Research Ethics Flashcards
Beecher’s article
1966, argues that medical research involves unethical designs. such studies shouldn’t be published. researchers need consent and also need to be responsible.
- Article gave people a sense of the reality of vulnerable populations.
- Rothman’s article comments on this, argues that WW2 laid foundation for utilitarian mentality and thus exploitative research.
SFNO and criteria for choices
stakeholders, facts, norms, options
procedural. layus out elements of ethical situation, but doesn’t give solution. prudence needed.
disagreement comes bc diff pppl have diff interests, disagree about facts, disagree about norms. Criteria = effectiveness, necessity, proportionality, least infringement, transparency
Jonas on research
research uses subject as a passive thing therefore is problematic. we overstate the justification for research. if people are going to self-sacrifice, it is because it comes out of a sense of the holy.
MEdical progress is not necessary, but an optional goal. if progress slows, society is not harmed. proceed by rule of descending order (researchers are subjects first, poor last)
phases of research
1 - toxicity or side effects
2 - tests efficacy and safety; small number of subjects
3 - tests safety and efficacy on larger number of subjects; RCTs
4- post-marketing surveillance trials after drug is on the market.
Eisenberg
research is a social imperative bc it relieves suff and death. it shouldn’t be regulated to constrain progress, because this will also be harmful to people.
belmont report - principles and their applications
respect for person (acknowledge self-determination and protect those with < aut)
beneficence - don’t harm, maximize benefits minimize burdens
justice - distribution of research benefits and burdens; selection of subjects should be fairi; those who have burdens need access to benefits.
—–
applications:
Informed consent; balance benefits and burdens; select subjects so that there is ind justice (all have chance to participate) and social justice (no group has disproportionate benefits/burdens)
Informed consent (belmont report vs Levine)
Belmont: information (disclosure), comprehension, voluntariness
levine: IC has four requirements: competent. voluntary, informed, comprehending
Miller on therapeutic misconception
TM is of incremental, not categorical, concern. It’s ideal to have no TM, but it does not necessarily invalidate consent or exclude participation.
TM is unacceptable when the person would ahve made a different decision without TM.
beachamp on autonomous action
conditions for autonomous action:
intentional (corresponds to what is willed, including neg consequences)
understanding
voluntariness (no one is totally voluntary)
*understanding and voluntariness vary in degrees. we can make cut-offs, but not clear categories. we have to balance.
Jansen on unrealistic optimism
optimistic bias is present and is not related to misunderstanding. optimism is an independent cause of expecting therapeutic effects of research, and can’t be addressed with more information. it has to be addressed at a higher level about research
Miller and Wertheimer on consent
The central Q about consent is not if its valid, but whether a bilateral consent transaction between 2 people is morally transformative. moral transformation comes from fair transaction, not autonomous authorization.
- consent transaction = interaction that lets a do x to b.
- lock and key /autonomous authorization model - valid consent is necessary and sufficient. valid consent is voluntary, informed, understanding, competence
- *They propose fair transaction model - A can proceed on basis of B’s consent if A has treated B fairly and responds reasonably to B’s consent/appearance of B’s consent. Flawed but fair consent is OK.
- criteria for moral transformation is different in diff contexts
application - their model allows for research when there is a ther misconception, whereas valid consent (lock and key) model does not.
Miller and Wertheimer on paternalism in research
research regs are paternalistic (Soft) and justafiably so.
Soft paternalism justifies protective regs. warns against nonvoluntary judgment of subjects and defect in DM that make a choice less autonomous.
paternalism doesn’t necessarily conflict with atunomoy
kinds of paternalism (Miller and WErt)
soft - limiting liberty bc person’s DM is nonvoluntary or nonautonomous
hard - limiting liberty of someone not impaired for their own good
direct / individual- limiting liberty of person you want to protect
indirect / group - limiting liberty of one person to give choices/protection to another
minimal risk research
probability and magnitude of harm is not greater than those ordinarily encountered in daily life (the q is whose daily life)
Kim argues that IRB review is not of sufficient value for min risk studies.
