Experimental

studies

Study designed to test the

Efficacy or Efficiency of a
particular agent.
 Effica Efficien
cy cy

Study of particular
Study of particular
agent that is
agent that is
whether work or not
whether work or not
under standardized
under everyday
factors(Age, SES,
geographic location,
conditions (normal
life conditions) with
race). varying degrees of
the disease.
 Experimental
studies

 Are carried out among

patients, frequently  Are carried out with
hospital in – patients. people in the community
( Therapeutic T, Intervention who may not necessarily
T, and Preventive T) be patients ( DHE project,
Xylitol regimen for
 Consider the gold
expecting mothers)
standard for causality
Clinical trial :-
• Is a controlled experimental
study of group of comparison.

• It designed to test the

hypothesis that certain agent or
regimen favorably alter the
natural history of a disease
experience.
• It compare two or more segments of a
single population.
• Essentially they must be similar in
distribution of age, gender, race, SES, and
disease
• The group receiving the agent or regimen
under study is the TEST group.
•
• The comparable group which not subjected to the agent or
regimen is the CONTROL group.

• Factors affecting the validity of the trial:-

• Random assignment to study and control groups will give
a confidence that the groups are comparable, which
prevent bias of data.
• Bias :- Errors and mistakes that can happen during the
study by :-
• The investigator.
• The participants.
• The assessment and outcome.
• Provide the control group with placebo, which is a
substance similar in appearance and all properties to that
being tested except in respect to the active ingredient.
• Apply the concept of blind or double blind assessment,
nowadays tertiary blind.
• Adequate number of subjects for the attrition, least
researcher will fail to show by statistical tests whether a
difference exists between the group.
• Reliability in diagnosis can be achieve by the inter & intra
examiner reliability tests
•
• Clinical trial must be continued long enough to permit
detection of new disease or extension of lesion already
present during the trial.

• For caries trial, minimum duration 2 years.

• For plaque inhibiting agents trial, 8-21 days.
• To demonstrate gingivitis reduction, 6 month
• For calculus preventing agents should last 90 days for
supra gingival calculus & longer for sub gingival calculus.
Stages in testing New Therapies:

• Preclinical: studies in cell cultures & animals.

• Phase 1: unblinded, uncontrolled studies in a few volunteers to test safety.

• Phase 11: relatively small RCT to test tolerability and different intensity or
dose of the intervention on clinical outcomes.
• Phase 111: relatively large RCT to test the effect of the therapy on clinical
outcomes.
• Phase 1v: large trials studies conducted after the therapy has been approved
by the FDA to assess the rate of serious side effect and evaluate additional
therapeutic uses.
 The systematic tendency of any factors
associated with ( the design, conduct,
analysis, evaluation and interpretation of the
results ) to make the estimate of the effect of
intervention deviate from its true value

Anthony J. Viera, Shrikant I. Bangdiwala, Eliminating Bias in Randomized Controlled

Trials:Importance of Allocation Concealment and Masking. Fam Med 2007;39(2):132-7
 Comparison in observational studies
conducted between independent groups,
while in RCT, it is done between
dependent groups (allocated randomly)
to eliminate the differences.
Also, if sample size is sufficient, all factors
influential in the outcome are likely to be
distributed equally between the groups which were
randomly allocated.

Therefore, any difference in the observed outcome

between the groups is likely to be due to the
intervention.
 How
should • Research question
an RCT formulation.
• The key component of a
be sound Research question
designe should include:
d?
• In brief, question need to be well
defined, clear, feasible, specific,
measurable, ethical and clinically
important before initiating the
experiment.
 A hypothesis has to be formed
for Research question to be
answered using an RCT , and to
be precise

Null Alternati
Hypothesi ve
s Hypothes
is
 The Null
Hypothesis
Is a statement that there is no actual
relationship between variables, and that
any such observed relationship is only a
function of chance, or sampling
fluctuation.

It is stated as a formal
proposition, using the
following symbols :
H0 : UA = UB
 Alternative
hypothesis
My also be stated in similar
terms :
HR : UA differ UB

Although null hypotheses are either not

reject or reject on the basis of data from
a sample, the hypothesis is made about
population values.
Experimental studies
Clinical Trial

Trial with independent control.

