Randomized Clinical Trials: The Versatility and Malleability of the Gold Standard

1
Randomized Clinical Trials The Versatility and
Malleability of the Gold Standard
Get your facts first, then you can distort them
as you please. (Mark Twain)

• Wing Institute
• Jack States
• Ronnie Detrich
• Randy Keyworth

2
Randomized Clinical Trials (RCT)History

• The use of RCT in agricultural experiments was
  pioneered by R. A. Fisher in the 1920s.
• The Tuberculosis Trials evaluating the
  effectiveness of streptomycin, is generally
  accepted as the first RCT - Post World War II

3
Randomized Clinical TrialDefinition

• A research study in which the participants are
  assigned groups to objectively compare different
  interventions.
• The use of chance to assign groups results in
  groups being similar
• The effects of the intervention can be evaluated
  based upon the statistical assumption that any
  error
• (a). can be calculated and
• (b) can be made very small

4
Why Use Randomized Clinical Trial?

• To make statements of cause and effect
  regarding an intervention
• RCT is recognized as a sound scientific method
• By the 1960s RCT had established a track record
  as the most accepted methodology

5
How Does Randomized Clinical Trial Achieve Its
Designed Purpose?

• Control of Bias
• Subject Selection Bias
• Increases internal validity by minimizing subject
  differences.
• Study Evaluation Bias
• Limits inadvertent tampering with results

6
Strengths of RCT Design

• Statements of internal validity
• Statements of external validity (Limited to the
  populations studied)
• Demonstrates strong social validity as the most
  commonly accepted method for evaluating research
• Governmental entities - World Health Organization
  and US Department of Education
• Professional associations - AMA and APA
• Advocacy organizations - Cochrane and Campbell
  Collaborations

7
Benefits of Randomized Clinical Control Design

• Valuable tool for making policy decisions in the
  selecting interventions with a strong evidence
  base
• When combined with a public health approach RCT
  offers a cost effective strategy to influence
  behavior on a broad scale through selection of
  effective universal interventions
• A way to avoid politically motivated,
  fad-of-the-month educational practices

8
Limitations

• Philosophical and Ethical Issues
• RCT creates a moral dilemma When the need to
  advance science is weighed against the obligation
  to offer each student optimal treatment
• Practitioner has an obligation to chose an
  intervention in the best interest of subjects
• A process must exist to protect subjects
• Achieving cultural acceptance

9
Limitations

• Strategic Issues
• Single studies are not sufficient to establish
  causation
• Difficult with low incident populations
• Requires specialized expertise to obtain
  approval, implement, and analyze data
• If you torture data sufficiently, it will
  confess to almost anything Dr.Fred Menger

10
Limitations

• Strategic Issues
• Limited ability for use making conclusions for
  individuals
• Substantial time required to complete studies
• Limited number of data points per condition
  (you cant see how independent variable works
  only that it works)

11
Limitations

• Tactical Issues
• Issues of treatment fidelity and integrity
• Timely response to proliferation of new practices
• Logistical Issues
• Challenges finding adequate numbers of subjects
• Expensive requiring substantial resources
• Limited number of studies currently available

12
Issues for Behavior Analysis (BA)

• RCT is considered a social science methodology
  vs. the natural science methodology preferred by
  BA
• BAs focus is on the individual and the
  methodology of choice is Single Subject Design
• Question Should BA adopt RCT as a part of its
  research methodology?

13
Recommendations

• BA needs to participate in use of RCT
• To remain a player in discussion and policy
  making
• Increase the acceptance of BA
• To make the general contributions of which our
  science is capable, behavior analysts will have
  to use methods of wider generality, in the sense
  they affect many people at the same time.
• Murray Sidman
• The Behavior Analyst - Fall 2006

14
Recommendations

• Increase the number of studies
• Increase education and training of consumers and
  practitioners in strength and weakness of RCT
• Integrate RCT into an effective Hierarchy of
  Evidence model that takes advantage of different
  methodologies.