Full transcript
0:06So uh good afternoon everyone. Uh first
0:09of all I'd like to thank the research
0:11office for inviting me to give this
0:12training. Um I'm
0:16as told you I work as an investigation
0:19manager at the center of excellence for
0:22bio banking in biomedical research at
0:24the university. Um and uh you heard
0:28about my research and today
0:33uh we we're going to look at how
0:36researchers can navigate their ethical
0:39and legal responsibilities and horizon
0:42Europe uh focusing on topics such as
0:44ethics review uh process and GDPR
0:48compliance and informed consent. So the
0:51session will last about 1 hour and uh
0:54I'll try to include two three practical
0:57scenarios at the end uh to help
0:59illustrate how these principles apply in
1:01real projects. Uh you'll also have time
1:04at the end to ask questions. I'm happy
1:06to receive questions from you uh to my
1:09email and I will also share the slides.
1:13Um so this is a short time in this short
1:17time together we'll focus on the
1:19essentials uh the key ideas and steps
1:22that uh every researcher should
1:24understand while giving you a clear
1:27picture of how to approach more complex
1:30issues later on. Uh the goal isn't to uh
1:34make you an expert in regulations but to
1:36show how uh how legal and ethical
1:38frameworks actually support uh good
1:40science.
1:42Uh so let's start
1:45um
1:47uh and explore how law and ethics come
1:50together to make research both compliant
1:52and meaningful.
1:54So before we examine how ethics are
1:57addressed uh within horizon uh and what
2:01they mean how they are evaluated and
2:03what researchers must do to ensure
2:05compliance and integrity uh throughout
2:08their projects. you will first consider
2:10very briefly uh from where researchers
2:13responsibilities arise by explaining uh
2:16the background of ethical and legal
2:19frameworks. So uh research ethics didn't
2:22emerge overnight. Uh it evolved through
2:25history and shaped by uh responses to
2:28harm and injustice. Uh each milestone uh
2:31we see here at this slide uh from
2:34neurenburgg code of yellow convention
2:37and GDPR. Now the AI act have
2:40strengthened uh and protected uh values
2:44such as dignity, integrity and
2:46accountability in science. So today
2:49these uh principles guide all research
2:52across Europe.
2:55So the major uh principles
2:58are of course um uh autonomy, dignity uh
3:03informed consent and we always act with
3:07beneficience
3:09and justice trustworthiness
3:11and avoiding harm. So the ali convention
3:15reinforces it values as legal
3:17obligations uh not just moral ones as it
3:21is binding.
3:24So um human rights and research ethics
3:28are deeply connected
3:31uh by protecting individuals uh their
3:34privacy, their health and freedom of
3:37thought. So researchers must uphold
3:39these rights in every step uh of data
3:42collection, consent and engagement. Uh
3:46and these ethical practices that both a
3:49legal and a human uh responsibility. The
3:52most important rights uh are that are
3:55relevant to what we're going to talk
3:57today. Of course, the right to privacy
3:59and data protection, the right to life
4:02and health, the right to freedom of
4:04expression, and importantly, the right
4:07to benefit from scientific progress, and
4:10to participate in cultural intellectual
4:12life.
4:16So uh the most relevant human right uh
4:20in contemporary uh research science is
4:23of course uh uh the right to private and
4:27family life under article 8 of the
4:30European convention on human rights. Uh
4:32this means our work with personal or
4:35health data must always balance public
4:38benefit with individual protection. So
4:40privacy is not a technical addon. It is
4:44a fundamental rights.
4:48Uh and of course for setting the
4:50background um uh one right we cannot uh
4:56continue without talking is
4:59uh the right to scientific progress.
5:02This is an important right but it's a
5:04bit forgotten. Uh so the European uh so
5:09the universal declaration of human
5:11rights introduces the right uh in 1948.
5:15Uh and the right uh uh says that uh you
5:20have the right to benefit from
5:22scientific progress and to be recognized
5:25for one's contribution. This is a dual
5:28right and ensures both public access to
5:31knowledge and fair credit to
5:33researchers. Um
5:36so this right has uh now uh been um
5:42applied in article 15 of the
5:44international convention on economic
5:46social and cultural rights uh and states
5:50uh and and obliges the states to make
5:54science available accessible and of
5:57quality. Uh it appears that scientific
6:00freedom and responsibility go hand in
6:02hand. uh nonetheless researchers are
6:06stewards of of this freedom and must use
6:09it ethically.
6:13So what is the right to scientific
6:15progress? Uh the right scientific
6:17process is that everyone has the right
6:18to receive the benefits of a scientific
6:21advancement and in the same time
6:24researchers have the right to uh proper
6:27attribution and the recognition of their
6:29intellectual and innovative ideas. So
6:32the right scientific process about uh
6:35access quality and inclusion uh and
6:38remind us that ethics isn't the buyer is
6:41what enables trust and participation.
6:46So responsible research ensures everyone
6:48benefits from this innovation.