• Randomized ( RCCT)
• Pseudorandomized.
 Experimental studies
Clinical Trial
Trials with
self control
• Before and after.
• Split mouth technique.
Cross – over design
 Crossover study design

• Crossover designs are often applied to

chronic conditions (palliative rather than
curative) such as recurrent gingivitis, where
outcome is measured in terms of short term
relief of signs and symptoms.

• Cross over design is more vulnerable to

drop out and attrition bias
• First, a pre- randomization run- in period
is usually required. During this period the
patient commonly receives a placebo
treatment, such as standard

• Prophylaxis and oral hygiene instruction,

and the condition of the entire study sample
is monitored until the physiological condition
of each patient has reached a stable
baseline.
 • Each patient is scheduled to receive a
sequence of both the experimental and
control treatments in opposite order, thus
signifying the cross over design.

• Second feature is that a wash-out period

is needed after the treatment has been
stopped to allow each patient physiological
condition to return to its baseline level.
• Long enough, as therapeutic effect of the
1st TTT does not carry over into the 2nd
TTT time period.
 Experimental studies
Clinical Trial
Trial with external
control

Using results of another

investigator as a comparison.
 Factorial design
Aims to answer two or more separate
research questions in single cohort of
participants.
• e.g effect of low dose of aspirin, vit E on
risk of cardiovascular events on healthy
women
The Key components of RCT design are:

 Random allocation
 Blinding
 Allocation Concealment
 Training and calibration
 Ethical considerations
 1-Random allocation
Each of the eligible
participant should have an
equal chance to be allocated
the intervention or No,
which will be achieved by In Cross-over design, all
parallel group design. participants receive both
intervention in
sequential manner and
the order of intervention
is randomly assigned,
thereby eliminating
individual participant
differences.
 Generation of random
sequence should be done
by some independent
personnel, usually a
statistician.
 RA does not mean random selection
of patients to include in the trial

 Random assignment of intervention is

done after subjects have been
assessed for eligibility and recruited,
but before the intervention begin.
 Why do we
use random
allocation?

Random allocation means that the two

study groups will end up comparable in
terms of factors such as age, sex, and even
other things that you do not know about
(such as their reaction to intervention).
 N.B: Randomization

Random selection of a
sample ensures that the
sample is representative
of the broad population;
it is typically used in an
observational study
 "RCCT" versus RCT

“RCCT” used only for trials that contain

control groups, in which group receiving the
experimental treatment are compared with
control group receiving previously tested
treatment (positive-control study)
, or Placebo (placebo-controlled study) or
Negative Control
 RCT "randomized clinical trials"
omits mention of controls and can
describe studies that compare multiple
treatment groups with each other
( comparative design )
(in the absence of a control group)
 2- Blinding
• The three types of blinding helps to
eliminate the unconscious information
bias.
• It is not always possible to blind either
the participants or investigators due to
the nature of the RCT, so the trial is
conducted as an Open-label RCT.
Everybody
knows which
intervention is
given to each
participant
Patient or
evaluator is
blinded as to
treatment,
but not both
Both patient
& investigator
are blind to
treatment
assigned
Patient,
investigator, &
Data analyst are
blinded to
treatment
assigned
  Neither investigators nor
participants are aware of whether
the next eligible participant will be
receiving treatment or control
intervention.
 This should be masked until the
time when participants are ready to
3- Allocation receive intervention, so
concealment unnecessary adjustments can be
avoided.
 This is very important in situation
when blinding of intervention is not
possible.
  Holding translucent
envelops up to bright light.
 Opening a well sealed
opaque envelop.
 Determining different
weights of the assignment
envelops.

Anthony J. Viera, Shrikant I. Bangdiwala, Eliminating Bias in Randomized Controlled Trials:Importance of

Allocation Concealment and Masking. Fam Med 2007;39(2):132-7
 Training on:
 The indices & how to
conduct examination
Calibration on indices to:
 Increase validity (success
6-Training in measuring what you
set out to measure)
&  Increase intra-reliability
Calibration (reproducibility within
examiner)
 Increase inter-reliability
(reproducibility between
examiners)
 • An interval of at least
a few days is
desirable between
training and
calibration to allow
Training & the examiner time to
Calibration assimilate knowledge
of the assessment
method and to
practice the
procedures.
 7- Ethical
considerations

 Rigorous ethical principles must be applied

to all RCTs involving experiments on animals,
humans or human biological material.