Edgar and Rothman on IRBs
current process needs to be more nationalized; Suggestions:
*we need a national ethics committee to address the broader political sanction of research
improve irbs to account for entrepreneurial nature of research
strengthen outsdie elements of IRBs but leave review in the institution.
Rid, Emmanuel, and Wendler on assessing risk
right now, we rely on intutiion to assess risk. they propose SERR method to assess risk in a more standardized way. (Systematic Evaluation of Rsrch risks)
- identify harms
- categorize magnitude of harms with harm scale
- estimate likelihood of potential harms
- compare potnetial harms with harm of comparable activity. if similar, risks may be acceptable.
Slovic on risk
most people perceive and assess risks without an emphasis on mortality rates. empirical evidence won’t convince them.
people assess risk with characteristics like familiarity, control, catastrophic potential, threat to future genertions. people tolerate more risk for things they see as beneficial.
Williams on IRBs
IRBs are bad at assessing risks and benefits compared to their ability to protect rights.
group vs. community
group = exists bc a shared trait, doesn't have structured leadership community = group that has organizational structure and leadership regardless of research; may form around a shared trait.
privacy vs confidentiality
privacy = what people don’t want researchers to have or know. deals with what can be withheld, what should not be gained without permission.
confidentiality = management of private information.sharing private information appropriately, protecting it.
HIPAA and research
hipaa privacy rule doesn’t apply directlyt o research, but it does specify circumstances under which providers can disclose info. HIPAA is not harmonized with the Common Rule. hipaa forms discourages subjects from enrolling in research. hipaa does not give privacy protection to human subjects, but creates many obstacles to recruitment (and therefore creates a selection bias)
deidentification
info that has been altered to remove certain data elements with an individual. between anonymous (has no direct/indirect identifiers and can’t be linked to an ind) and personal information (which is identifiable)
deidentification may not be sufficient bc:
who is doing this?
easy to re-ID people
group harms
objectionable uses (abortion, cloning, etc)
commercial exploitation (Henrietta Lacks, etc)
trust/transparency
public opinion of research
Rothstein - on deidentifiable information
identifiable info has legal protections; deidentified info does not.
Rothstein we needs a continuum of protections, not black and white categories.
Privacy rule and common rule are inconsistent about deidentified info - common rule - anonymous or deidentified info is exempt. privacy rule - it’s exempt only if consent is first given. and its’ exempt only for a limited purpose.
Rothstein argues that de-IDed info has lots of risks to inds and groups; so, current system needs revision and clarification.
certificates of confidentiality
make researchers exempt from disclosing PHI in civil, criminal, administrative, legislative, or other proceeding.
in reality, they haven’t always worked this way. courts don’t always honor them.
HAvasupai
DNA samples given to ASU researchers to help with diabetes. then used for other things, like mental illness and discovering the tribes’ origins. found that they were not consistent with tribal stories.
- IC was not sufficient
- IRBs have to consider group harms
- Ended in a settlement
Tuskegee
syphilis study
Coleman on vulnerability
people are vulnerable to consent, risk, and justice.
typically, vulnerable is applied to groups - pregnant women, kids, prisoners, poor, etc. the assumption is that you are vulnerable bc you are part of a group or population. OR, bc of consent (you aren’t competent)
But, there are consent based vulnerability (which soemtimes can be remedied)
risk based vulnerability
justice based vulnerability - ex, rsrch in other countries
Kipnis on vulnerability
we shouldn’t think of vulnerability in subpopulations, but through an analytical approach that considees circumstances.
Six kinds of vulnerabilty:
1 cognitive (capacity to consent)
2 juridic (is participant subject to authority of others who have interest in their participation?)
3 deferential (does subject have readiness to accede bc of external factors or patterns of deference/submission?)
4 medical - (subject is selected bc she’s in desperate health condition)
5 allocation - subject lacks social goods, and this lack leads her to participate.
6 infrastructural vulnerability - does the political, organizational, social, or economic context contain integrity? (ex, rsrch in communities without medical access)
IOM report on prisoners
there’s a need to expand protections to prisoners.