6:52So uh we always have this challenge um
6:56how to balance the necessity of either
7:00utilizing healthcare or having human
7:04subjects um in our research with uh
7:09safeguarding trust and the rights of uh
7:11human participants. Uh and we have to do
7:14that even when there is uncertainty
7:17within uh laws. uh for example uh in
7:22using sensitive data to advance health.
7:25Um
7:27so informed consent and data protection
7:30and transparency
7:33are not optional. They are the the
7:35bridge between progress and production.
7:38And
7:43of course um
7:46uh all research disciplines uh uh and
7:50research uh no matter what the field
7:54whether it is health engineering uh
7:56social science or uh culture the ethical
7:59foundation remains uh the same remains
8:02constant and we adapt uh these
8:05principles to context but never
8:08compromise. on uh dignity, justice,
8:11integrity and so and so forth. Uh this
8:15shared framework grounds the ethics
8:18review process we will discuss next for
8:20Horizon uh 2020.
8:25So
8:28uh where are horizon Europe? Um so uh
8:33let's start with the legal foundation
8:36which is you can see that in article 19
8:39uh of the horizon Europe regulation
8:42uh it establishes that all funded
8:44actions must comply with ethical
8:47principles and with relevant EU national
8:50international
8:52uh law including EU and EU law of course
8:56uh including the EU charter of uh
8:59fundamental rights and the European and
9:01conventions on human rights. And from
9:04this first sentence alone, uh it's clear
9:07that ethics are not um an afterthought.
9:10They are deeply interwined uh with human
9:13rights. Uh the regulation explicitly
9:17highlights principles such as
9:18proportionality
9:20uh respect for privacy and data
9:22protection uh the physical and mental
9:24integrity of persons non-discrimination
9:27and protection of environment and human
9:30health. Uh these princ uh principles
9:33form the ethical foundation of every
9:36horizon uh Europe project. Uh so when
9:39you design uh a study proposal you are
9:42expected to demonstrate how your
9:45research upholds these values.
9:50Uh so now uh that we've seen where
9:53ethics are introducing horizon uh Europe
9:56regulation uh let's move to the uh model
10:00brand agreement which is the second key
10:03legal reference uh the document that
10:06formally defines the relationship
10:07between
10:08uh your project uh consulting in the
10:11European Commission uh within the GRA
10:14agreement article 14 specifically
10:16addresses ethics and research integrity.
10:19It enforces the same principles as the
10:22regulation but provides uh more
10:26operational detail about how they apply
10:28in practice. uh it states that all
10:31beneficiaries must carry out their
10:33activities in full compliance with uh
10:35ethical principles including uh the
10:38highest standards of research integrity
10:40and in line with uh all applicable EU
10:44national and international laws
10:46including the uh charter of fundamental
10:50human rights. Uh another important
10:52requirement is that all activities
10:54funded under Horizon Europe must have an
10:58exclusive focus on civil applications.
11:01Uh this mean that research should not
11:04have a military or non civil purpose. Um
11:08even if it uses dual technologies like
11:11AI, cyber security tools or drones. Uh
11:15these are areas where ethical oversight
11:17is especially critical. Uh and finally,
11:21article 14 uh reminds beneficiaries that
11:25participation in her in Europe also
11:28implies a commitment to fundamental
11:30values of the European Union which are
11:33of course respect for human dignity,
11:36freedom, democracy, equality, the rule
11:38of law and human rights, including the
11:41rights of minorities.
11:43So in other words, ethics in Horizon
11:46Zero go far beyond compliance. They
11:48express to use a broader vision of
11:51responsible uh human centered research.
11:55Uh whether your work is in medicine, AI,
11:58social science or engineering, these
12:00principles apply equally. Ethics is
12:04everywhere in research and every project
12:06is expected to reflect these values in
12:09its design and implementation.
12:13So um what cannot be funded very
12:16quickly. Uh
12:19so um Ryzen will not find any under
12:23circumstances
12:25uh whatever is defined in article 18 of
12:27the regulation
12:30um and that uh include human uh cloning
12:34uh for reproductive purposes, genetic
12:37modification of human beings that could
12:40become heritable uh creation of human
12:43embryos solely for researcher or for
12:46stem cell procurement
12:49uh somatic cell nuclear transfer
12:51techniques or any activity that leads to
12:54the destruction of human embryos. Uh
12:57research involving human uh stem cells
13:00both adult and embryotic may be funded
13:02only if it complies with national
13:05legislation and UF standards. uh this
13:09means that uh a project involving such
13:13research can proceed only in the
13:16countries where it is legally permitted.
13:19So always check uh three levels of
13:21compliance. The national EU and
13:24international uh laws um bad compliance
13:28with EU and international human rights
13:31law remains mandatory.
13:41So there um there are nine ethics issues
13:44in horizon Europe. Um
13:49uh so uh the European Commission has
13:52identified these nine main categories of
13:55ethics issues that may arise in
13:57research. uh the first is human
14:00embryotic stem cell and human embryos
14:03followed by human participants and then
14:06human cells and tissue. Next we have
14:08personal data which cover all research
14:11involving the collection uh storage or
14:14processing of personal or sensitive
14:16information. Then of course uh there's
14:18animal animals uh non-U countries uh
14:22environmental
14:24uh environmental health and safety and
14:28uh new horizon uh is of course
14:31artificial intelligence.