 Evaluation of risk and benefit to the

participants and society, obtaining ethical
approval and informed consent are crucial.
Where there is previous evidence
showing superiority of an intervention
over that of doing nothing, an RCT using
a placebo (or doing nothing) is unethical.
(Water fluoridation =Quasi)
 Research

• A scientific effort to find

solution to a problem.
• If there is no problem, research
is not needed.
 Research problem

When there is a discrepancy between what is

and what should be ; i.e. between a present non
satisfactory situation and a future ideal satisfactory
one.
Research problem could be

A basic science problem ( biochemical,

bacteriological, biomaterial…..etc)
It could be a clinical problem
(diagnostic test, treatment
The solution
efficacy, maymanagement
patient become in the
or
future
health acare
nondelivery
satisfactory
system
situation
problem ). and research will again
be needed.
• On handling clinical research, there are two view points that run
together :

• Anatomy of the research.

• “ What it’s made of “….. This includes the elements of the research
plan :
• Research question, Design, Subjects, Measurements, and Sample size
calculation.
• An investigator’s goal is to create these elements in a form that will
make the project feasible, efficient, and cost effective.
• Physiology of the research:
• “ How it work “……. What happened in the study sample, how these
findings generalize to people outside the study.
• An investigator goal is to minimize the errors.

Separating the tow themes will clarifies our thinking about a complex
topic.
Research Question
• All studies should start with a research question that addresses what
the investigator would like to know.

• A good R . Q arise from medical articles, conferences, and from

critical thinking about clinical practices and problems.
1-Research Question
1-Research Question
Research Question
 It usually begins with a general concern.
 The question should be focused before planning the study
 This is better be done by breaking the question into more
specific questions
 Then we select one or two to build the protocol
 This is the Aim/aims of the study
This question …..
1. Leads on to inclusion and exclusion criteria.
2. Helps build up a search method.
3. Helps thinking about what data to abstract
FORMULATION OF HYPOTHESIS
FORMULATION OF HYPOTHESIS

• Based on your initial observation and gathered information,

you formulate your own hypothesis.
• A hypothesis is a proposed solution to the problem.
• It is a prediction that can be tested through experiment.
• Because it is a speculation It may start with the word
“If…then” format.
• For example, “if teeth are treated with Brand X sealant, then they
will be more caries resistant than the teeth without”. It named as “
theoretical hypothesis “
Types of Hypotheses
 Study examining 5 years tooth loss for 1000 individuals after
endodontic treatment comparing when a crown has been placed
or not.
 In exposure treatment:
 Yes = received crown……No = not received.
 In outcome:
 Yes = loss of tooth…….No = not loss tooth.
 Tooth loss Tooth loss

Crown placed yes No total

yes A 50 B 600 A+B = 650

No C 250 D 100 C+D = 350

total A+C = 300 B+D = 700 1000

RELATIVE RISK (RR):

 Indicate likelihood that someone exposed to a risk factor/treatment

will develop the disease/benefit as compared with one who has not
been exposed.
 This expressed as EER/CER.
 Experimental Event Rate/ Control Event Rate.
 (A/A+B) / (C/C+D).
 (50/650) /(250/350)= 0.08/.71= 11.3= RR.
 If RR = 1 means there is no apparent effect on risk or benefit at all.
 If RR > 1 means a person is estimated to be at an increased risk or
benefit.
 If RR < 1 means the person may be at decreased risk or benefit.
 In our example the RR of tooth loss is 11% for those who have
received a crown.
Odd RATIO: ( OR ).

 Is the proportion of patients with the target event divided by the

proportion without the event.
 (A/B) /(C/D)………..OR AD/BC = 3.3 in the example.
 This means that the odd of loosing a tooth are more than three times
greater for those who do not receive crown than that who do
receive after endodontic treatment.
 OR = 1 indicates the effect of treatment are not different than control
.
 OR > 1 indicates the effect of the treatment are better than the effect
of control.
 OR < 1 indicates the effect of the treatment are opposite.
Three kinds of papers
published in journals :-
• \primary research….reports of
original clinical basic or
epidemiological research.

