Experimental Studies

1
Experimental Studies

• Two primary categories of experimental studies
• General difference is the unit of observation
• Community intervention trials
• Focus is groups or community outcome
• For evaluating large scale efficacy of new
  programs/policies
• Ex Fluoridation Trial, Stanford Five-City Trial
• Clinical trials
• Focus is on individuals
• More rigidly controlled than community trial

2
Experimental Studies

• Types of clinical trials
• Drug trials
• Treatment trials
• Example Aspirin trial
• Prevention trials
• Example HDFP (high BP detection and follow-up
  trial)
• Behavioral intervention trials
• Vaccine trials
• Hep-A Vaccine Trial in Monroe, NY
• Surgical trials
• Example Radical mastectomy vs. lumpectomy

3
Experimental Studies

• General features inherent to methodology
• prospective
• control group
• Intervention (drug, vaccine, surgical technique,
  etc.)
• investigator controls/ manipulates conditions
  studied (hence an experiment)
• assignment of exposure

4
Experimental Studies

• From Modern Epidemiology, 2nd Ed.
• Epidemiologic experimentation reduces variation
  by extraneous factors in comparison with the
  study factors

5
Efficacy vs Effectiveness

• Efficacy can the treatment work under ideal
  conditions
• more good than harm among those who receive it
• established by restricting study to patients who
  will follow medical advice

6
Efficacy vs Effectiveness

• Effectiveness can the treatment work under
  normal (ordinary) conditions
• more good than harm
• established by offering treatment to patients and
  allowing them to decide to accept or reject
• if treatment ineffective, could be due to lack of
  efficacy, acceptance, or compliance

7
Efficacy vs Effectiveness

• Compliance - the extent to which patients follow
  medical advice
• intervenes between an efficacious treatment and
  an effective one
• not just willful neglect of advice
• failure to understanding instructions, failure to
  keep prescription filled, money, and insurance
  coverage
• limit effectiveness regardless of efficacy

8
Strength of Evidence

• Weak Strong
• Case reports Randomized
• Case series trials

Observational studies
9
Study Design

• Randomized trials
• Randomized control study
• Cross-over trial
• Withdrawal study
• Factorial design study
• Group allocation study
• Large simple trial
• Multi-center trial

10
Study Design

• Non-randomized trials
• Concurrent non-randomized trial
• Historical control trial (nonconcurrent)
• Hybrid study
• Equivalency trial

11
Study Design

• Generally, the appropriate study design is
  dictated by the research question
• Regardless of design, assignment of exposure done
  by investigator (hence experiment)
• Ethical constraints may limit feasibility or
  dictate study design

12
Study design

• Origins from non-clinical settings (agricultural,
  industrial, laboratory) where investigator has
  greatest degree of control over
• subjects
• structure of study
• size of study

13
Study design

• Clinical trials on humans differ
• greater variability of response to treatments
• ethical constraints
• simultaneous testing of subjects not feasible,
  hence longer duration studies

14
Study design Phase I, II, III

• Licensing of a product generally requires three
  phases
• Phase I small group (lt100)
• Vaccine trial demonstration of response safety
• Drug trial establish maximally tolerate dose
  (MTD) dose range identification of toxic
  effects
• Phase II larger group (100-200)
• Vaccine trial assess antibody response and
  reactions
• Drug trial - efficacy in dose range validate
  Phase I
• Phase III assess efficacy and effectiveness

15
Randomized Control Trials

• Assignment of intervention by randomization
• Comparison of effect among intervention group to
  no effect in control group
• Equal probability of participants to be assigned
  to intervention or control group
• Gold standard to which all other trials are
  compared

16
Randomization

• Simple definition
• each participant has equal probability of
  assignment to intervention or control group
• random allocation of subjects is the basis for
  statistical inference

17
Randomization technical considerations

• Comparability assure comparability with respect
  to known and unknown risk factors
• Bias minimize (or eliminate) bias
• removing control of assignment from investigator
  non-discoverable process
• guarantees that treatment assignment not based on
  patient prognostic factors
• Validity quantify errors attributable to chance

18
Bias associated with randomization

• Selection bias occurs if method of allocation
  is predictable could influence participation if
  anticipate treatment assignment
• Accidental bias if balance on risk or
  prognostic factors is not achieved more a
  problem in small studies
• Miscellaneous sources of bias
• treatment administration
• outcomes assessment

19
Confounding and randomization

• Confounding prevented by randomization
• some argue randomization is unnecessary since can
  be controlled for in the analysis
• Assumption 1 all important confounders are
  known and measured
• Assumption 2 statistical models, adjustment
  procedure assumptions are correct
• better to prevent than control

20
Random allocation methods

• Fixed allocation
• Simple
• Blocked
• Stratified
• Adaptive procedures
• Baseline adaptive
• Response adaptive

21
Masking (blinding)

• To avoid systematic differences between the
  treated and control groups during the course of
  the study in factors other than the one under
  study
• controlling for information bias
• Double-blind
• Patient, care provider, and investigator (person
  collecting data) all unaware of treatment
  assignment
• Single-blind
• Patient unaware of treatment assignment