14:33AI was not considered as a separate
14:35ethics category by June 2020
14:39uh initially but it has now been
14:41recognized as a distinct and complex
14:43area especially because um of the
14:46potential of bias discrimination and
14:50challenges to accountability and
14:52transparency.
14:54Uh so finally there is a ninth more
14:56flexible cateorium. Uh and you can see
14:59that in the ethics
15:01um called other ethics issues. Uh this
15:05is where you would plug anything that
15:07does not keep uh an input into the
15:11predefined boxes. for example, social or
15:14behavioral uh research
15:18that could be uh misused or that raises
15:22new ethical questions to MIT
15:25technologies. Uh and then um cutting
15:28across all these areas there are two
15:31important underlying considerations that
15:33are always addressed. One is the risk of
15:36crosscutting uh or overlapping issues
15:40where several ethics dimensions interact
15:43and the other is the potential of misuse
15:46of results. Uh in other words, even if
15:49your project is fully compliant and well
15:52intention, it's important to think about
15:54how your results might be um supplied in
15:58a way that causes harm or violates
16:01rights. So designing categories form the
16:04backbone of ethics process and
16:07recognizing where your project fits
16:10among them is the first step forwards a
16:13strong and credible proposal.
16:20So um
16:24another cost adding concern is in use
16:27when research outcomes could be applied
16:30for harmful or unethical purposes. This
16:33includes knowledge material or knowledge
16:36material or technologies that might be
16:39repurposed for criminal or terrorist
16:43activities or contribute to chemical
16:46biological ideological or nuclear weapon
16:49development. So misuse can also be uh
16:54sub in sub in forms. For instance,
16:58surveillance technologies that undermine
17:00civil liberties uh or uh datadriven
17:04profiling tools uh that risk uh
17:08discrimination or stigmatization of
17:10vulnerable uh groups. Uh so in horizon
17:14Europe these fall under ethics issues
17:17table but may also overlap with uh
17:21security review if they pose potential
17:24safety threats. So researchers are
17:27expected to anticipate and mitigate
17:30these uses uh through carefully design
17:33and transpiring governance and ethical
17:36inflection.
17:44So why ethics matter?
17:47And now that we've seen um what ethics
17:51in horizon Europe looks like in
17:53practice, let's take a moment to talk
17:56about why it actually matters. What's
17:59the motive behind all this? Uh so ethics
18:02in research uh is something as an
18:06administrative task uh something uh you
18:09complete at the end of a proposal just
18:11before submission. But in reality,
18:14ethics and integrity are not above
18:17exercise. They are core components of
18:20excellent research and innovation. Uh
18:23the European Commission sees ethics as a
18:25foundation of trust pass and science
18:28innovation and the people behind it.
18:32And uh this is why uh the commission
18:36promotes what we call ethics by design.
18:39Means thinking about ethical aspects
18:41from the very beginning and integrating
18:44them uh into your research planning and
18:47design and not adding them later as an
18:51afterthought.
19:01So uh what is the practical process uh
19:05the ethics issue table and self
19:07assessment. Um so now uh we understand
19:11the ethical and legal foundations of
19:14Horizon Europe. Uh which is the why. Uh
19:17let's turn to the practical process. How
19:19ethics are identified, evaluated and
19:23monitored within a proposal. Uh the
19:27first and most important step is
19:29completing the ethics self assessment
19:31which is found in part A. uh section
19:34four that need forms in the EU funding
19:38uh independence portal. Uh this process
19:41has two simple steps but both are quite
19:44crucial. Uh the first step uh you answer
19:48a series of questions in the ethics
19:50issue table. Uh this question help
19:53identify whether your research involves
19:56any activity with ethical implications
19:59such as human participation.
20:02Um
20:06and the second step is whether you have
20:10uh if you have answered yes to any
20:12question you must provide additional
20:15information explaining how you will
20:17address uh that issue and how you appro
20:21your approach ensures compliance with
20:23all relevant laws and ethical standards.
20:26Uh for instance, if the project involves
20:28human participants, you must describe uh
20:31your informed consent procedures. Uh
20:33your ethics approval process and your
20:36data protection strategy. If your
20:39project includes AI systems, uh you must
20:42explain how you prevent bias, ensure
20:45explanability and maintain human
20:47oversight.
20:49Um this part your written explanation
20:52forms the ethics self assessment itself.
20:56It becomes part of the proposal and will
20:59uh uh will be examined by ethics expert
21:02during the evaluation.
21:09So here you can see an example of how
21:12the ethics issue table looks in
21:14practice. Uh this one focuses on the
21:17sections about the environment, health
21:19and safety which is one of the main
21:21ethics categories. Uh for example, you
21:24may be asked whether your research uses
21:27uh substances or technologies that could
21:29harm humans, animals or the environment
21:33or if it involves endangered species or
21:36protected areas. If you answer yes,
21:40uh to any of this question, you need to
21:42include a bit more detail in your
21:44proposal. Uh things uh things like a
21:47risk benefit analysis and how you apply
21:51the precautionary principle and what
21:54safety measures you plan to implement.
