Transcript Randomized Controlled Trial

Randomized
Controlled Trial
Subodh S Gupta
Dr. Sushila Nayar
School of Public Health
MGIMS, Sewagram
Type of study
Alternate name
Observational studies
Unit of study
Descriptive studies
Analytical studies
Ecological
Correlational
Populations
Cross-sectional
Prevalence
Individuals
Case-Control
Case-Reference
Individuals
Cohort
Follow-up/ Longitudinal
Individuals
Experimental/ intervention Studies
Randomized Controlled Clinical Trial
Studies
Patients
Field Trial
Healthy person
Community Trial
Community intervention
studies
Communities
Progression of Study Design:
Clinical Research







Isolated Case Reports
Case Series
Cross-Sectional study
Case-Control Study
Cohort Study
Randomized Clinical Trial
Meta-Analysis
EXAMPLE:
 The role of oxygen in retrolental fibroplasia RLF among
premature infants.
Progression of Study Design:
Clinical Research




First Case - Feb. 14, 1941, Dr. Clifford, Boston
Case Series - 1941 (Silverman 1980)
Ca-Co Study (53 RLF Children, 298 Normal Children)
Association was observed. Still, it was postulated that
poor health of infants necessitated longer hours of
oxygen. Poor health and not oxygen use caused RLF.
Progression of Study Design:
Clinical Research




Cohort Studies: Contradictory Results
I RCT: Gallinger Muncipal Hospital, Washington, DC
II Collaborative Multi-centre Trial
Confirmed the role of oxygen in the etiology of
Retrolental Fibroplasia
Progression of Study Design:
Community Research






Ecological Study
Cross-Sectional Study
Case-Control Study
Cohort Study
Randomized Community Trial
Meta-Analysis
EXAMPLE:
Lipid - Atherosclerosis Association
Progression of Study Design:
Community Research





Analysis of Death Rates from CAD according to per
capita fat consumption in 20 countries  Hypothesis of
L-A association.
CS Studies: Framingham and Evans County Heart
Studies (Dawber et al 1971, Cassel 1971)
Case-Control Studies confirmed Association.
Cohort Studies (Truett et al 1967, Tyroler et al 1971)
Community Based Controlled Trials of Lipid Reduction
(Lipid Research Clinics Program)
Deciding which one to use
The investigator observes the
events without altering them
NO
Observational study
Example: Comparing the history of
needle sharing among IV drug abusers
who have HIV antibodies with those who do not
Decision # 1
Alter the events under study?
Example: Impact of health education
Yes
The investigator applies an intervention,
& observes the effect on the outcome
on needle sharing habits
Experimental study
Deciding which one to use
Each subject is examined
on only one occasion
NO
For observational studies
Cross-sectional study
Example: Study of needle sharing habits
and HIV antibodies measured at the
same time
Decision # 2
Make measurements on more
than one occasion?
Yes
Each subject is followed over
A period of time
Example: Cohort study that assesses
current needle sharing habits of group of
IV drug abusers and observes who
subsequently develop HIV antibodies
Longitudinal study
Deciding which one to use






