CLINICAL TRIALS

DR DINESH KUMAR WALIA

ASSTT PROFESSOR
DEPARTMENT OF COMMUNITY MEDICINE
GOVERNMENT MEDICAL COLLEGE
CHANDIGARH
Types of Studies:-

Randomized Trails

(a) Experimental

Non Randomized Trails

Cross-sectional

Descriptive

(b) Observational Longitudinal

Case-control

Analytical

Cohort

Retrosp. Prosp.
 CLINICAL TRIALS
• Meaning be experimental studies:
• Problems in experimental studies:
– Cost
– Ethical issues
– Feasibility
• Types of experimental studies:
– Randomized
– Non-randomized
• Animal Studies & Human experiments:
– In animal studies disease in animals is experimentally
reproduced for testing efficacy of preventive and
therapeutic measures.
– Human experiments are conducted to investigate
disease etiology and efficacy of the measures.
 RANDOMIZED CONTROL TRIAL (RCT):
• RCT is an experimental study to evaluate methods of
treat/prevention involving random process of allocation
(to patients or randomly allocation of patients to
different measures).
DEFINITION:
• Carefully planned and ethically designed study
(experiment) with the aim of answering questions
concerning effectiveness of different regimens, a
surgical procedure or method of treatment, or
therapeutic regimen administered to patients.
NEED:
• Evaluation of safety and efficacy of therapies.
• Opportunities to screen new drug
• To detect differences between drugs or methods of
treatment
 TYPES OF RCT
Therapeutic:
– To compare efficacy of a new drug with the best of
current treatment (diet, surgery, physiotherapy,
ionizing radiation)
Prophylactic: To measure effectiveness of preventive
measures (weight reduction, contract, immunization,
fortification of food stuffs)
Why Placebo? To separate the effect of the therapy on
trial from the suggestive element introduced giving any
treatment, placebo a pharm. inert or harmless substance
or procedure made to resemble the drug or medical
procedure under evaluation is used
CLINICAL TRIALS:

• A type of RCT (a prospective planned, ethically designed

expt) to compare the effectiveness of different regimens
or methods of treatment in human subjects.
NEED FOR CLINICAL TRIALS:
– Evaluation of safety and efficacy
– Opportunity to screen new drugs
– To detect differences of responses and advantages
BIAS IN CLINICAL TRIALS:
– Systemic difference at admission
– Differential practice in follow-up
– Differential assessment of outcome
– Differential exclusion or withdrawal
METHODS OF REDUCING BIAS:
– Randomization
– Blinding
– Uniform handling of procedures

Single Blind Trail : the patient is unaware which

treatment he / she is receiving.
Double-blind trail: Neither the patient nor the
doctor assessing the response is aware which
specific treatment is given. Double blind trials
avoid physician’s or patient’s potential bias.

7
The basics: What is a randomized controlled trial?

• Simplest definition: Individuals are allocated at random

to receive one of several interventions (at least two
total).

• RCT’s are comparative studies (in contrast to case

series studies that do not make comparisons between
groups)

• RCT’s are experimental—the intervention is controlled

by the investigator

8
 Phase I, II, III, IV (human) trials

• Phase I
– Conducted after animal safety established
– Tend to focus primarily on human safety
– Focus on proper dosing and metabolism
– Participants sometimes include the investigators,
terminal patients, employees—often NOT people with
the disease the drug is designed to treat
– Often neither randomized nor controlled—these are
more like case series studies

9
SEQUENCE / PHASES OF CLINICAL TRIALS
New Drug / Surgical
Procedure

Animal
Phase – I : (on
Experiments
human
volunteers)

Phase – I
To obtain MTD for
safety / non Toxicity

Worthy for further Not worthy for further

Phase – II
expt expts

Phase – II Yes Phase – II (a)

To estimate Whether effective to
efficacy continue

Phase – III
Compare efficacy with
standard drug Phase-III
Terminate the Expt.
Introduction in
practice
Phase – IV
Post marketing
surveillance (large
scale)
 LAYUOUT OF AN RCT
Ref / Target Popn

Sample
Non eligible

Exclusion

Non
volunteers
Randomization

Expt. Group Control Group

Follow-up

Assessment of
Outcome
The essential feature of a trail is comparison of
one experimental group receiving the treatment
being evaluated and another group (control) being
given the standard treatment, or if there is no
generally accepted therapy, no treatment or
placebo.