21:57You should also have supporting
21:59documents ready such as the laboratory
22:01safety classification or any necessary
22:04authorization or permits. Uh the key
22:07idea is that these materials shouldn't
22:09be prepared at the last minute. Um
22:18so some um concrete examples of what the
22:22table includes
22:24um regarding artificial intelligence. Uh
22:28there might be questions and does the
22:30activity involve AI based systems? Could
22:33AI potentially stigmatize or
22:35discriminate against people? Does it
22:38replace or influence human decisions? Uh
22:41similarly for personal data, you will be
22:44asked whether your research involves
22:47processing new or existing uh personal
22:49data or transferring data outside of
22:52you. Uh so if you answer yes to these
22:55questions are there going to be required
22:58a detailed explanation how um you're
23:02going to minimize any risks. Um and uh
23:12so uh let's move to uh the ethics
23:16screening. Uh so after the ethics of
23:19assessment uh that you have completed at
23:23proposal stage the next step is the
23:25ethics screening which is carried out by
23:28the European Commission. Uh this uh
23:31screening only applies to shortlisted
23:33proposals of course uh those on the main
23:38list or the reserve list. uh it's uh
23:42conducted by a British independent
23:44ethics experts and it's completely
23:46separate from the scientific evaluation.
23:49uh so while the scientific review looks
23:52excellent and impact this screen focuses
23:56only on ethics.
23:59Uh so the screen takes place before uh
24:02the grant agreement is signed. uh if the
24:06reviewers identify serious or complex
24:08ethics issues the proposal moves to full
24:11ethics assessment uh which is deeper and
24:14more uh detailed process. Uh so in such
24:18case the signature of the grant
24:20agreement may be slightly delayed until
24:22all ethics requirements are uh clearly
24:26defined. Uh for proposal that don't
24:29raise any major ethics concerns, uh the
24:32process is much simpler and they are
24:36cleared with no additional ethics
24:38requirements.
24:45So um
24:48as mentioned when a project is found to
24:51raise serious or complex uh ethics
24:54issues it must go through a full ethics
24:57assessment but when does a proposal
25:00qualify as serious? uh it's considered
25:04serious when the research has the
25:06potential to violate fundamental rights
25:09or freedoms uh as protected by the EU
25:12charter of fundamental rights or the
25:15European convention of human rights.
25:17It's also serious if the project could
25:20undermine key EU values u the values
25:23that we already talked about human
25:26dignity freedom and democracy and the
25:29rule of law. Uh another case is when uh
25:33uh the research could cause significant
25:35harm to participants
25:37uh researchers, the public or to animals
25:39or the environment. And finally, if
25:42there are concerns about integrity of
25:44the data or research practices meaning
25:48uh doubts about honesty, reliability or
25:52transparency, the proposal will also be
25:54flagged as serious. In such cases, the
25:57commission may request full ethics
25:59assessment before the project can
26:01proceed.
26:07So some proposal are considered complex
26:10from an ethics perspective.
26:13Uh this happens when a research uses
26:16very advanced or untested methods or
26:20technologies that might create
26:22uncertainty about uh participant safety
26:25or the impact on uh fundamental rights
26:28or research integrity. A proposal can
26:31also be complex if it raises significant
26:33ethics issues at scale. For example,
26:36when it involves large number of
26:38participants, uh controversial methods
26:42or high-risk technologies or when the
26:45work is carried out is sensitive or
26:47higher is in sensitive or higher
26:50regions.
26:52Uh so finally projects can be complex
26:55when there is multiple or overlapping
26:58ethics issues. uh for instance research
27:02with marginalized or vulnerable groups
27:05uh where risk like stigmatization or
27:09exclusion could combine and increase the
27:12ethical impact.
27:15So in short complexity arises when a
27:18project scope design or context makes
27:21its uh ethical risk more significant or
27:24harder to manage and that's when uh a
27:27full ethics assessment is usually
27:28required.
27:31So there is also uh a third situation
27:34where a proposal can be considered both
27:37serious and complex from an ethics point
27:40of view. Uh this happens when the
27:42research area itself is the subject of
27:45major ethical debate. Uh for example
27:49when methods or technologies involved
27:51are still being uh widely discussed
27:54among scientists and ethicists.
27:57uh it also applies when uh there are
28:00serious doubts about researchers or
28:03institutions ability to properly manage
28:06or mitigate the risk linked to the
28:08project. Uh another case is uh when the
28:12research uh poses a high risk or of use
28:17and no clear or effective measures exist
28:19to uh contain that risk. Uh finally if
28:23the proposal shows a lack of awareness
28:25of the ethical issues it will
28:28automatically be flagged as serious or
28:30complex. So the most important point is
28:33uh the last point um where if ethics are
28:38not properly addressed uh from the start
28:40the project will undergo uh full ethics
28:44assessment and later face closer
28:47monitoring by the consumer.