- expand def of prisoner to include those on parole or probation
- shift from category to risk based approach to research review (and therefore disallow phase 1 and 2)
- enhance oversight on prisoner research
Chwang on prisoner research
additional risk-benefit constraints on prisoners are not necessary. Common Rule and IOM recs are enough.
-IOM compares prisoners and kids. Prisoners are not kids.
Subparts of federal policy for protection of research subjects
Subpart a - common rule
b pregnant women, fetuses, neonates
c - additional protections for prisoners
d - kids
undue inducement
the perceived benefits are such that an individual’s ability to judgment rationally (or at least in a way that we see fit) is impaired.
Wilkinson, Martin and Moore on inducement
money is not an undue inducement, bc it’s not forcing participation. choosing to accept inducement for research makes good sense in some cases. it may be a very rational choice for a poor person to participate in research. Desperate need does not undermine consent, because consent is about autonomy, not freedom. Freedom requires multiple options, whereas one can choose to act autonomously without multiple options. (taking a med to save your life is autonomous, and you have no other choices).
- Emanuel also argues that undue inducement claims are bogus and actually usually mask three more legitimate concerns (excessive risks were missed by IRB, poor informed consent, unfair subject selection/justice).
- In contrast, McNeill argues that society has an obligation prior to protecting research subjects autonomous action, to establish basic safeguards that are fair. It’s not fair that the poor are encouraged to expose themselves to harm more than the not poor populations.
- Lemmons and Elliott even argue that healthy human subjects should be treated like laborers - they could even unionize and demand things.
models of payment for research subjects (Dickert and Grady)
market model - pay well, be competitive so that you attract subjects
Wage payment model - pay like this is a job
Reimbursement model - pay for expenses incurred
Wertheimer on exploitation
there’s a difference between harmful nonconsensual exploitation and mutually advantageous/consensual exploitation.
**we have a prima facie duty not to interfere with autonomous choice to be exploited. we must prohibit harmful and nonconsensual exploitation – this does not mean that the poor should be excluded.
In fact, participation in research amongst the poor for their own benefit can be mutually advantageous and thus it is hard to justify interference.
vivisection
cutting into a live organism
human vivisection - experiments on humans not for their benefit - used until 1930s
Emmanuel’s principles to ensure research is ethical
social/scientific value, sci validity, fair subject selection, favorable risk ben ratio, ind review, informed consent, respects subjects,
(informed consent is not enough and it’s not necessary. all criteiria are necessary except IC and IRB review.
what does empirical research reveal about informed consent?
give info in parts
education improves DMC
written material and pics inc understanding
ppl can’t guess the wishes of another
people can’t understand risks, benefits, or standards
therapeutic misconception
types of review IRB
exempt
expedited (1 reviewer)
full board
types of approval IRB
full approval
approval with contingencies
tabled/deferred
disapproved
sources of variability in IRBs
local circumstances
members’ background / judgment
why are IRBs bad at balancing risks and benefits?
bias in favor or approval
composition of IRBs - they arlready value rsrch
psychological effect of committee decisions (don’t want to polarize, they know researchers)
what are the differences btw hipaa and common rule?
hipaa - more demanding, implies > risk than actually exists; hipaa limits use of identifiable data sets.
common rule - anonymous or deidentified info is exempt. privacy rule (hipaa) - it’s exempt only if consent is first given. and its’ exempt only for a limited purpose.
deidentified vs anonymous
D - info altered to remove certain data elements with individuals;
A - no direct or name, can’t be linked to me
Degrees of protected information (high to low)
anonymous, anonymized (made anonymous), deidentified, linked by source, linked by researcher, direct identifiers.
Kipnis - How should we respond to vulnerability?
reduce vulnerability, involve less vulnerable subjects, distribute benefits to participants, community consultation
Kass’ recs for pregnant women
allow phase 2, expand phase 3, place burden of proof on IRB to include women.
