Transcript Outline

Ethics in Clinical Trials
• Ethical concerns in general
• Ethical concerns in trials
• Interesting examples
Ethical Concerns in all Research
• RQ should be important and answerable
• Benefit should outweigh risk
• Participants should give consent
• Privacy should be protected
• Research report should be fair and honest
Ethical Perspectives
• Idealism
– human beings are special
– can never be a means only
• Utilitarianism
– greatest good for the greatest
number
Roles
• Physician - Investigator
• Patient - Participant
• Treatment - Research
HISTORY
• Nuremberg Code (1947)
• Declaration of Helsinki, WMO (1964 - 97)
• NIH Ethical Review Policies (1966)
• National Commission for the
Protection of Human Subjects (Belmont
Report, 1979)
• DHHS Regulations, 45 CFR 46 (1981-91)
Nazi Experiments
THE NUREMBERG PHYSICIANS’ TRIAL
• “Participants” were placed in freezing
water and time to death was measured
• “Participants” drank sea water and
adverse effects were measured
The Nuremberg Code (1947)
• Voluntary participation
– legal capacity to give consent
– free of force, fraud, deceit, duress
– free to withdraw at any time
• Fruitful results for society
– unprocurable by other means
– conducted by qualified persons
• Avoid unnecessary risk to subjects
– risk not greater than importance of RQ
Post WWII Trials
• US Atomic Energy Commission
tests of the adverse effects of
radiation exposure
• Clinical trials in federal and state
prisoners
Tuskegee Syphilis Study (1932-72)
• Prospective cohort funded by USPHS
• 600 poor, illiterate, black men
– 399 with syphilis; 201 without syphilis
– followed for 40 years for tertiary syphilis
• Never informed of condition
• Never treated
Ethical Principles
• Beneficence
• Respect for autonomy
• Truth-telling
• Justice
• Promise-keeping
• Privacy
World Medical Association
Declaration of Helsinki (1964)
• Voluntary participation with consent
in writing
• Design described in written protocol
• Review by an independent committee
• Responsible scientific publication
• Protection of vulnerable populations
Institutional Review Boards
• NIH required ethical review of internal studies
in1953 and funded studies in 1966
– led to establishment of local IRBs
• Oversight by NIH Office of Human Research
Protections based on Federal regulations
– risks to subjects minimized and reasonable
– informed consent in writing
– provisions for privacy
– safeguards for vulnerable populations
– selection of subjects equitable
Federal and Local Regulations
• UCSF - Committee on Human Research
– you can volunteer to serve!
• NIH - Office for Human Research Protection
–http://ohrp.osophs.dhhs.gov/polasur.html
• Code of Federal Regulations Title 45, Part 46
–http://ohrp.osophs.dhhs.gov/
humansubjects/guidance/45cfr46.htm.
What’s Special about RCTs?
• Randomization - equipoise
• Intervention - relatively safe
• Placebo - acceptable clinical option
• Measurements - safe and tolerable
• Interim monitoring - careful and timely
Equipoise
• Question important and not answered
• Evidence of benefit, not conclusive
–trial of new drug treatment for advanced
breast cancer
–trial of treatment for common cold
Intervention and Control
• Maximize benefit, minimize harm
–intervention (minimal effective dose)
–control (placebo acceptable?)
• Qualified staff and protections for
known potential harms
–manage known adverse effects
–pay costs of known adverse effects
• Identify associated harm
Measurements Safe and Tolerable
• Trial of estrogen for fracture prevention
–substitute TVUS for endometrial biopsy
• Trial of accuracy of spiral CT for PE
–all get spiral CT and pulmonary angiogram
• Trial of effect of estrogen treatment on
coronary atherosclerosis (ERA)
–randomized to estrogen or placebo
–coronary atherosclerosis on angiograms
Fecal Occult Blood Testing
Kronborg, et al., 1996
• Randomized, controlled trial
• 60,000 persons in Denmark
– identified via central records
• FOBT biannually or usual care
– request for FOBT mailed by PMD
• Outcome = colon cancer
– based on national registry
Informed Consent
•
•
•
•
•
•
•
•
Purpose of trial
Why asked to participate
Visits, procedures, time and costs
Discomforts or risks
Benefits to subject and society
Alternatives
Confidentiality
Contact for questions, problems
Active Compression-Decompression for
CPR, Schwab et al., 1994
• Randomized, controlled trial
• 860 persons with cardiac arrest
• ACD CPR or standard CPR
• Outcome = discharged alive
• No informed consent
• Trial halted by FDA
SPECIAL POPULATIONS
• Children
• Fetus
• Mentally disabled persons
• Institutionalized persons
• Prisoners
• Unconscious or severely ill persons
Alternatives to Informed Consent
• Waiver of consent
– life threatening situation
– consent not possible
• Permission from parent or guardian
• Deferred informed consent
– enter study without consent
– later consent or participation terminated
• Prospective consent
Cumulative Meta-analysis
Effect of beta-blockers
on mortality after MI
Antibiotics for Abortion
Sawaya, et al, 1996
• Cumulative meta-analysis 12 RCTs
– after 5 trials (1985), summary RR 0.5,
p<.05
– 7 additional trials performed
• findings of 5 trials non-significant
• trials continued up to 1993
Zalospirone for Depression
Rickels, et al, 1996
• Randomized trial
• 287 people with major depression
• Placebo or 3 doses of drug
• Outcome - change in severity of
depression
• High dose effective; two lower
doses not
Prevention of AIDS in Africa
• Standard care for HIV+ pregnant
women in US
– zidovudine orally before delivery
– IV during labor, then orally for newborns
– RR .33 for infection in newborn
• Pregnant HIV+ African women
– randomized to oral AZT or placebo
– most funded by US agencies
Cardiac Arrhythmia Suppression
Echt, et al., 1991
•
•
•
•
•
Randomized trial
1498 patients post-MI with PVCs
Encainide, flecainide, or placebo
Outcome = death
Trial stopped after 1 year due to
increased deaths in treated
group; p = .004
MORE Trial
Grady, et al., 2003?
•
•
•
•
•
Randomized, placebo-controlledtrial
7704 women with osteoporosis
raloxifene or placebo
Outcome = morphologic spine fx
Increased risk of VTE (RR=2.5)
• first year RR=6
• second year RR= 5
• no increased risk thereafter
Conflict of Interest
• Professional judgement unduly
influenced by other interests:
– reputation
– promotion
– financial interests
• ownership, stock, gifts, consulting
fees, travel and entertainment,
research support
Ethics in Randomized Trials
• Informed consent isn’t enough
• Important ethical issues in all stages
of the design, conduct and
presentation of findings of RCTs
• Easy to get sucked into thinking that
your research is more important
than participant safety