Transcript RANDOMIZED CLINICAL TRIALS

RANDOMIZED CLINICAL
TRIALS
What is a randomized
clinical trial?
 Scientific investigations: examine and evaluate the safety and
efficacy of new drugs on human volunteers.
 Compare a new drug vs. existing drugs or placebo.
 designed in an ethical: voluntary consent.
 The results: the most valued data in the era of evidence-based
medicine.
Classification of clinical trials
Phases:
 Phase I, safety – tolerability – pharmacokinetic/dynamics –
dose/ response.
 Phase II, Dosing – Efficacy
 Phase III, Effcacy compared to gold standard treatment.
 Phase IV, post-marketing - pharmacovigilance.
Phase
I
Phase
II
Phase
III
Phase
IV
Classification of clinical trials
Trial design:
 Parallel-group, followed-up to determine the effect of each treatment
in parallel groups.
 Crossover, patients to different sequences of treatments.
 Factorial, assign patients to more than one treatment-comparison
group.
 Cluster randomized, larger groups are randomized instead of
individual patients.
Classification of clinical trials
Number of centres:
 Single-center.
 Multicenter.
What the study was designed to prove:
 Superiority study, new drug is more effective than
placebo.
 Equivalence study, how that two drugs have the same
clinical benefit.
 No inferiority study, new treatment cannot be said to be
significantly weaker.
Erroneous results
Bias/Systematic errors
 Design, conduct, analysis, and the results  associate
the occurrence of systematic errors for influenced by
bias.
 The bias can also make the trial-derived estimate of a
treatment effect deviate from its true value.
 Types of bias in clinical trials:
 Related to subject selection.
 outcome measurement.
 estimate bias.
Erroneous results
Random error
 The sampling error is produced because the
sample might yet show a chance false result.
 Reduced:
 choosing a very large group of patients.
 special analytic techniques (meta-analysis).
 Statistical testing involves an evolution of the
probability of obtaining the observed treatment
difference.
Erroneous results
Confounding
 Distortion of the real relationship: treatment and outcome by
another factor.
 Eliminate bias: the solution is randomization in conjunction.
 Before start the trial  delete the bias of confounding across
randomization techniques:
 stratified randomization confounding.
 regression analysis.
How develop a randomized
clinical trial?
1) Objectives and endpoints
 Clear objectives that are measured by endpoints.
2) Study design
 Type of trial design
 Minimize bias
3) Patient population
 Define number of patients.
 Exclude some patients.
 Informed consent.
How develop a randomized
clinical trial?
4) Sample size calculation
 used for minimize random error.
 For calculate: statistical models that take account of
the recruitment of patients into the trial and the type
of statistical test to be used.
5) Conduct of the trial
 Define the initial dose used of the study drug.
 the investigators recruit patients for the trial.
 Management, and coordination.
 Visits will scheduled every for X time.
How develop a randomized
clinical trial?
6) Interim monitoring
 Randomly assign patients: an independent statistician centre
(DSMB).
 DSMB has to make sure that the new drug that patients are
taking is not damaging.
7) Final data analysis
 All endpoints can analyse: the log-rank test, stratified and
displayed on Kaplan–Meier plots.
 Statistical analyses are usually stipulated in the trial protocol.
8) Trial reporting
 Results publish different reports, which follow the CONSORT
(Consolidated Standards of Reporting Trials).