4 levels of risk w kids
minimal risk
greater than minimal risk but gives benefits to subjects
minor increases other than minimal risk but will increase gen knowledge,
and not otherwise approvable
Manson and O’Neill on protecting genetic information
genetic information needs to be protected; however, rather than worry about regulating genetic information, we should promote an agency based model of informing nad communicating about genetic info. we should focus on the actions that lead to the acquisition and use of genetic knowledge. This focus does not give gentic information a different status, but focuses on obligations that are broader and more thorough.
3 ethical considerations of human whole-genome sequencing (McGuire)
circumstances under which research results are disclosed to participants
obligations, if any, that are owed to participants close genetic relatives
options regarding how future uses of samples and data taken for sequences are managed.
problems with giving patients access to genetic research (McGuire)
- results are subject to misinterpretation
- process of disclosure and follow up care requires > training for providers
- how do you store this info in health record?
*his rec is to only put info of known clinical relevance in health record.
kinds of consent to use of biospecimens (Maschke)
broad consent (blanket consent for all uses) specific consent (only for a certain purpose) tiered consent - permit use in current study only but allow contact for future use
Challenges to standard consent to genetics/biobanking
- right to withdraw consent
- separation in time and space btw collection and research
- risks/bens are often group based
database vs biobank
database - contains only genetic/clinical info
biobank - contains actual samples, with or without information attached
genetic exceptionalism; arguments for and against?
genetic information is special and must therefore be treated differently from other types of medical information
Greenberg v miami childrens
inds do not have ownership rights in specimens obtained in rsrch
wash u v catalona
donors don’t have control over samples beyond consent and withdrawal
unidentified vs unlinked
anonymous
anonymized
CLIA
Clinical laboratory improvement amendments; CMS regulatory system that covers lab tests on humans in the US; lacks quality control.
How does Ruth Macklin propose that we understand ethics in international research
it’s not the cultural relativity of values that is the problem, but different facts and circumstnaces surrounding the research. So, we should have the same ethical principles, but adapt different procedural requirements for diff circumstances. So principles do not change; they are standard in all cultures and times. there’s a major difference btw universal principles and the idea of exceptionless principles. she’s saying there are universal principles.
–so the Q is not is it okay to do experimental treatment in developing country if that isn’t approved in developed country, but do the risks, benefits, and circumstances in that country make the experiment morally justifiable.
Quinacrine trials
.
interaction principle
(Wertheimer)
an investigator is ethically bound to provide super-contractual obligations to a subject by entering into a rltshp w the subject, even if the subject doesn’t agree and even if the investigator wouldn’t have an obligation if he didn’t enter the contract.
Those who promote the interaction principle argue that it could be morally worse for the researcher to interact with the subject than to not interact with the subject even if the interaction is beneficial and consensual (so they reject the nonworseness claim)
Those who promote the principle also argue that the researcher has added obligations (greater obligation claim)
-there are 5 types of greater obligations - responsiveness (people in diff countries may agree to research that doesn’t benefit them), standard of care, post-trial obligations, ancillary care (feeding starving people, treating conditions you notice during study), fair benefits (building infrastructure and capacities in the place we are testing)
Wertheimer rejects this. He argues that mutually advantageous and consensual transactions may be morally ok if they are better for all than non-interactions (International research paper)
nonworseness claim
(Wertheimer on international research) it cannot be moreally worse of A to interact with B than not to interact with B if the interaction is beneficial to b and b consents
belmont report
- The Belmont Report [1] summarizes ethical principles and guidelines for research involving human subjects. Three core principles are identified: respect for persons, beneficence, and justice. Three primary areas of application are also stated. They are informed consent, assessment of risks and benefits, and selection of subjects
Prompted in part by problems arising from the Tuskegee Syphilis Study (1932–1972)
weakness: does not specify how its three ethical principles should be weighted or prioritized
belmont report
The Belmont Report [1] summarizes ethical principles and guidelines for research involving human subjects. Three core principles are identified: respect for persons, beneficence, and justice. Three primary areas of application are also stated. They are informed consent, assessment of risks and benefits, and selection of subjects
Prompted in part by problems arising from the Tuskegee Syphilis Study (1932–1972)
OHRP
Office for Human Research Protections
research in ED in 1980s
some relied on deferred consent (get consent later); this was banned in 1993
some turned to argument that there was minimal differential risk of doing research
1996, final rule about waiver for emergency research
When can a waiver be obtained for Emergency Research
Per the Final Rule, it can be obtained if:
- subjects are in life-threatening condition and can’/t consent
- no proven available treatments
- collection of evidence is necessary to determine safety and efficacy of intervention
- there’s no way to ID ppl ahead of time
- rsrch will directly benefit ppl and can’t be done another way
- Then, community consultation, independent data monitoring, and public diclosure are required
Morrison on emergency research
standard of care in ED is old, based on animal studies, and vulnerable populations suffer disproportionate injuries and therefore are most disadvantaged by lack of progress.