Secondary research…..summarize
knowledge in a particular areas.

Commentaries ,where some

documented facts are used as basis
for program development.
Secondary Scientific
Papers :-
• Is a review paper which summarize other paper.
• 2 types of review :-
• 1- Narrative review in which the studies reviewed
have not been identify or analyzed.
• 2- systematic review, it contain explicit statement
or objectives with spelt out research question.
• The data source are stated as well as the methods
of selection.
• Research conducted according to an explicit
methodology.
• A systematic review includes
• Abstract .
• Introduction :- A well clear systematic
review answers a questions or closely
related questions, which should be made
clear at the beginning of the review.
• Methods :- Should fully describe the
methods used for locating , selecting ,
extracting and synthesizing the data.
• Body of the review :- The sequence should
have logical basis. The argument should e
critical.
Meta – analysis :-

• A special type of systematic review

that combines results from more
than one investigation to obtain a
weighted average of the effect of a
variable or intervention on a defined
outcome.
• Combining data increases sample
size and the power of the study to
provide statistically significant
conclusions.
Study design :-
• Health research studies may be broadly
divided as follows :-
Non intervention studies ,where the
researcher just describes and analyzes
researchable objects or situations but
does not intervene.
Intervention studies ,where the researcher
manipulates objects or situations ,either
by preventive or therapeutic measures
,and determine the outcome of his
manipulation.
Case report

Case series
• Descriptive studies :-
• Involves the systemic collection and
presentation of data to give a clear picture of
a particular situation. Involve 3 types
• Descriptive case studies :
• In-depth description of the characteristics of
one or limited number of cases (patients
,health center ,village ,community).
• Cross –sectional studies :-
• Quantification of the distribution of
certain variables in the study population
at one point of time.
• It may quantify the demographic
variables (age ,education ,gender..ect)
and the events that has occurred in the
population.
• Such study called prevalence study.
 3- Cross sectional Studies

Advantages
• Relatively easy Disadvantages:
• Inexpensive to • Establish an
carry out association, not a
cause and effect
• Ethically
relationship.
acceptable,
• Longitudinal studies :-
• Is one in which the same group of people
is studied on two or more occasions to
determine the progress of the condition.
• Such study also call incidence study.
• Dental caries increment could be
determine through that study.
Causality & Risk

• Causality meaning that certain

exposure results in a particular
outcome. Analytical studies have the
general aim of seeking out cause
and effect association.
• It cannot directly adresses cause
and effect, they seek to quantify the
degree of disease risk in specified
circumstances.
• Risk means the probability that a
specified event will occur, e.g
disease or death.
• Risk factor is defined as an attribute
or exposure that is known, from
epidemiologic evidence, to be
associated with a health condition,
can be modified and considered
important to prevent.
• Risk factor for a disease must be
demonstrated as such longitudinally.
• If a suspected R.F cannot be
confirmed as such because the
necessary longitudinal studies are
impractical or unethical, the factor
may be classed as a risk indicator.
• Risk indicator is a factor shown to
be associated with a disease in
cross sectional studies.
• Research experience has shown that risk
indicators which ensure from cross
sectional studies can disappear in more
rigorous longitudinal analysis indicating
that it is not a true R.F.
• Risk marker is an attribute or exposure
that is associated with the increased
probability of disease although it is not
considered part of the causal chain.
• It also be called a risk predictor when
included in predictive statistical model
whose purpose is to predict disease
occurrence.
• Recently the term demographic R.F is use
to refer to these factors such as age a,
gender, race and socio-economic status.
• They clearly are of no use when
considering disease prevention as they are
not modifiable
• The following conditions must be satisfied
before particular exposure can be
accepted as the cause of a disease called
the Bradford Hill Criteria. It can be reached
from non experimental epidemiologic
analysis.
• 1- The sequence of events :- To be causal,
an exposure must precede the occurrence
of a disease which need longitudinal study
to demonstrate it.
• 2- Consistency of association :- Good
number of studies proved that with
fairly similar positive results.
• 3- Strength of association
statistically :- In valid studies, the
stronger the association between
exposure and outcome, the more
likely it is that the association is
causal.
• 4- Specificity of association :- If a given
exposure is related to other disease as
well as the disease in question, it is less
likely to be seen as causal. However, lack
of specificity by itself does not justify
rejecting causality.
• 5- Degree of exposure ( Dose response )
that is the risk of disease should be related
to the degree of exposure except toxin.
• 6- Biologic plausibility :- Association
must make biologic sense from our
knowledge of the disease.
• 7- Absence of error :- The observed
association must not be the result of
error, whether that be bias, sampling
error, analytical errors, or the
intrusion of other extraneous factors
( confounding factors ).
• The most important factors are
:-
• Time sequence.
• Statistical association.
• Absence of error.
Analytical studies :-
• Analytical studies attempt to
determine association between
disease and possible risk factors or
demographic R.F, and to quantify the
degree of risk.
• The determination of risk factors for
a disease allows the potential for
preventing the disease by removing
or modifying the risk factors.
Types of analytical
studies :-
• A – Cohort study ( prospective study ).
• Observational analytic design, in which
group of individuals are defined on the
basis of presence or absence of exposure
to a suspected risk factor for a disease.
• The participants are then followed over a
period of time to assess the occurrence of
that disease.
• Aiming to provide sound evidence of
whether there is a valid statistical
association between exposure and
disease.
Prospective Cohort study
• They are followed over time
(longitudinal) to see at what
frequency they develop the
disease/condition as compared with
unexposed control group.