22
Administration

• Randomization and treatment assignments should be
  administered by an independent unit
• Larger trials use computerized log-in and
  eligibility verification prior to randomization
• Regardless of system used, randomization and
  subsequent treatment assignment should be done
  when participant ready to begin the intervention

23
Strengths

• Potentially the most rigorous design for
  evaluating the effect of a single controlled
  exposure or intervention
• When well executed, less subject to issues of
  interpretation than non-experimental studies
• More reproducible than non-experimental studies
  (theoretically)

24
Limitations

• Applicable only to certain types of questions, at
  a particular stage of knowledge about the problem
• Complex planning
• Usually expensive
• Ethical considerations may limit selection of
  participants and generality of results
• Rigor of the conduct may fall far short of the
  intent

25
Main Features of Clinical Trials

• Question
• Design
• Randomized
• Blind
• placebo-controlled
• setting
• Study Population
• time, place person, sample size
• inclusive, exclusive criteria

26
Main Features of Clinical Trials

• Intervention (independent variable)
• End point, Outcome (dependent variable)
• definition, measurement
• follow-up protocol (data collection)
• Analysis
• description (comparability, adherence)
• effect on outcome variables
• Conclusions (Interpretation)

27
Selection of population

• The population consists of those individuals who
  meet the eligibility requirements as to
• the presence of the condition
• risk of events for treatment evaluation
• free of contraindications to the treatment
• available (including specified consent
  procedures) for the duration intended for
  observation
• They may be selected in regard to likelihood of
  compliance with the intervention.
• Once the population is selected, individuals are
  randomly allocated into treatment groups

28
Selection of population

• Comparison of outcomes between treated and
  control groups is the essence of a clinical trial
• Historical controls may be used in those rare
  instances where the natural history of the
  disease has been invariant, e.g., tuberculosis
  meningitis
• Concurrent controls are required for most
  questions
• Often the control group is given the "best
  available treatment" which is evaluated in
  relation to the new treatment under evaluation

29
Data collection

• Independent variable (treatment)
• Intervention to be investigated, with placebo or
  alternative active treatment
• Dependent variable (outcome)
• Events or conditions of importance are those to
  be affected by the treatment under investigation,
  i.e. the end point(s)

30
Data collection

• Hard end-points
• Objective measures can be measured
  objectively, blood tests, death, etc
• Soft end-points
• Subjective measures reported by the observer or
  patient (examples degree of disability, pain,
  toxicity, etc.)
• Often of greater importance to the patient,
  largely neglected by researchers
• Self-reporting of adverse events more likely to
  reveal adverse effects, especially if rare

31
Data collection

• Target population is narrowed down considerably
  based on pre-determined criteria to ultimately
  establish the study population
• Has potential of magnifying problems due to loss
  to follow-up, competing risks/outcomes over the
  course of the study, unexpected treatment changes
• To assess this potential effect, data collected
  as to comparability of groups at entry, adherence
  to the intervention, and status concerning
  outcomes as indicated by the occurrence of end
  point events or conditions and possible side
  effects/toxicity of the intervention

32
Data analysis

• Quality control
• addresses reliability of observations, extent of
  missing data, departures from design
• Description - provides evidence on
• comparability of groups at entry
• adherence to intervention
• frequencies of recognized side effects/toxicity
• end point rates
• All categorized as at entry or otherwise in
  accordance with the design

33
Data analysis

• Hypothesis testing
• characterization of study participants,
  especially as to comparability of groups at entry
• direct measures of incidence of end points or
  mortality/survival using life tables or survival
  analysis
• may involve multivariate analysis to adjust for
  differences at entry changing status of
  participants over time (time-dependent
  covariates)
• special analytic methods (sequential analysis) to
  determine basis for early termination versus
  continuation may be employed

34
Interpretation of results

• Conclusions are chiefly centered on the validity
  of the comparison of groups over the course of
  the trial,
• the nature and importance of the differences
  observed
• the relation of benefit to risk
• the generality of findings to others

35
Interpretation of results

• In a clinical trial, observed differences in
  outcome between the treated and control groups
  can be assigned to one or more of the following
  categories
• Sampling variation (chance)
• Pre-existing differences in factors affecting
  outcome (selection bias and confounding)
• Differences in the management and evaluation
  during the course of the study (information bias)
• True effects of the treatment (truth)

36
Hepatitis A Vaccine Trialin Healthy Children

• The first double-blind, placebo-controlled trial
  in an American community with recurrent outbreaks
  of hepatitis A
• The study community (Monroe, NY) was
  characterized by rapid growth, large families,
  and a high seropositive rate (68) for hepatitis
  A antibodies among adults gt19 years of age
• Following demonstration of protection with a
  single dose, booster doses were administered to a
  subset of study participants at 6, 12 or 18
  months after the primary dose of VAQTA

Werzberger, A. et al. N. Engl. J. Med.
327(7)453-457, 1992
37
Hepatitis A Vaccine Trial