28:52very briefly to see what are the
28:55possible outcomes of ethics aiming. Um
28:59so if during uh the assessment you
29:03answer no to all questions meaning your
29:05project raises notical concerns you
29:07receive a straightforward ethics
29:09clearance. Uh if you do answer yes to
29:13some questions uh the concern and ethics
29:16expert will check whether these issues
29:18are serious or complex. Uh if uh not and
29:24your approach complies with national
29:26European international legislation, you
29:29can receive ethics clearance and no
29:30further requirements
29:32are asked. Um however if the project is
29:36uh ethically sound but still needs extra
29:39oversight uh the commission may ask you
29:42to appoint an independent ethics advisor
29:46or ethics work. Uh in that case you see
29:50the conditional ethics clearance. Uh so
29:53finally if your proposal does a series
29:55of complex ethic issues it will go
29:58through a full ethics assessment.
30:01So uh as you can see the more robust you
30:04address ethics early on the sooner uh
30:08this process uh uh we completed
30:22so um
30:26uh once a party goes to full ethics
30:29assessment ment. Uh there are several
30:32possible outcomes. Uh if you do receive
30:35ethic clearance, the grant agreement can
30:37be finalized without delay. Uh if the
30:41proposal receive conditional ethics
30:43clearance, it means that uh certain
30:45ethics requirements must still be met.
30:48Uh so these requirements uh might be
30:52detailed in the ethics summary report
30:54and must be uh fulfilled before the
30:57grant agreement is signed or included as
31:00contractual obligations in the grant
31:02agreement. So often these will involve
31:06creating a dedicated ethics work package
31:09uh with specific ethics deliverables uh
31:13for example providing uh consent
31:15procedures, copies of ethics approvals
31:18and other documentation as part of
31:21reporting and monitoring.
31:34So uh this process doesn't stop once you
31:38get fun. Um ethics continue during the
31:43whole lifetime of the project. Um an
31:46ethics review can be launched if your
31:48project raises complex or sensitive
31:51issues. uh in that case the commission
31:54will check whether um you actually doing
31:58what you promise in terms of consent uh
32:01data protection safety animals welfare
32:05AIS and so on. So this review is not
32:09optional. It's uh uh initiated by the
32:12project officer if something is unclear
32:15or if documents you provided are not
32:18convincing. uh it's then uh handled by
32:23up to five independent ethics experts.
32:25So the message is ethics is not a box to
32:29pick at proposal stage. Uh for highest
32:33project compliance is monitor all the
32:36way through implementation.
32:43So this slide is a clear overview of how
32:48the ethics process unfolds in Horizon
32:50Europe. It's a snapshot. It starts with
32:53the ethics self assessment as we said uh
32:56completed by the applicant during
32:58application phase. uh if the proposal
33:01passes the scientific evaluation and is
33:05uh placed on the main or reserve list it
33:08under goes ethics by consu ethics
33:12experts to evaluate the level and
33:14seriousness of potential issues.
33:17uh projects flagged as relevancies or
33:20complex ethics issues are then subject
33:23to full ethics assessment uh where
33:26additional details may be requested on
33:30ethics work package uh or advisory board
33:33may be needed to be added. Uh finally
33:36for projects with ongoing ethical uh
33:40sensitivities the commission may uh
33:42contact ethics review checks or audits
33:44during the implementation of project.
33:53So um
33:56to probably address ethics uh in your
33:59proposal and project the first step is
34:03uh for ethics self assessment
34:07and ideally discuss with the consult
34:11early uh the project design
34:15and not left until last minute
34:20a bit more faster because
34:23don't have much time. So some takeaways
34:28um is uh the key message is to consider
34:32ethics early from the very beginning of
34:36uh your project ideation
34:39and not just right before submission.
34:42Ethics should be embedded uh in your
34:44research design from the start. Always
34:47refer to relevant national and
34:49international laws and standard as what
34:52is permissible in one country may not be
34:54in another country. And make sure you're
34:57aware of these differences when during
34:59your activities.
35:05So you can find more information about
35:08all this uh in the grant agreement in
35:12horizon program guide and uh so forth.
35:16You will receive these uh slides and add
35:21the links. So there's also a guidance on
35:24the views.
35:30So I now turn to how to minimize uh the
35:35ethics complexity of the project. Um
35:41there is a need for cross project
35:44compliance strategies and good
35:46practices.
35:48Uh so in general documentation and
35:51compliance clear documentation is the
35:54backbone of ethical compliance. So every
35:57project must maintain a privacy policy
36:00and data protection impact assessment uh
36:03to show accountability and transparency.
36:07uh also uh before any data or Pakistan
36:10involved in ethics review IRB or
36:13national expertise uh must be obtained
36:18and of course um uh if it's needed there
36:23must be oversight for this uh during uh
36:27the lifetime of the project and uh
36:31training of uh LC issues and GDPR issues
36:36of the researchers and people
36:41that are involved in research.
36:44Um so very briefly
36:47uh when working with um uh patients,
36:51children or anyone in vulnerable
36:53positions, you need to you need to uh
36:57keep consent clear, layered and boring.
