(Paracelsus)… pattern in the data… respiratory illness when they were in mines… correlation
John Snow… dot on a map… data collection without consent… did he stop the epidemic?
Removed the pump… gun show epidemiology
Epidemiology in the era of big data. Epidemiology 2015
More human lives will be touched by technology than anything else…
Privacy vs science
Privacy and confidentiality were never seriously considered to be hard barriers to sharing and analysis
Reductio issue (left toe example)… can you help others with same problem?
Biomedical research has long relied on the work of trusted entities to collect health information
Security, de-identification, anonymization, pseudonymisation
Cloud is a metaphor
Data is used sloppily—knowledge, information etc
Ethical concerns should focus on decision support- given variable data and database quality, uncertainty
Appropriate user and users
Data sharing and interoperability
5mb computer in 1956 pic.
Is not an absolute right
Must therefore be balanced again other right, including a right to benefit from science
Infrastructure support refusers
Health and privacy
Smart laws and policies
Recognition of duties to collectives
Learning healthcare systems
Public health analogue “duty to treat”
Role of Ethics
Illuminate the force, scope and limitation of rights
Identify and balance conflicting rights, and rights and duties
Identify and justify duties
Management and governance
Balance health, data, privacy
Identify best practices
Develop revised IR-like review entities
Consultation capacity for risk communications, decisions under uncertainty
Agreement and civil society
DNA, epigenome, life going through illness, and exposure to all thinigs in world…
Social determinants of disease…
Scandanavia—clinical care/ research? Health insurance provision?
Health care provided for everyone? They affect the health of populations. Buy insurance?
Understanding cases within profession (Wade Robinson)
Edward Tufte’s compelling mistaken reading of what went wrong with the Challenger.
Challenger ‘O-rings’ safety in the Challenger: the astronauts were killed by impact not by explosion
Tufte blamed the engineers- if they had done a scatterplot they could have worked it out
You can see increasing risk of damage
The ascending curve of risk
If proper scatterplot done, no one would have risked to launch the Challenger in such cold weather
“They didn’t know what they were doing, but they were doing a lot of it”
Big thing is that NASA launched BELOW 50 C at 40C.
Research Misconduct in FDA Regulated Clinical Trials: What Not to Do
What is research misconduct?
The falsification of data or results or recording and reporting
Deviation from established protocol- data doesn’t reflect what you were studying
Violation of human subjects and rights
How did definition happen?
World Med Assoc Declaration of Helsinki
Tuskugee Syphilis study and Belmont Report
US Government Oversight
DHHS (office of human research protection)
How did FDA get authority?
Thalidomide incident in 1950s (Francis Kelsey)
1962 Amendments (Kefauver-Harris Amendment)
What does FDA do to ensure research misconduct does not occur?
What is FDA looking for in regards to ethical standards and regulatory requirements
Required reports by IRB
FDA Research Observations
FDA investigators cite observed deviations from regulatory requirements
What happens after observations are reported?
Observations by FDA investigators are passed through multiple levels of review before a final classification
FDA Clinical Investigator Action
Warning letter and nidpoe
Get in trouble—this is big issue
University of Pennsylvania Gene Therapy Trial
Objective: treatment for OTC deficiency
Prevents proper elimination of ammonia
Death in the trial within 96 hours
1999 Nov. 1.5 month Clinical Investigator James Wilson
Based on all the citations that happen
Did not follow protocol
Should have STOPPED clinical trial… 5 increases of therapy before subject died in cohort 6!
They did not tell the patient ALL the information.
University College London Medical Device Implantations
Strict regulatory oversight—safe for humans
Results: Deaths of guinea pig subjects
Termination of researchers employment
Regulatory and criminal investigations
Bial-Portela Clinical Trial
The investigators didn’t do anything wrong, this was just a poorly designed trial
Cohorts were overlapping
Real time data wasn’t coming in and subject died, and others ended up with brain lesions
Written to prevent issues that have previously occurred
Human subject rights protection
Assurance of data validity
Inspections and Audits (FDA, OHRP)
Dr Sheldon Krimsky
Monsanto Litigation Documents
Integrity of Research journals
Crisis in Credibility
Contested issue between science and commerce
Can it affect public health
Corporations have a different view of science- one of many inputs to production
Science just one of inputs into production function …. Lead, asbestos, BPA, tobacco (lobby against)
IARC, WHO Report 2000
DSM-IV: panel members had conflicted interest when they produced categories
100% panel on mood disorders, schizophrenia etc.
“On the Take”
Conflict of interest and scientific journals
Can you believe what you read
Funding effect in science
The Monsanto Litigation
Specialised cancer research arm of the WHO, reached a determination chemical glyphosate-based herbiside
More than 270 of the cases have been consolidated multi-district litigation for oversight by 1 judge in the US District court of San Francisco
Multi district litigation
Monsanto’s Web site on Glyphosate—does not have adverse effects to humans, wildlife and environment (hodgkins lymphoma IHL)
Rightoknow web site
Ghost writing: that is how they do business. EPA referenced. Determination 2016, glyphosate was not likely carcinogenic
They hired Intertek Scientific & Regulatory Consultancy
Undo influence of regulatory agencies
Undo scientific journals…
Unethical to ghost write journals!
It was published by Toxicology articles
William Heydens… disclosed this
Editor in chief got money for this from Monsanto!
JFCT- retract or remove the journal… paper retracted after 1 year
“results were not definitive”… paper retracted but as soon as it was another journal published the article
Who should a journal editor review—an employee of Monsanto review a paper about Monsanto
When vital public health reports are published in refereed journals there is a heightened expectation that they meet professional standards of scientific integrity. Those standards include full disclosure of conflicts of interests…. Sources of funding….
The Lancet--- 3 positive reviews, and one so-so… “not a priority to publish”
Ethical Guidelines for Authorship- Subrata Saha
Gift authorship; ghost writing
Millikan Oil Drop Experiment: https://www.youtube.com/watch?v=XMfYHag7Liw
Fletcher and Millikan share but each sole authorship
The Journal of Bone and Joint Surgery
Henry R Cowell 1998 wrote an editorial
Inform patient care
Goal should not be to add to CV but to help
Ideas should be new
Don’t waste space, for redundant data… should be consolidated, not salami?
Prior to the experiments
Informed consent, clinical aspects, cost, statistical method, begun BEFORE we begin
Manuscript written with same patient care in mind than before
Career advancement, get grants
Rights to copyright
Pressure to publish first
Who was first? Discovery
Tenure and professional standing
Who should be an author?
Advisor and graduated students…
Now many groups collaborate, in group and within nation
All work builds on multiple achievements
Expectation: author wrote it; author did work that was written down; always real data; and results and claims should be accurate; author should disclose bias or potential for COI.
Contributions—interpretation, acquisition of data, drafting, critical, final submission, author should know about it and carefully review it
Financial disclosure form
Should not be allowed to publish
Who is not the author?
Just because someone got funding but played NO role in design or rationality of study then should not be author
General responsibility for lab—not enough for author
Just because they have the data—not enough for author
Other authorship issues
Still a problem: divvy up work, order of authorship, etc Avoid conflict by meeting from outset and publishing paper.
What are external regulations?
Self-plagiarism—is this plagiarism?
Biomedical Device Risks and Non-Medical Implantable Risks by Katina Michael
Slides available here.
Audio of my presentation here.