Can you alter the events under the study?
How strong is the hypothesis?
How common is the disease or health event which is to
be studied?
How common is the exposure/ determinants of the
health event?
Do you want to study the different factors/
determinants of a health event or disease? Or;
Do you want to see the multiple effect of an exposure?
How much resources do you have?
Randomized controlled trials
”An epidemiological experiment in which
subjects in a population are randomly
allocated into groups, usually called study
and control groups to receive and not receive
an experimental preventive or therapetuic
procedure, maneuver, or interventition”
John M.Last, 2001
Randomized Controlled Trial
• A true experiment
• Key features
– the classic way to evaluate efficacy or
effectiveness of drugs (or exercise, diet,
counseling)
– patients are followed over time (prospective)
• Properly done, an RCT can be used to
determine cause and effect.
Why RCT?
• ”Gold standard” in epidemiological research
• Makes study groups comparable
– Controls for confounding (known and unknown)
– Prevents selection bias
“RANDOMIZED, DOUBLE-BLIND, CONTROLLED
TRIAL” is considered as research design par
excellence and “GOLD STANDARD” amongst
research designs with which results of
other studies are often compared.
Deviation from this standard has potential
drawbacks
Advantages
• Most efficient for investigating causality
• Ensure ‘ONLY ONE’ factor is different:
confounding factors do not confuse the results
• Ensure that treatments are compared efficiently
• Look for effects of combinations of treatments,
interaction between treatments and personal
characteristics
• Only study design which can help us evaluate a
new treatment (medicine, other procedures etc.)
Disadvantages
• Share many of the disadvantages of cohort
study
• Ethical concerns
• It may not be possible for all kinds of
questions that we have
• Intervention studies screen out ‘problem’
subjects, such as the very young, the elderly
and pregnant and lactating women
Ethical Considerations
•
•
•
•
•
•
•
Major issue for ‘Randomized Controlled Trial’
Proper information to all the study subjects
Informed consent
The trial is conducted ethically
Avoid bias in results
Sample size is adequate to give the results
What if, before the study is completed, there is
evidence that one treatment is better than the
other one
ETHICS
IMPORTANT ISSUE IN CLINICAL TRIALS
ETHICAL CLEARANCE
*
INSTITUTIONAL REVIEW BOARDS
*
ETHICAL COMMITTEES
*
ICMR GUIDELINES
*
FEDERAL/STATE GUIDELINES
19
Types of Randomized Controlled Trials
1. Clinical Trial:
Diagnostic, Therapeutic, Prophylactic, Devices,
Procedures, Regimens, Protocols
2.
3.
4.
5.
6.
Preventive Trial
Risk Factor Trial
Cessation experiments
Trial of etiologic agents
Evaluation of health system
Types of Randomized Controlled Trials:
1. Clinical Trial
- Concerned with evaluating therapeutic
agent, mainly drugs
eg. Evaluation of beta-blockers in reducing
cardiovascular mortality
- Not all clinical trials are susceptible to being
blinded
21
2. Preventive Trials:
- Trial of primary preventive measures eg.
Vaccines
- Analysis of preventive trials must result in
clear statement about benefits to community,
risk involved and cost to health
22
3. Risk Factor Trials:
- Investigator intervenes to interrupt the usual
sequence in the development of disease for
those individuals who have risk factor for
developing the disease
- Primary prevention of CHD using clofibrate to
lower serum cholesterol
23
4. Cessation Experiment:
- An attempt is made to evaluate the
termination of a habit which is considered to
be causally related to disease
- Cigarette smoking and lung cancer
24
5. Trials of Etiological Agents:
- To confirm or refute an etiological hypothesis
6. Evaluation of Health Services:
- Domiciliary treatment of PTB was as effective
as more costlier hospital or sanatorium
treatment
25
MULTICENTER TRIALS
Reasons for Multi-center Trials :
1. To recruit necessary number of subjects within
a reasonable time.
2. May assure a more representative sample of the
study or target population
3. Enables investigators with similar interest and
skills to work together on a common problem
26
CLINICAL TRIALS
• Prospective study comparing the effect and value
of one of more interventions against a control in
human subjects with a given medical condition.
• Measures causality in terms of the effect of an
intervention: If one alters the risk factor, does one
alter the occurrence of the event/injury?
• "...the most definitive tool for evaluation of the
applicability of clinical research.“
WHAT IS CLINICAL TRIAL
• A clinical trial is defined as a prospective study
comparing the effect and value of
interventions against a control in human
beings
• Study participants must be followed forward
in time. They need not all be followed from an
identical calendar date.
• Must contain a control group against which
the intervention group is compared.
• At baseline, the control group must be
sufficiently similar in relevant respects to the
intervention group so that differences in
outcome may reasonably be attributed to the
action of the intervention.