37:01uh make withdrawal uh simple and
37:04transparent and always plan how to
37:06handle incidential findings. One of the
37:09most uh uh often finding of the ethics
37:13review is that they don't have a clear
37:17uh policy about institutional findings
37:20uh for minors or cognitive impair uh
37:23participants
37:25uh involve guardians and uh use um age
37:32appropriate material. Of course uh if
37:35procedures are invasive uh you need to
37:37justify the need and strengthen safety
37:41monitoring.
37:45So uh when human samples and bio banks
37:49are involved you need to double check
37:52that existing consent really allows for
37:55new uses of international or
37:58international sharing. If not uh you
38:02need to consent or request an ethics
38:05waiver. Uh so maintain a sample uh
38:10provenence record showing consent scope
38:13and approvals and for uh important human
38:17tissue make sure all EU permits are in
38:20order.
38:23Um so also treat data like living
38:28organisms. So you need to protect it,
38:31track it and and know where it travels.
38:34Uh you need to run a data protection
38:37path assessment when you're dealing with
38:39uh especially with sensitive data or
38:41special data hardware called under the
38:44GDPR.
38:46uh if you can you should apply
38:48minimumization techniques and crypton by
38:51default and uh audit who has access uh
38:56if partners are outside you need to use
39:00uh GDP safeguards and sign
39:04the standard contractual process
39:07and so forth. So regarding
39:11artificial intelligence, uh you need to
39:14keep it explainable. Uh decisions must
39:17always have a human in the loop. Uh test
39:21for bias, update models regularly and
39:25document their purpose uh inputs and
39:28leads in a clear uh fact sheet.
39:33So of course for animals
39:36um the use of animals
39:40uh should be done only when no
39:42alternative uh uh research type exist.
39:47Uh design studies use ars in mind reduce
39:52time and publish all results even
39:56neutral ones provide a recognition.
39:59uh and document every lab based uh
40:03safety protocol
40:05uh and the GOP uh compliance step. This
40:09shows reviews that the project considers
40:12human and environmental safety as
40:14important efforts.
40:17So as we said for new collaboration
40:22uh you need to track all cross water
40:25flows of samples of data. If you
40:29collaborate outside the EU
40:31um apply for the safeguards and update
40:34consent language needs to be included in
40:38the consent that transfer uh to non
40:42countries will happen.
40:45uh of course for governance and
40:47oversight
40:48uh set on independent ethics board early
40:52if it's needed uh include the uh uh
40:56legal GDPR AI and patient experts uh
41:00they need to meet regularly and actions
41:04and keep ethics visible
41:06uh through a dedicated work package with
41:10clear deliverables
41:12um and forg or high work uh you need to
41:18add a data set keyboard and use this
41:22part.
41:27So uh regarding user insecurity, we need
41:31to think ahead about your uh use risks
41:35uh and profiling tools, AI news uses,
41:38sensitive data and so forth. Uh and
41:41create a short meditation plan uh as to
41:45who controls access uh how publications
41:48are reviewed and how issues are
41:51escalated.
41:54So finally um you need to involve
41:58patients and public integrate gender
42:01analysis where relevant and community
42:04open science within GDPR limits uh and
42:08RI isn't for it's how we make research
42:14meaningful.
42:20So in some instances you might need to
42:23um have um
42:28uh access committees for example in bio
42:31grants. Uh it was uh uh compulsory to
42:35create an access committee where
42:38uh when researchers ask for uh samples
42:43or data, you have to do the proposal and
42:46ask for certain um um certain documents.
42:52For example, um Rex or BIC
43:03And
43:05what else
43:07we going to
43:10have much time? You always sign MTS and
43:14between EU countries
43:17when uh data are transferred.
43:23So now we don't really have much time to
43:25talk about the GDPR in consent. I'm
43:28going to go very
43:30uh very quickly as we all know that GDPR
43:34is the main legal instrument that
43:36regulates uh research about when it
43:41involves data either processing storage
43:44or sharing uh personal data. It applies
43:48to all you uh residents worldwide.
43:52uh there are certain uh key principles
43:54affecting research that are establishing
43:57these little instrument data
43:59minimization content transparency and
44:03confidentiality.
44:04Uh it also states all the types of data
44:08and uh depending on what data you use uh
44:14specific measures need to uh be followed
44:18um especially if you use uh sensitive
44:21data or special data as they're um
44:24stated in the
44:26uh in the regulation.
44:29Uh some examples of what special
44:31categories of data are racial oriented
44:34origin, sexual orientation, of course,
44:36health data, genetic data and so forth.
44:40Uh it also includes the data subject
44:42rights that you have to include in the
44:45informed consent forms. The more the
44:47more the most important uh right of the
44:50data subject is of course the right to
44:52withdraw
44:56and
44:59Yeah.
45:01Uh so the most important thing is that
45:05consent forms should uh specify
45:09uh when and how can someone withdraw
45:13from research
45:15and it should be as easy as it was to uh
45:20uh be part of the research.
45:24So there are different types of consent
45:28and each research has to take in
45:30consideration the specifics of the
45:34project and uh which consent is
45:38considered for the project. Uh I'm not
45:41going to go to uh all these uh types but
45:46you can have a look at them later.
