Transcript The double blind, randomized, placebo

Kevin Nesh
THE DOUBLE BLIND, RANDOMIZED, PLACEBOCONTROLLED TRIAL: GOLD STANDARD OR
GOLDEN CALF?
OBJECTIVE: TO THINK ABOUT WHAT WE ARE TAUGHT
“SOME FACTS MAY NOT EXIST INDEPENDENT OF THE
APPARATUS OF THEIR PRODUCTION”
Hierarchy of evidence for questions about the effectiveness of an intervention or treatment.
A K Akobeng Arch Dis Child 2005;90:840-844
Copyright © BMJ Publishing Group Ltd & Royal College of Paediatrics and Child Health. All rights reserved.
INTRODUCTION
•
The general adoption of the Double-blind RCT was based
on theoretical reason and intuitive attractiveness rather
than a compelling body of data
Randomization: aim to eliminate unconscious
and deliberate human influence
• Blinding: treatment and analysis of outcomes
are not colored by prejudice
Where is the evidence that our evidence is real?
•
THE DISCREPANCY ARGUMENT

Primary empirical evidence for the objectivity of
the double-blind RCT lies in the differential
outcomes it detects compared with other
research designs
 The
more rigorous the methods the less efficacious
the treatment
HISTORY- 1950-1954
SINGLE- BLIND PILOT STUDY
19 patients demonstrated that
the cardiac drug Khellin was
dramatically better than
placebo pill
DOUBLE BLIND PILOT STUDY

39 patients demonstrated
no difference between the
two groups
The history of the innovation of blinding starts well before this in the 1700’s
with a debate about mesmerism; but in contemporary medicine this is the
sole piece of data that began the adoption of our modern arguments
From 1960-1970 many other reports mostly in the psychiatric realm
continued to see less efficacious treatments with the blinded approach
and further used this as validation of the RCT approach
Whats the take away point?: RCT’s produce smaller efficacy thus are
more true
WRONG!

Post 1970’s metanalysis:
“the unpredictability paradox: review of empirical
comparisons of randomized and non-randomized clinical
trials” – kunz
 Objective To summarise comparisons of randomised
clinical trials and non-randomised clinical trials, trials
with adequately concealed random allocation versus
inadequately concealed random allocation, and high
quality trials versus low quality trials where the effect of
randomisation could not be separated from the effects
of other methodological manoeuvres
WRONG CONTINUED

Eleven studies that compared randomised controlled
trials with non-randomised controlled trials (eight for
evaluations of the same intervention and three across
different interventions), two studies that compared trials
with adequately concealed random allocation and
inadequately concealed random allocation, and five
studies that assessed the relation between
qualityscores and estimates of treatment effects, were
identified. Failure to use random allocation and
concealment of allocation were associated with relative
increases in estimates of effects of 150% or more,
relative decreases of up to 90%, inversion of the
estimated effect and, in some cases, no difference.
SO NOW WE CHANGE SYNTAX

Modified Discrepancy Argument

Instead of consistently adjusting for bias in the
direction of inflated estimates of effects, the
methodological safegaurds of randomization and
blinding are now considered the “best protection
against the unpredictability… of bias”
Note: there are studies that have found observational
studies to be more homogenous than RCT’s
 Thought: Scientific syntax is very persuasive- if you
instead of saying double-blinded RCT use the words
intentional ignorance RCT does it sound as persuasive?

CAN AN “UNBIASED” METHOD PRODUCE ITS
OWN DISTORTIONS?

- Circular Logic:
“the truth is what we find out in such a way”
 We recognize it as truth because of how we find it
 How do we know that the method is good?

 Because

it get at the truth
Randomly subjecting a person to a milieu of
hidden exposures and then spolighting him or her
with relentless observation does not nurture
normalcy, nor does it isolate humans from their
mental processes. They are not immune to any
“unconscious” processes that subvert science
MASKING
Knowledge that one has a chance of receiving
placebo may introduce uncertainty sufficient
enough to decrease the magnitude of the
response or extreme vigilance increasing its
effects
 “resentful demoralization” – participation may
create ambivalence, confusion, passivity and
voluntary submission

PLATINUM STANDARD
Trial in where both physician and patient were
unaware that they were even involved in a blind
RCT
 France 1990:

30 matched pairs of pt with insomnia DB-RCT
 1 group knew of trial for new Benzo with hypnotic
effects versus placebo; other group unaware
 Both groups given placebo only

 Control
group had higher hypnotic activity (p<.05)
 Does informed consent mitigate the drug?
PLATINUM STANDARD CONT.

France: 49 consecutive patient with mild/mod
cancer pain randomly chosen to be informed or
not informed
25 not informed pts received naproxen and placebo
randomly thinking they were under standard of care
 18 informed patients entered the identical DB-RCT

 Informed
group placebo and naproxen significantly more
effective than any intervention of non informed (placebo even
better than naproxen of non-informed)
 Knowledge can affect placebo and treatment
disproportionately!!!
 If placebo in treatment and non treatment arm are not the
same- how can you do any significant analysis???
PLATINUM STANDARD CONT AGAIN

US study




100 volunteers; all told they would be in a trial receiving placebo or real
drug (caffeine)
50 then told they would be in Double Blind RCT (real coffee versus
decaff)
50 were deceived saying would receive real coffee
Alertness, tension, pulse, BP, certainty of consumed caffeine were
monitored
 Deceptive
administration of placebo resulted in
higher pulse rate and alertness (p<.05 p<.003)
 Conclusion: certainty of receiving a drug maximizes
effects and doubt diminishes effects
PLATINUM STANDARD SOME MORE



Another caffeine experiment
100 subjects; 3 arms ( informed receive caffeine, informed
no caffeine, informed RCT) then all given random caffeine or
decaf
Results: cognitive states interacted with drug and placebo on
several outcome points




No caffeine reported less alertness than those told received
caffeine or double blind
Told caffeine and did not get caffeine reported greater tension
Double Blind lower BP than other groups and only reliable
discriminate group
Conclusion: DB studies may lead to erroneous conclusions about
clinical effects of particular drugs
Thank you for letting me enthrall you with my words
TAKE AWAY MESSAGE?