• Most often a new intervention is compared
with best current standard therapy.
Stages of experimentation
– Phase I: dose-finding
– Phase II: preliminary evidence of efficacy
– Phase III: comparisons to standard therapy
– Phase IV: post-marketing surveillance
PHASES OF TRIALS
Phase I Trials:
• Initial studies to determine the metabolism and
pharmacologic actions of drugs in humans, the
side effects associated with increasing doses,
and to gain early evidence of effectiveness;
usually conducted on healthy volunteers
31
PHASES OF TRIALS
Phase II Trials:
• Controlled clinical studies conducted to evaluate
the effectiveness of the drug for a particular
indication or indications in patients with disease
or condition under study and to determine the
common short-term side effects and risks
32
PHASES OF TRIALS
Phase III Trials:
• Expanded controlled and uncontrolled trials
after preliminary evidence suggesting
effectiveness of the drug has been obtained,
and are intended to gather additional
information to evaluate the overall benefit-risk
relationship of the drug
33
PHASES OF TRIALS
Phase IV Trials:
• Post-marketing studies to delineate additional
information including the drug’s risks, benefits,
and optimal use
34
Steps in conduct of RCT
1. The protocol
2. Selecting reference and experimental
populations
3. Randomization
4. Intervention
5. Follow up
6. Assessment
1. The Protocol
-
-
Rationale
Aims and objectives, Research questions
Design of the study: selection of patients, drugs
and doses, assessment, withdrawals, data analysis,
data discharge
Ethics: patient consent, adverse events
Documentation
Procedure
43
2. Selecting Reference and Experimental
Populations
a. Reference or target population - population to
which the findings of the trial, if found successful,
are expected to be applicable (eg. drugs, vaccines,
etc.)
b. Experimental or study population - actual
population that participates in the experimental
study
44
Participants must fulfill the following criteria:
-
Must give informed consent
-
Should be representative of the population
-
Should be qualified or eligible for the trial
45
SAMPLE SIZE
Clinical trials should have sufficient statistical
power to detect differences between groups
considered to be of clinical interest. Therefore,
calculation of sample size with provision for
adequate levels of significance and power is
essential part of planning.
3. Randomization
- Heart of the control trial
- Procedure: Participants are allocated into
study and control groups
- Eliminates bias and allows comparability
- Both groups should be alike with regards to
certain variables that might affect the
outcome of the experiment
- Best done by using table of random
numbers
47
RANDOMIZATION
Randomization tends to produce study groups
comparable with respect to known as well as
unknown risk factors, removes investigator
bias in the allocation of subjects and
guarantees that statistical tests will have valid
significance levels.
48
4. Manipulation / Intervention
- Deliberate application or withdrawal or
reduction of a suspected causal factor
- It creates an independent variable
49
5. Follow Up
- Implies examination of the experimental and
control group subjects at defined intervals of
time, in a standard manner, with equal
intensity, under the same given circumstances
- Attrition: Inevitable losses to follow up
50
6. Assessment
- Positive results
- Negative results
- Biases: Subject variation, Observer bias,
Evaluation bias
- Can be corrected by blinding
51
Randomized Controlled Trials
Example:
Randomized Controlled Trials
Avoidance of bias
•
•
•
•
Use of a control group
Blindness
Randomization
Consent before randomization
Control group
•
•
•
•
•
Placebo
Most widely accepted treatment
Most accepted prevention intervention
Usual care
Accepted means of detection
Randomization: Definition
• Not a random sampling
• Random allocation
– Known chance receiving a treatment
– Cannot predict the treatment to be given
• Eliminates selection bias
• Similar treatment groups
Only one factor is different
• Randomization tries to ensure that ONLY ONE
factor is different between two or more
groups
• Observe the consequences
• Attribute Causality
Randomization
• We want to assign a group of subjects to one
of two groups—Treatment A or Treatment B
– How can we do this in a random manner?
Randomization
Random assignment
• Flip a coin
– “Heads”—Tx A
– “Tails”—Tx B
Randomization
Random assignment
• Roll a six-sided dice
– Even number—Tx A
– Odd number—Tx B
Randomization
Random assignment
• Table of random numbers
• Computer generated list
‘Almost’ Random assignments
Random assignment
• Alphabetical
– Tx A = patients with last name A–M
– Tx B = patients with last name N–Z
• Telephone number/social security number
– Tx A = last digit odd
– Tx B = last digit even
• Sequential
– Tx A = morning patients
– Tx B = afternoon patients
• Bed number
– Tx A = odd bed number
– Tx B = even bed number
‘Almost Random’ assignments
There are potential problems in the “Almost
Random” assignment scheme
– Do you see a potential problems with these
‘Almost Random’ assignment scheme