45:50So uh I'm going to
45:54very very quickly say that
45:57um a valid consent consists of these uh
46:02five uh facts which is that you have to
46:05make sure that uh the participant has
46:08capacity. So by default we say that
46:11participant has capacity but if it comes
46:14to the attention of the researcher uh
46:17that the participant is not
46:19understanding
46:20uh the purpose of the research or why uh
46:23is participating uh it has to go through
46:27specific steps to make sure that it has
46:31capacity.
46:32Uh, of course there must be
46:34voluntariness.
46:36Uh, and this sometimes uh it is very
46:39difficult to uh understand where it was
46:43voluntary when uh relationships between
46:47participants and uh doctors exist as Dr.
46:53also the researcher.
46:55Um, of course there needs to be a
46:59provision of sufficient information
47:02and uh this also depends on the research
47:05project and sufficient information means
47:08that the participant will be able to uh
47:12make a clear decision whether wants to
47:15participate in the uh project and of
47:18course uh all information should be
47:21delivered in lay language not for the
47:24participant to understand comprehend and
47:26this should be documented to a consent
47:28form and sign. So you always give
47:32sufficient time to the participant to um
47:36reflect and understand the project
47:40before signing the consent form and also
47:43uh very importantly you have to give
47:45time for questions.
47:48So I'm going to skip all this
47:52and
47:54I don't know how much time do we have
47:56from this scenario.
48:02>> So uh summary of the key points um
48:07uh
48:09so ethical and legal frameworks are
48:11dynamic consent should be understood as
48:15a process.
48:16uh the most suitable model of consent
48:19should be adopted uh and uh must be as
48:23easy to withdraw consent as it is to uh
48:26consent and data protection is critical
48:30to ensuring trust and acceptability.
48:33Uh so governance and oversight ensure
48:36accountability. So I have some scenario
48:40that we can have a look at it was
48:43requested.
48:45Um so the first one it's about
48:52uh a consortium
48:54that validates a CFS blood free
48:59biomarket panel to a black early
49:07vascular amino in a box of 55 year
49:13reduces historical bio bank samples
49:16uh that are used
49:20um and uh unexplainable
49:24AI model uh combines proteomics and MRI
49:29and one of the data analysis partners
49:31outside the EU.
49:34So what what questions what ethical
49:38questions might be raised from this? Uh
49:42the first one of course is does the
49:44original consent cover the secondary
49:46use?
49:48Uh does it cover international sharing
49:52and and does it cover uh a potential
49:55contact of the participant? So this is
49:59uh something that would be evaluated
50:01based on uh the consent forms.
50:05Um
50:07and uh of course how do we inspect the
50:11right not know while managing uh
50:14clinically relevant essential findings.
50:17Uh what uh triggers at DPIA and what are
50:23are your musts in it? Uh is data
50:27transfer on your partner lawful local
50:32and appropriately safeguarded?
50:34um what human in the loop business is
50:37necessary
50:39risk reports.
50:41So
50:43uh in these cases
50:47uh you should use uh layered plain
50:50language consent
50:53add ongoing consent checkpoints and
50:55offer opt out for individuals
50:58uh adopt a written incidential finding
51:01policy. This is very important if you
51:04are planning to um
51:07to give results.
51:10Uh complete the DPI BPIA
51:15um data impact assessment um before
51:20large scale processing. Um
51:24and uh for non transfers uh you need to
51:28implement GDPR act 46 safeguards which
51:32is to sign the standard contractual
51:35clauses. Sometime contractual causes is
51:38a contract an agreement which is the
51:41same as uh the MTA is resigned between
51:44the EU countries but um the clauses
51:49there are standard so that the member
51:52that is outside the EU guarantees that
51:55you follow uh all the rules under the
51:58GDPR.
52:03Um
52:06so move to next one. There are more that
52:08you could do here.
52:12So this is uh a nonrandomized
52:16case to trial in children with the labs.
52:20Um
52:21and all test genomics guide regime. Um
52:26there's a smart there is a smart uh
52:29phone app that collects uh pros and
52:32activity data between visits uh and
52:36samples between us sites um and results
52:40inform those adjustments.
52:43And here we need to ask how do we
52:47structure as consent so minors
52:50understand
52:52and and guidelines through the uh
52:56consent.
52:58So uh depending on the age uh there are
53:02ethics guidelines how to create an
53:05ascent process and it's different for
53:08very small children and
53:12under 12 and it's different for 12 year
53:15olds and it's different for seven 16 17
53:19year olds and it must happen in par with
53:23the consent that is given by the
53:26guidelines.
53:28Um
53:30so uh what uh the other thing is because
53:34they are minus uh what the question is
53:37what safety oversight is proportionate
53:41uh in this instance
53:44um
53:45and uh does the create systematic
53:48monitoring prop
53:51under the GDPR? How do we handle
53:53withdrawal especially if data already
53:56uploaded or analyzed
53:59uh which equity checks ensures models
54:03don't disadvantage subgroups? So we said
54:06the first thing is that we need a double
54:08consent model
54:10uh which is age of copy consent plus the
54:14graphing consent. Um
54:18and uh also for the app they should
54:24before deployment
54:27uh and
54:29local and crypton and clear notices and
54:33so forth. And of course define uh the
54:37withdrawal practice.
