What is an experimental trial

1
What is an experimental trial?

• A planned experiment in humans
• Designed to test the efficacy of a treatment or
  intervention
• Comparison of outcomes in a group of participants
  receiving the intervention and those in a
  comparable group of patients with a control
  therapy
• Participants in both groups are enrolled, receive
  intervention, and followed over same time period

2
Gold standard study design

• Purposes include
• To test hypotheses
• to test effectiveness of interventions
• to suggest public health policies and programs
• to assess costs and benefits of programs
• to evaluate efficacy of health services

3
What is the question?

• Primary question
• Stated in advance
• Clearly defined
• Have adequate power to detect differences
• Usually stated as a hypothesis
• The outcome may be a beneficial action such as
  saving life, ameliorating illness, reducing
  symptoms, improving quality of life, or
  modification of an intermediate characteristic
• May be stated as testing no difference between
  groups

4
What is the question?

• Secondary question
• May examine a different response/outcome
  variables
• May examine subgroups of the population
• Adverse effects
• Natural history

5
Intervention

• Drugs, procedures, lifestyle modification, etc
• Must be standardized across treatment groups
• Potential intervention must be maximized
• Must take into account dose, time of initiation,
  duration, logistics, etc..

6
Basic Trial Design
7
Study Design

• Use of a control/comparison group
• Randomization
• removes potential selection bias
• produces comparable groups
• validity of statistical test guaranteed
• although groups compared are never perfectly
  balanced for important covariates in any single
  experiment, the process of randomization makes it
  possible to ascribe a probability distribution
  to the difference in outcomes between the two
  groups receiving equally effective treatments and
  thus assign significance levels to observed
  differences

8
Randomization

• Simple
• Blocked
• Stratified
• Multi-utility attribute

9
Other Designs

• Non-random concurrent controls
• Crossover
• Historical controls/databases
• Factorial design
• Controlled clinical trial vs community or
  group-randomized trials

10
Cancer Trials

• Prophylactic trial
• Designed to test the effectiveness of a substance
  to prevent a disease (chemoprevention trials)
• Phases of clinical trials
• Phase I -
• Initial testing new vaccine or drug, small
  sample, identify dose-limiting toxicities,
  optimal dose range, pharmacology of substance
• Phase II
• Testing on specific tumors, slightly larger
  sample, determine efficacy on tumor size,
  validate toxicity , dose, may be randomized
• Phase III
• Randomized, testing efficacy, outcomes are
  survival, QOL, etc..

11
Outcome/Response variable

• Total mortality, death from a specific cause,
  incidence of disease complication or specific
  adverse effect, symptomatic relief, clinical
  finding, laboratory measure, cost
• May be stated as a change from one state to
  another (alive to dead) or from a level of a
  continuous variable to another
• Defined in advance with operational definitions
• Capable of being assessed in all participants
• Sometimes participation ends in the study when
  response variable occurs while other times there
  is a fixed endpoint
• Capable of unbiased assessment
• Response variable should be capable of being
  ascertained as completely as possible
• Measured independently of treatment and control

12
Masking or Blinding

• Concealing knowledge of treatment assignment to
  reduce bias
• Patients/participants
• Observers (investigators)
• Data monitors/ collectors/ analyst
• Primary purpose is objectivity
• Single masked- participant is masked regarding
  intervention
• Double masked- participant and investigator masked

13
Study Population
Population w/o condition
Definition of condition
Ineligible
Entry Criteria
Enrollment
Not enrolled
14
Study Population- Inclusion criteria

• Participants who have potential to benefit from
  the intervention
• High likelihood you can detect the hypothesized
  effect
• Weigh adverse effects against possible benefits
• Competing risk
• Likely to adhere to study protocol

15
Generalization

• Participants are usually nonrandomly chosen from
  the population
• Participants in trials are considered special
• Volunteer population may have certain
  characteristics
• Always attempt to characterize the population
  and how different from the target population
• Generalization appropriate in some cases but not
  in others

16
"Final Report" On The Aspirin Component Of The
Physician's Health Study by Harris Gelberg, M.D.
Department of Cardiology In the Spring 1988
issue of the Highlights (Volume 11, Number 2) the
preliminary report on the effects of aspirin in
the prevention of heart attacks was reviewed. In
the July 20, 1989 issue of the New England
Journal of Medicine, the final report was
submitted. The information gleaned from this
study that involved 22,071 participants over 60.2
months revealed a significant reduction in heart
attack risk in those participants taking one
adult aspirin every other day. This group
demonstrated a 44 reduction in heart attack risk
when compared to the group taking a placebo.
Although there was no overall reduction in
cardiovascular mortality, the author concluded
this was due to the very low incidence of heart
attacks among physicians in this study This
reduced risk was seen only in participants over
50 years of age. Furthermore, aspirin had the
greatest protective effect in those with the
lowest cholesterol levels.
17
(No Transcript)
18
Other issues

• Intention to treat principle
• Effectiveness vs. efficacy
• Effectiveness refers to extent to which specific
  treatment produces intended results (also
  intention to treat analysis)
• Efficacy refers to potential effect of treatment
  under optimal conditions

19
Websites

• Society for Clinical Trialshttp//www.sctweb.org/
  
• Controlled Clinical Trials Journalhttp//www.el
  sevier.nl/inca/publications/store/5/0/5/7/5/8/
• Clinical Trials Discussion Threads and FAQs
  http//pharminfo.com/conference/clintrial/clnarc.h
  tml