• Study Participants
• 1,037 healthy children, ages 2-16, who were
  seronegative for anti-HAV antibodies were
  randomized to receive 25 U/O/0.5 mg/L of VAQTA
  (n519) or placebo (n518) before the beginning
  of a summer-fall outbreak of hepatitis A

38
Hepatitis A Vaccine Trial

• Efficacy
• 100 protection against hepatitis A was
  demonstrated in 519 susceptible children (ages
  2-16) following a primary dose of VAQTA (plt0.001)
• Efficacy was based on clinically confirmed cases
  of hepatitis A occurring 50 days or more after
  vaccination to exclude children incubating the
  infection before vaccination
• In fact, no cases of clinically confirmed
  hepatitis A were observed in the vaccine group
  after day 16, while 34 cases were confirmed in
  the placebo group

39
Efficacy and Safety of Recombinant Human
Activated Protein C for Severe Sepsis

• Early Release (from online journal) Because of
  its possible clinical implications, this article
  is being released before its publication date.
  The final version of the report will be published
  on March 8 (posted February 9)
• The New England Journal of Medicine
• Volume 344 March 8, 2001 Number 7
• Original Article
• Gordon R. Bernard, Jean-Louis Vincent,
  Pierre-Francois Laterre, Steven P. LaRosa,
  Jean-Francois Dhainaut, Angel Lopez-Rodriguez,
  Jay S. Steingrub, Gary E. Garber, Jeffrey D.
  Helterbrand, E. Wesley Ely, Charles J. Fisher,
  Jr., for the Recombinant Human Activated Protein
  C Worldwide Evaluation in Severe Sepsis (PROWESS)
  Study Group

40
Efficacy and Safety of Recombinant Human
Activated Protein C for Severe Sepsis

• Research question
• Does activated protein C (drotrecogin-?
  activated) reduce the rate of death from all
  causes at 28 days in patients with sever sepsis
  and have an acceptable safety profile?
• Phase III Trial
• Study Design
• A randomized double-blind placebo controlled
  multi-center trial
• Outcome
• Mortality (any cause) at 28 days post treatment
• Coordinating Center Vanderbilt

41
Efficacy and Safety of Recombinant Human
Activated Protein C for Severe Sepsis

• Methods
• Population
• 1,690 patients with systemic inflammation and
  organ failure due to acute infection from 7/1998
  through 6/2000
• 164 centers in 11 countries
• Randomization
• Stratified-block at each center on a 11 basis
• assigned to receive either drotrecogin-?
  activated _at_ 24 ?g per Kg body weight or placebo

42
Efficacy and Safety of Recombinant Human
Activated Protein C for Severe Sepsis

• Methods
• Evaluation
• Followed for 28 days or until death
• Blood samples at days 0, 1-7, 14, and 28
• All measurements performed by a central
  laboratory
• Statistical analysis
• Stratified based on
• APACHE II Score (Acute Physiology and Chronic
  Health Evaluation II) 4 strata
• Age 2 strata (lt60 vs 60 yrs)
• Plasma protein C level 4 strata

43
Efficacy and Safety of Recombinant Human
Activated Protein C for Severe Sepsis

• Methods
• Statistical analysis (continued)
• Designed to enroll 2,280 patients
• Two planned interim analyses _at_ 760 and 1,520
  enrollees
• Guidelines to suspend trial if treatment was
  significantly better than placebo were determined
  a priori
• Analysis performed by independent DSMB (data
  safety monitoring board)

44
Efficacy and Safety of Recombinant Human
Activated Protein C for Severe Sepsis

• Results
• Enrollment suspended at second interim analysis
  (data from 1,520 patients)
• Efficacy
• Statistically significant higher rate of survival
  in treatment group compared to placebo group
  (plt0.001 at the ?0.05 level, two-sided test)
• 19.4 reduction in Relative Risk of death
• Absolute reduction of 6.1
• Safety
• Incidence of bleeding higher in treatment group
  (3.5 vs 2.0) although not statistically
  significant (p0.06)

45
Efficacy and Safety of Recombinant Human
Activated Protein C for Severe Sepsis
46
Efficacy and Safety of Recombinant Human
Activated Protein C for Severe Sepsis
TABLE 1 BASE-LINE CHARACTERISTICS (CONTD)
47
Efficacy and Safety of Recombinant Human
Activated Protein C for Severe Sepsis
48
Efficacy and Safety of Recombinant Human
Activated Protein C for Severe Sepsis
FIGURE 2 KAPLAN-MEIER ESTIMATES OF SURVIVAL
49
Efficacy and Safety of Recombinant Human
Activated Protein C for Severe Sepsis

• Conclusion
• Treatment with drotrecogin-? activated
  significantly reduces mortality in patients with
  severe sepsis
• Treatment may be associated with increased risk
  of bleeding
• In this population, 1 in 16 patients would be
  saved

50
Additional References

• Text
• Fundamentals of Clinical Trials, 3rd Ed.,
  Friedman, Furberg, DeMets
• WWW
• http//www.clinicaltrials.gov
• http//www.fda.gov/cder/