• The exposure has to precede the

outcome
(development of the
disease/condition)
• Advantages of cohort study :-
• Allow the investigator to examine a
large number of hypothesis at one
time.
• Gives valid results.
• Disadvantages :-
• Take time to get the out come data.
• Very expensive ( long time ).
• Sample size must be large ( drop out
?? )
 Prospective Cohort study

Useful when:
- The disease/
condition of interest
occurs frequently.
Disadvantages:
- Subjects can be
readily obtained.
- The time it could take to
- The risk factors are develop the disease or
known or thought to condition of interest (lung
cause harm (tobacco) cancer)
- When there are
ethical - The cost of follow-up
considerations.
- The potential for losing
subjects over time.
Prospective cohort study.
Non experimental

• Analytical, typical cohort study.

• Ask question “what will
happened”
• Longitudinal.
• Drop out bias.
Historical cohort or retrospective
cohort studies.

• By using information collected in the

past and kept in records or files.
• The event being evaluated actually
occurred before the onset of the
study. the cohort of subjects has
been assembled for other purpose.
• Direction of inquiry is still forward in
time, from possible cause or risk
factor to an outcome
Historical Cohort
• B – Case – control study(Retrospective)
• Are relatively simple and economical to carry out,
commonly used to investigate causes of rare
diseases.
• It include people with a disease of interest ( case )
and suitable unaffected group ( control ).
• Investigator collects information on exposure
history from diseased study subjects ( case ) and
non diseased ( control ).
• Aiming to provide sound evidence of whether there
is a valid statistical association between exposure
and disease.
 Case control study
• Make observations about possible associations
between the disease of interest (lung cancer) and
one or more hypothesized risk factors (tobacco
use).

• Retrospective in that subjects already have a

certain disease or condition and are compared with
a matched disease-free controls from the same
population.
Case – control study.
Non experimental
• Analytical study that ask “what
happened”.
• Begin with the absence or
presence of an outcome, then
look backward in time
(retrospective) to try to detect
possible cause or risk factor.
• Matching concept is essential to be
sure that both groups will be similar
with respect to important variables
that may otherwise cloud or
confound the conclusion.
• longitudinal
• Recall bias.
 Case control study

Disadvantages:
Useful in :
• That investigators are
• Studying the etiology of looking back in time & rely
rare diseases because on:
they are difficult to study • The subjects’ recall
on a population basis. (recall/ memory bias)
• Incomplete sources of:
• Information for
• Allows multiple etiologic exposure histories
factors to be studied
concurrently.
 Case control study

• The assumption is that the differences should

explain why the cases developed the
condition/disease of interest and the controls did
not.

• Less reliable a statistical relationship

between two conditions does not mean that
one condition actually caused the other.
(confounding factors).

• Researchers should confirm the results with a

RCT or a prospective cohort study.