EXPERIMENTAL EPIDEMIOLOGY
▪ Also known as intervention studies
▪ Involve an active attempt to change a
disease determinant (ex. Exposure,
behavior) or the progress of a disease
through treatment
▪ Similar in design to experiments in
other sciences, but subject to additional
constraints
▪ Effects of intervention measured by
comparing outcome in the experimental
group with that in a control group
▪ Ethical considerations are very
important
▪ Informed consent is needed
▪ Randomized controlled trials using
patients as subjects (clinical trials)
▪ Field trials in which the participants are
healthy people
▪ Community trials in which the
participants are the communities
themselves
A. Randomized Controlled Trials
• Studies effects of a particular
intervention
• Random designation of subjects (by
chance) to intervention and control
groups
• Results assessed by comparing
outcomes
• Randomization reduces bias, provides
rigorous tool to examine cause-effect
relationships between an intervention
and outcome (Hariton and Locascio,
2018)
Features of a well-designed RCT (Kendall, 2003)
• The sample to be studied will be
appropriate to the hypothesis being
tested so that any results are
appropriately generalizable.
• There will be effective (concealed)
randomization of the subjects to the
intervention/control groups.
• Both groups will be treated identically
in all respects except for the
intervention being tested, and to this
end patients and investigators will
ideally be blinded to which group an
individual is assigned.
• The investigator assessing outcome will
be blinded to treatment allocation.
• Patients are analysed within the group
to which they were allocated,
irrespective of whether they
experienced the intended intervention
(intention to treat analysis).
• Analysis focuses on testing the research
question that initially led to the trial
B. Field Trials
• Involve people who are healthy, but
presumed to be at risk
• Data collection occurs “in the field”,
usually among non-institutionalized
people
• Logistically complicated and expensive
• Used to evaluate interventions aimed at
reducing exposure without measuring
the occurrence of health effects
• Can be done on a smaller scale and at a
lesser cost – no lengthy follow-up or
measurement of disease outcomes
C. Field Trials • Measurement error

• Treatment groups are communities, not **Sampling error

individuals
• Due to the fact that a small sample does
• Appropriate for diseases influenced by not represent all of the population’s
social conditions, & for which variables
prevention efforts are focused on group
• Proper and unbiased probability
behavior
sampling
Limitations
• Increase sample size
• Only a small number of communities
***Sample size
can be included
• Essential information before calculation
• Random allocation of communities
can be done:
usually not practicable
• Required level of statistical
• Difficult to isolate communities where
significance of the ability to
intervention is taking place from
detect a difference
general social changes that may be
occurring • Acceptable error, or chance of
missing a real effect
• Definitive conclusions about overall
effectiveness of community-wide • Magnitude of the effect under
efforts are not always possible investigation

