POPULATIONS FOR ANALYSIS

Intention to Treat Analysis

All randomized patients are included in final

data analysis. Even if some do not receive the intervention or
drop out before completion of treatment. ‰
• Least biased and most conservative „

Per Protocol Analysis

Only patients who actually complete the trial

according to protocol are analyzed
‰ Potentially biased by selection of the most compliant and often lowest
risk population 39
 40

Do we always need an RCT?

• “Perception that parachutes are a successful

intervention based largely on anecdotal evidence”

• No RCTs identified in systematic review

• Under exceptional circumstances apply common sense
 41

Ethical Issues in RCTs.

When should we do a trial?
Trials must have: a rationale based on prior
observational data or biologic evidence

• an explicit, testable and potentially falsifiable hypothesis

• an uncertainty as to whether an intervention is efficacious
(“equipoise”) rue equipoise provides a rationale for equal
assignment proportions.

• Reasonable expectation that benefits will exceed risk

• An intervention that potentially can be randomized

 42

RCTs may not be possible or practical

• Not ethical/possible to assign intervention
• Cigarette smoking and lung cancer
• H. pylori infection and ulcers
• Impractically large sample size
• Very low-incidence outcome
• e.g., rare side effect of medication
• Impractically long duration
• Outcome requires many years to develop
• e.g., development of cancer
• Known effective treatment
 43

RCTs may not be possible or practical

• Cannot use a placebo (e.g., trials to prevent mother-to-

child HIV transmission).
• Need to provide standard of care
• Personal choice
• Cannot randomize very different interventions. For
example, trials of different types of contraceptive (e.g., pill
vs IUD), are ethically questionable because women have
the right to select a method of their choice
• Can randomize within method type e.g., pill A vs pill B)
„Risks of new treatment likely to exceed risks of existing
treatment
 44

Placebo
• Placebos are inert treatment intended to have no effect
other than the psychological benefit of receiving a
treatment, which it self can have a powerful effect.

• If there is no accepted treatment for the condition being

studied, when ethical constraint allow it. Placebos are not
necessary when the objective of the trial is to solely
compare different treatment with another.

• Equipoise restrict use of placebos. The best available

therapy/interventions should be there.
 45

Informed consent
• When the participant understand
• They are participating in research study of a stated duration
• The purpose of research , the procedure that will be followed and
which procedures are experimental
• The participation is voluntary and it can withdraw anytime
• Potential risk and benefits associated with participation
 46

Coercive inducement and full disclosure „

• Participants should not be coerced by:

• Denying treatment or benefits to persons who refuse (i.e.,
there should be some alternative treatment available)
• By providing excessive compensation for participation
(i.e., money or gifts)
• By applying pressure to participate
• There must be full disclosure of:
• Reason for doing the trial, reason a person was selected
for participation, who is funding the trial
• Procedures entailed (eligibility, randomization, treatments)
• Risks and benefits and measures taken to reduce risks
Confidentiality assurances
 47

Generalizability issues in RCTs

• Validity: In the studied population, was the study
performed in a way that the results are valid?

• Generalizability: Are these results applicable to my

patients?

• Assumptions required to transfer findings from an RCT to

another policy population. i.e. external validity has to do
with whether the result that is established in the study will
be true elsewhere.
 48

Hazards to external validity/generalizability

• Selective enrollment
• Non adherence/non-compliance
• Discrepancy between treatment assigned and actual treatment
received
• Hawthorne and John Henry Effects
• Hawthorne and John Henry effects might occur if the participants in an
RCT know or notice that they are part of an experiment and are under
observation. It is obvious that this could lead to altered behavior in the
treatment group (Hawthorne effect) and/or the control group (John Henry
effect).
• Specific sample problems: Problems that occur when the
treatment in the RCT is provided with special care compared
to how it would be implemented under real-world conditions.
•
 49

Reproducibility in RCTs
• Study design of high methodologic quality
• Minimizes bias: better estimate of “truth”
• Transparent (full and clear) presentation of methods and
analyses
• Enables assessment of methods and results
• Allows duplication of study, re-analysis of results
• Registration before trial begins
• Prevents changing design or pre-specified
outcomes/analyses without explanation
• Trials reported per international standards
• All appropriate elements included
 50

References
Texbook References:
Rothman J, Greenland S. Modern epidemiology.
Third edition. Lippincott - Raven Publishers, 2008.

Additional Readings:
1. Porta M. A dictionary of epidemiology. 5th edition.
Oxford, New York: Oxford University Press, 2008.
2. Bhopal R. Study design. University of Edinburgh.
3. NLM. An introduction to Clinical trials. U.S.
National Library of Medicine, 2004
4. Songer T. Study designs in epidemiological
research. In: South Asian Cardiovascular Research
Methodology Workshop. Aga-Khan and Pittsburgh
universities.