-although there’s a waiver, only 21 such studies done in last 10 years bc the hurdles are so cumbersome to get the waiver.
**We should take utilitarian not deontological approach to this research, and we need to do more of it. it’s a disservice to the disadvantaged if we don’t.
Gostin on human subjects research
ethics usually focuses on rights and duties; should focus on moral boundaries for populations.
We need principles to protect poulations:
-don’t harm
-respect self-determination of populations
-protect vulnerable pops, justify rsrch on them by their needs
-protect privacy, integrity, and self-esteem of pops
-distribute bens equally to pops and build infrastructure
population vs behavioral and social sci rsrch
population - all rsrch and practice performed on/affecting groups of ppl/pops (epidemiological research, surveillance, field trials for drugs)
Jim: quality and assessement service studies don’t count as BSS. trend is to include qualitative studies as research.
NIH OBSSR:the term “behavioral” refers to overt actions; to underlying psychological processes such as cognition, emotion, temperament, and motivation; and to biobehavioral interactions. The term “social” encompasses sociocultural, socioeconomic, and sociodemographic status; to biosocial interactions; and to the various levels of social context from small groups to complex cultural systems and societal influences.
Basic research in the behavioral and social sciences is designed to further our understanding of fundamental mechanisms and patterns of behavioral and social functioning relevant to the Nation’s health and well-being
According to Jim, what are ethical issues in BSS research?
- researching that doesn’t address basic human needs
- kinds, magnitudes, and probabilities of harm
- confidentiality concerns
- consent vs assent, can minors consent?
- it can reinforce stereotypes!
- *BSS complain about IRBs, but we actually allow a lot more BSS rsrch than other countries so they should chill the hell out. Plus, because we are careful, people will trust us and participate in rsrch.
finder’s fees and bounty-hunting fees
FF - small payments to non-researcher (physician) who identifies/enrolls subjects
BHF - large, per capita payments from industry to PIs for testing their products
- Christensen and Orlowski argue that FF should never be allowed, and BHF monitored very closely. Block payments are better than per capita.
- Lind compares FF to fee-splitting (phys pay other phys for referral, this is banned)
Bayh-Dole Act
1980 - univeristies, researchers, and research institutions can patent NIH funded work.
-passed to promote research through commercial incentives. since the law, hundredfold increase in patents, copied by other countries.
Emanuel and Steiner on institutional CoIs
instit CoI = conflict btw mission (care, teaching, rsrch) and profit
COI could cause distortion in institutional $ distribution, underemphasis in enrollment of studies that don’t have > $ gain, biased promotions
there should be prima facie norm against CoIs, could be overturned in rare circumstances without other options.
Emanuel and Steiner on institutional CoIs
instit CoI = conflict btw mission (care, teaching, rsrch) and profit
there should be prima facie norm against CoIs, could be overturned in rare circumstances without other options.
Nancy Oliveri
whistleblower who noticed an industry sponsored trial of deferiprone was toxic and inefficacious; reported research to drug maker (Apotex) who said she signed confidentiality agreement. She disclosed anyways.
shows tension btw academic freedom and control of rsrch by pharma.
