- Information
- AI Chat
Was this document helpful?
Research ethics 2 - Lecture notes 9
Course: Bio-Medical Ethics (PHI 325)
53 Documents
Students shared 53 documents in this course
University: North Carolina State University
Was this document helpful?
Research ethics (2)
How we do medical research now: the science of clinical trials.
Most medical research involving human subjects takes form of CLINICAL TRIAL.
1. Clinical Trial: description
Scientific study designed to test a medical intervention in humans (Vaughn, 194/222).
Aim is to determine treatment’s effects, especially efficacy and safety.
Effectiveness and side effects of treatment are tested by administering it to a human subject.
Clinical trials provide strongest and most reliable evidence of treatment’s efficacy (better
than animal trials or clinical observation).
oThey are carefully designed to eliminate bias, avoid errors.
The “gold standard”: Randomized clinical trials (RCTs)
oIn era of scientific medicine: RCT is standard technique for testing and changing
diagnostic and therapeutic methods.
oProcedures to rule out bias:
“Double-blind” test design:
Neither investigators nor subjects know who is receiving drug being
tested: “blinding” (addresses “treatment bias”).
One group of patients given drug being tested (experimental
group), and the other (control group) placebos (nowadays, more
often established drug that is the “standard of care”).
To get useful information, researchers must study the relevant
differences between the two groups that emerge AFTER
intervention:
oWithout control, can’t tell whether result is due to treatment
or some other factor (including “placebo effect”); with
control group, there is more reason to believe that relevant
effects in experimental group due to treatment.
oFor treatment to be judged effective, subjects in treatment
group must show significantly greater improvement than
those in placebo group (placebo-controlled trial) or those
receiving standard treatment (active-controlled trial).
Randomization: assigning subjects randomly to groups.
oControl and experimental groups share relevant
characteristics. (Minimizes bias resulting from assigning
subjects to particular groups, i.e., selection bias.)
oMinimizes possibility that unconscious bias leads to
placing preferred subjects into a particular group.