Potential problems with
Simple randomization (flip a coin)
Randomize individuals to one of two treatments
• If n is big, works great
• If n is small
– May be imbalanced with respect to . . .
• Sample size
• Other variable
Unequal sample sizes
• If study has very small sample size, there is no
guarantee two groups will have equal sample size
using simple randomization
Block Randomization
• Ensure that # of patients assigned to each
treatment is not far out of balance
• Variable block size (permuted)
– An additional layer of blindness
• Different distributions of a trait like gender in
the two arms possible
Block randomization
•
•
•
•
•
•
AABB
ABAB
ABBA
BABA
BAAB
BBAA
Six different
ways to
arrange two As
and two Bs
Block randomization
Roll a die (#1–6) to determine pattern
• Each pattern has same probability of being
chosen (one in six)
• Guarantees balance after every four patients
Potential problems with
Simple randomization (flip a coin)
Imbalance on a key variable
• If study is very small, no guarantee groups are
“comparable”
• Solution—stratify
Stratified randomization
• A priori certain factors likely important (e.g.
Age, Gender)
• Randomize so different levels of the factor are
BALANCED between treatment groups
• Cannot evaluate the stratification variable
Stratified randomization
• Stratify, then do block randomization
ABBA
BBAA
BABA
ABAB
BAAB
Female; 45-60 yrs AABB
ABBA
BBAA
BABA
ABAB
Male; 25-44 yrs
BLINDING
•UNBLINDED, OPEN TRIAL
•SINGLE BLIND
•DOUBLE BLIND
•TRIPLE BLIND
71
Double-Blinded
Single-Blinded
Types of Randomized Studies
•
•
•
•
•
•
Parallel group
Sequential trials
Group sequential trials
Cross-over
Factorial designs
Adaptive designs
Parallel Group
• Randomize patients to one of k treatments
• Response
– Measure at end of study
– Delta or % change from baseline
– Repeated measures
– Function of multiple measures
Sequential trials
• Not for a fixed sample size/ period
• Terminates when
– One treatment shows a clear superiority or
– It is highly unlikely any important difference will
be seen
– Special statistical design methods
Group Sequential Trials
• Popular
• Analyze data after certain proportions of
results are available
• Early stopping
– If one treatment clearly superior
– Adverse events
• Careful planning and statistical design
Factorial design
• Each level of a factor (treatment or condition)
occurs with every level of every other factor
• Vitamin A and Vitamin E for prevention of
Hypertension:
Vitamin A Placebo Vitamin A
Vitamin E Placebo Vitamin E Placebo
Vitamin A Placebo
Vitamin E
Vitamin A
Vitamin E
Incomplete/ Partial/ Fractional
Factorial Trial
• Nutritional Intervention Trial
• 4X4 incomplete factorial
• Did not look at all possible interactions
– Not of interest
– Sample size prohibitive
Cross-over Trial
• Two treatments, two period cross-overs
• Use each patient as own control
• Must eliminate carryover effects
– Need sufficient washout period
INTERVENTION
PERIOD 1
SUBJECTS
RANDOM
PERIOD 2
ALLOCATION
CONTROL
CROSS OVER DESIGN
82
Adaptive designs
• Smaller overall sample size (potential)
• Run-in; then analyze data continuously or
fixed intervals
• Act like group sequential design
• Close an arm early
• Re-estimate sample size based on variance
Flow diagram of the progress through the phases of a randomized trial
Flow diagram of the progress through the phases of a randomized trial
Exercise 1
• A multicentric randomized double blind study
to assess the efficacy of Probiotics for
reduction of risk of sepsis among neonates
• Five sites have been selected for this study.
• Describe a suitable method for allocating
hospital patients to intervention groups.
Exercise 2
• A study planned to assess the desirability, and
overall impact on the health services of day
surgery
• Several hospitals agree to take part in the study.
• Two groups will be compared using various
subjective criteria (self-assessed health) and
factual criteria
• Describe a suitable method for allocating hospital
patients to intervention groups. Assume that a
list of day surgeries have already been
established.
Exercise 3
• In a study of four treatments for eradication of H.
pylori, Tham et al. report the following eradication
results (expressed as ratios of eradication to number
treated)
–
–
–
–
Omeprazole + Amoxycillin + Metronidazole – 6/20
Ranitidine + Amoxycillin + Metronidazole – 8/20
Omeprazole + Placebo – 0/20
Omeprazole + Clarithromycin – 4/20
• Test whether there is a significant difference between:
i. The first two treatments in this list
ii. The third treatment (the only one not involving an
antibiotic) and all the rest combined
Exercise 4
• Refer to Cerebral Palsy data:
i. Test whether the addition of rhizotomy has a
significant effect on motor functions
ii. Summarize the effect of adding rhizotomy,
giving a 95% confidence interval for your
summary statistics
Exercise 5
• Refer to data given for Norwegian Multicenter
Study (Table C.8)
i. Construct a separate actuarial life table for
each treatment group
ii. Plot the estimates of cumulative survival
probability on a graph and interpret the
findings
iii. Compare the overall survival in the two
groups