12
 Types of RCT’s—classifcation schemes

• Based on the type of interventions being evaluated

• Based on how participants are exposed to interventions
• Based on the number of participants
• Based on whether goal is evaluation of superiority vs.
equivalence
• Based on whether investigators and/or participants know
which intervention is being studied (blinding)
• Based on whether the preferences of non-randomized
individuals are considered

13
 Types of RCT’s—classification schemes

• Based on the aspects of interventions being evaluated

– Explanatory and pragmatic trials
– Efficacy and effectiveness trials
– Phase I, II, III trials

14
 Explanatory vs. pragmatic trials

• Explanatory trials
– Address whether or not an intervention works
– Strict inclusion criteria; highly homogenous groups
– Example: study of hypertension that only enrolls 40-
50 year olds with no history of drug treatment
– Intended to yield as clean a result as possible
• Pragmatic (or management) trials
– Designed to test both whether the intervention works
but under circumstances mimicking clinical practice
– Sometimes will involve one drug vs. another rather
than placebo

15
 Efficacy vs. effectiveness

• Efficacy—does the intervention work in the people who

actually receive it?
– These trials tend to be explanatory
– Goal here is high compliance

• Effectiveness—how does the intervention work in those

offered it
– Tend to be pragmatic

16
 Superiority vs. equivalence trials
(equivalence trial slides courtesy of Starley Shade)

• Superiority trials
– Intended to determine if new treatment is different
from (better than) placebo or existing treatment
(active control).

• Equivalence trials
– Intended to determine that new treatment is no worse
than active control.
• We can never assess absolute equivalence.
• We can only assess no difference within a
prescribed margin.

17
 Why do an equivalence trial?

• Existing effective treatment

• Placebo-controlled trial unethical

– Life-threatening illness.

• New treatment not substantially better than existing

treatment.
– May have fewer side effects, greater convenience,
lower cost, higher quality of life, or provide an
alternative or second line therapy.

18
 Hypotheses

• Superiority trials
– Null hypothesis is that there is no difference between
treatments.
– Alternative hypothesis is that the new treatment is
different from (two-sided) or better than (one-sided)
control.

19
 Hypotheses

• Equivalence trials
– Null hypothesis and alternative hypotheses are
reversed.
• Null hypothesis is that difference between
treatments is greater than X.
• Alternative hypothesis is that difference between
treatments is less than X.

20
 Equivalence margin

• If confidence interval lies entirely within the equivalence

margin, then equivalent.

• If confidence interval lies entirely outside the equivalence

margin, then one drug is superior.

• If confidence interval crosses the equivalence margin,

then inconclusive results.

21
 How to set equivalence margin

• Superiority trials set sample size to detect a “clinically

significant” difference.

• Equivalence trials set sample size to establish “clinically

insignificant” difference.

• “Clinically insignificant” determined by information

outside of the trial.
– May be a source of great controversy.
– Has a large impact on sample size.

22
 Equivalence margin
ificance?
Yes.----- - •••• _____ .••• --l--- ...
I • - ···-· ---·. •••••
1 Not equivalent
- .- - - ~- - . . -- - -•

Yes - --------------------~---- ______ ____ _ _______ ◄ Unce~ain ►

L.. .. . ______ ___ _
Yes ------- -- -- --- ------- .. ◄ Equivalent ► :
No -------- ___________________ J___________ E uivalent 1
E . I t
I
Yes ---------------------------- 7- ◄ qu1va en ►
I
:
Yes ·------- -------- ,,,~ . Uncertain
I .-► 1

Yes-------- ◄ Not equivalent► 1

1
1
1
I I
No ----------------- Uncertain 1

23 0
True difference
 Challenges in design

• Necessity of a gold standard for existing therapy (active

control).
– May not exist if multiple existing treatments.
• Necessity to establish equipotent doses of new
treatment and active control.
– Requires prior testing of multiple doses of each drug.
– Difficult to know if smaller prior study not conducted.

24
 Challenges in design

• Best condition for new therapy may not match previous

research for active control.
– Testing of new therapy as second line treatment.