• Amount of disease in the

population
Potential errors in epidemiological studies
• Relative sizes of the groups
*Random (chance) error
being compared
• Value of sample measurement diverges
• Sample size often determined
(by chance) from that of the true
by logistic and financial
population value; chance difference
considerations
between the observed and true values
of something • Compromise between the 2
must be made
• Causes inaccurate measures of
association • Sample Size Determination in
Health Studies (S.K. Lwanga and
• Can never be completely eliminated
S. Lemeshow, 1991)
since only a sample of the population is
studied • Make sure groups are of
appropriate relative size to
• 3 major sources:
improve precision
• Individual biological variation
• Often an issue in case-control
• Sampling error studies
**Measurement error
• Also called observational error
• Causes the recorded values of variables
to be different from the true ones
• Measurement error can be reduced by:
• enforcing stringent protocols
• making individual
measurements as precise as
possible
*Systematic error (systematic bias)
• Results differ from the true values in a
systematic manner
• “consistent, repeatable error associated
with faulty equipment or flawed
experiment design”
• Study with small systematic error is said
to have a high accuracy
• Accuracy is not affected by sample size
• Principal biases:
• Selection bias
• Measurement (classification)
bias
**Selection bias
• Systematic difference between
characteristics of the people chosen for
a study with those who are not
• Obvious source: participants select
themselves for the study
• There may also be selection bias in
studies of children’s health, where
parental cooperation is required
• All epidemiological study designs need
to account for selection bias
** Measurement (classification) bias
• Occurs when individual measurements
or classifications of disease/exposure
are inaccurate
• Many sources of measurement bias and
their effects have varying importance
• Less chance for systematic
measurement bias if specimens from
control & exposed groups are randomly
analyzed by different labs
• Recall bias: differential recall of
information by cases and controls
• Ex. Cases are more likely to
recall past exposure, especially
if it is known to be associated
with the disease under study
(ex. lack of exercise & heart
disease)
• can exaggerate or
underestimate the degree of
effect associated with exposure
• strategies to reduce recall bias:
• Careful selection of the
research questions
• Choosing an
appropriate data
collection method
• Studying people to
study with new-onset
disease
• Use a prospective
design
• Observer bias: measurements are
influenced if the investigator, lab tech,
or participants knows the exposure
status
• Experimenter or research bias
 • Blind and double-blind studies
• Blind: participants do not know
how they are classified
• Double-blind: both the
investigators and the
participants do not know how
the latter are classified
• strategies to reduce observer bias:
• Ensuring that observers are well
trained
• Screening observers for
potential biases
• Having clear rules and
procedures in place for the
experiment
• Making sure behaviors are
clearly defined
• Setting a time frame for:
collecting data, for the duration
of the experiment, and for
experimental parts
Confounding
• Can occur when another exposure
exists in the study population & is
associated both with the disease and
the exposure being studied
• Problem arises if extra factor is
unequally distributed between the
exposure subgroups
• Occurs when effects of 2 exposures (risk
factors) have not been separated &
analysis concludes that the effect is due
to one variable rather than the other
• Arises because non-random distribution
of risk factors in the source population
also happens in the study population
• Might appear as a bias, but in fact does
not result from systematic error in
research design
• Age, social class are often confounders
*Controlling confounding:
• Design:
• Randomization
• Ideal method for
ensuring potential
confounding variables is
equally distributed
among the groups
being compared
• Sample sizes have to be
sufficiently large
• Avoids the association
between potentially
confounding variables
and the exposure being
considered
• Matching
• Selecting study
participants to ensure
that potential
confounding variables
are evenly distributed
in the 2 groups being
compared
• Used extensively in
case-control studies,
but may cause
problems in controls
selection if matching
criteria are too strict or
too many =
overmatching
• Expensive, time-
consuming
 • Restriction
• Limit the study to
people who have
particular
characteristics
• Analysis:
• Stratification
• Involves measurement
of the strength of
associations in well-
defined and
homogeneous
categories (strata)
• Conceptually simple
and relatively easy to
carry out
• Limited by the size of
the study
• Cannot help to control
many factors
simultaneously
• Statistical modeling
Validity
• Expression of the degree to which a test
is capable of measuring what it is
intended to measure
• Study is valid if its results correspond to
the truth
• No systematic error
• Random error should be as
small as possible
*Internal Validity
• Degree to which the results of an
observation are correct for the
particular group of people being studied
• For a study to be of use, it must be
internally valid
• Threatened by all sources of systematic
error
• Can be improved by good
design and attention to detail
*External Validity
• Generalizability
• Extent to which results of a study apply
to people not in it
• Internal validity is necessary for, but
does not guarantee, external validity;
easier to achieve
• Requires external quality control
measures and judgments about the
degree to which the results can be
concluded
• Assisted by study designs that examine
clearly-stated hypotheses in well-
defined populations
• EV is supported if similar results are
found in studies in other populations
Ethical Issues
• Those involving actions and policies that
are right or wrong, fair or unfair, just or
unjust
• Ethical dilemmas frequently occur in
epidemiology
 • Ethical principles govern the conduct of
epidemiology
• Research and monitoring are important
to ensure that PH interventions do not
have serious unintended and harmful
consequences
• All epidemiological studies must be
reviewed and approved by ethical
review committees
*Ethical principles
1. Informed consent
• PH efforts to modify behavior of
vulnerable people may be unsuccessful
without assuring these people that their
interests will be protected
4. Scientific integrity
• Requires vigilance on the part of ethical
review committees and close attention
to peer review of publications
• Training and mentoring of
epidemiologists must include serious
and repeated discussion of these issues

• Free and voluntary

• Participants retain the right to withdraw

at any time

• Epidemiologists must respect personal

privacy and confidentiality at all times

• Obligated to tell communities what they

are doing and why, and transmit the
results and significance of the studies

• All proposals must be submitted to

properly constituted institutional ethics
committees before beginning research

2. Confidentiality

• Obligated to preserve confidentiality of

information obtained

• Also extends to the right of a person to

withhold information from others

• Permission is required before access to

data is given (medical records, case
registers, other data files & databases)

3. Respect for human rights

• Some argue that individual human

rights are key to preventing infection
because spread of disease was
facilitated by their denial