25
 Bias in equivalence trials

• In superiority trials, incomplete follow-up, low

compliance, and co-interventions tend to bias results
toward the null.

• In equivalence trials, these biases increase the likelihood

accepting the alternative hypothesis of no difference
between groups.

26
 Bias in equivalence trials

• Incomplete follow-up
– May limit observed response and therefore bias
results toward no difference.
• Low compliance
– May limit observed response and therefore bias
results toward no difference.
• Co-interventions
– May create ceiling in response and therefore bias
results toward no difference.

27
Outcome Measures: Response:

– Death, recovery, disability, pain, recurrence

Meaning by Placebo:

– Pharmacologically inert / harmless medical procedure

or substance, made to reasonable the drug /

procedure under evaluation.

Why Placebo:

– To separate the effect of treat on trail from the

suggestive element introduced by giving any

treatment.
 Trial classification: Efficacy

• Assessment of equivalence does not insure efficacy.

– Both drugs could be equally ineffective.
• Evidence of efficacy must come from:
– Outside the study, or
– An additional placebo arm.

29
 Efficacy

• The desire to establish efficacy using information outside

of the study
– Creates pressure to make equivalence studies mirror
the methods of the original placebo controlled trials of
the active control.
– However, this may not put the investigational drug in
the best light.
• The target of the two drugs may differ.

30
 Efficacy

• The desire to establish efficacy using an additional

placebo arm
– Continues the ethical debate about the use of
placebos.
• Some proportion of patients will not receive active
drug.

31
 Phase I, II, III, IV

• Phase II trials
– Intervention is given to those who actually have
disease
– Aim is to evaluate different doses
– Often not randomized

32
 How are quality assessments used?
• Clinicians may use assessments to determine how to
apply results to their practice
• Journals may use assessments to determine publication
• Researchers may use assessments to influence new
research
• General public (?) - recent example of anti depressants
and suicide among teens

33
 Essentials of reporting and interpreting individual trials
(Jadad ch 5)

• 30 years of empiric evidence supporting the idea that

there is a gap between what trials should vs. do report
for a reader to fully interpret the quality of a trial
• Key elements of a trial for a reader
– Is the topic interesting?
– Are the results likely to be unbiased?
– Can we use the results?
– Are the results important enough to remember?

34
 Key elements when reviewing a trial

• What was the sampling frame?

• What were the exclusion/inclusion criteria?
• Was the setting relevant (generalizability)?
• What were the interventions—how given and by whom?
• Are the details of randomization and blinding provided?
• What were the outcomes of interest and how were they
measured? By whom?
– Were the outcome assessors blinded?

35
 Key elements of a trial (cont.)

• Were the results properly analyzed?

– Statistical tests
– Statistical significance
– A priori sample size and/or post hoc power statement
(elaborate)
– Was intention to treat analysis used?
– Were sub-groups defined a priori?
• Is the flow of participants in the trial shown?

36
• Reducing bias in trials
– One example: efforts to limit direct funding by
the developers of the intervention
• More trials to address clinically relevant questions
• More precise results in trials
– Too many small, imprecise studies are done
– Would we be better off with fewer, more
definitive RCT’s?
• Improving the ways that trials are presented and
understood by the public
– Should (basic) interpretation of scientific results
be taught in schools?

37
 Improving trials (cont.)
• Registering trials at inception
• Publishing trials soon after completion, and
regardless of results
• Trials are more systematically captured in
systematic reviews and meta-analyses
• Make trials more easily accessible
• Decision makers need to learn how to interpret
trials
• Despite 50 years of improvement, most trials
are biased, too small, or trivial
38
 PREPARATION OF A PROTOCOL:
(Plan of RCT).
– Aims
– Review and significance of study

– Duration of the trail (period of entry &

follow-up)
– Patient population
• Inclusion
• Exclusion
– Experimental design
 – Treatment administration
Number of treatments involved,

Treatment allocation ratio in different groups (1:1 or

1:2 randomly)

Treatment management / allocation / administration

– Comparability of patients in different groups

(by randomization)
– Clinical, lab procedures and data to be
collected
40
 – Criterion for response and toxicity
– Frequency of interim analysis
– Statistical considerations [sample size,
randomization design of trail, analysis
strategies etc]
– Informed consent (ethical issues)
– Data collection forms
– References
– Responsible investigators

41