54:39What happens to sedonimize or uh
54:42irreversibly anonymized data if they
54:46have already been used in research and
54:50of course plant bias check uh sex age
54:55um and restrict AI to decision support
55:00with clinician overides.
55:08This is another scenario.
55:16The next one um
55:19so this is the transational consortium
55:22studies liver disease disease. I'm
55:25finding a humans IPSC derived liver to
55:31immunication mice. The goal is to
55:33understand hydrosis and test and type
55:38compounds before human trials. No human
55:40participants are recruited but human
55:43derived cells are used. Uh so questions
55:47arise about ethical status of human
55:49animals
55:51proportionality of animal suffering and
55:54governance of human cells uh provenence
55:57and consent for secondary use. So what
56:01are the ethical means of creating human
56:03animals
56:05uh for biomeical purposes? It's a
56:08question. Do the original donor's
56:10consent for this age line adequately
56:15cover animal implantation and commercial
56:18research. So this must be clearly stated
56:22in the consent form. How can uh we
56:25ensure compliance with the PRs uh about
56:28the protection of animals uh and be
56:31transparent um uh about the welfare of
56:34animals and monitor? Uh should
56:38experiments be uh pre-register and who
56:41provides ethical oversight when human
56:44cells are involved?
56:46Uh so the first thing is to confirm that
56:49IPSC lines come from owners with valid
56:52consent allowing secondary use including
56:54standard plant plantation, genetic
56:57modification and commercial development.
57:01um seek approval from both animal ethics
57:05committees and institutional review
57:07boards.
57:08Uh given the human animal interface uh
57:12comply with the PIs regarding uh the
57:16animals safety
57:19um and use in organizers before in in
57:25before in vivo.
57:27Uh so you might need to pre-register the
57:30studies for transparency reasons
57:34and um of course document
57:38um the origin and concerns
57:44and so
57:52I have
57:55Um
57:59so this is actually a low risk ethics
58:04research. Um a university uses existing
58:08anonymized plasma samples and associate
58:11clinical data from uh certified national
58:14environment to validate protein
58:17biomarker for cardiovascular disease. Uh
58:20all samples were originally collected
58:22under approval studies uh consented for
58:26secondary use and then now university
58:28anonymized uh before transfer. No new
58:31data collections, no uh invention, no
58:35crossber transfer. Is this considered a
58:38low ethics service research? Um what uh
58:43minimal documentation should still be
58:46provided to uh the fun or
58:50What can data be exacted as clear to the
58:52anonymized?
58:54Uh do the researchers still need ethics
58:57committee notification or just the way
59:01there? Uh so the first thing is to
59:04consider whether they are truly
59:05anonymous. If there is no key or code um
59:10that can
59:12u identify
59:15the participants blend to the anonymous
59:19um
59:21and of course there's need again risk to
59:25keep uh
59:28documentation also keep the varants
59:31appro
59:32uh some for access and data use license.
59:38Um
59:39so this is much more simple than
59:43um other
59:46cases.
59:48Uh so
59:50another scenario about engineering and
59:53AI artistic drone inspection for bridge
59:56safety. Uh this is a project where the
1:00:00engineering consulting builds upon
1:00:02drones to inspect aging bridges using
1:00:06high resolution cameras and AI image
1:00:10analysis detect plat. Some flights would
1:00:14be over public roads and near
1:00:16residential areas. So the questions are
1:00:19what privacy or data protection issues
1:00:21arise when drones capture signature
1:00:24images of people or vehicles? Um is
1:00:28specific airspace authorization needed?
1:00:31How can the project minimize
1:00:33surveillance concerns?
1:00:35So um familyization
1:00:38measures could be the use of uh
1:00:41geoencing and should learn to prevent
1:00:44capture of identifiable individuals.
1:00:48uh contact data protection impact
1:00:50assessment before field work and obtain
1:00:54of course aviation for the clearance and
1:00:57inform the public uh with visitors
1:01:01and on
1:01:03the project website as well.
1:01:08So how scenarios
1:01:12um
1:01:16Another
1:01:20scenario that
1:01:22was a different discipline um regarding
1:01:27this researchers record spoken stories
1:01:31in an endangered dialect among elderly
1:01:35uh speakers in rural areas. Uh so the
1:01:38data will be stored in an open access
1:01:41purpose. So the key questions are what
1:01:44considered valid from consent voice
1:01:46recording.
1:01:49uh now we have different
1:01:52um ways of collecting data from voices
1:01:56from smells and so forth and this
1:01:59creates issues as to what uh what um
1:02:03value from consent is uh for business
1:02:06and voice
1:02:08uh and how can participants retain
1:02:11control of their personal narratives. Uh
1:02:14when does open access conflict with
1:02:17privacy or c
1:02:18ownership. Uh so there needs to be used
1:02:22obtained language concern explaining the
1:02:26purpose and the access level and
1:02:29withdrawal rights of participants
1:02:33prior choice uh choices for example to
1:02:38be used only for for public uses for
1:02:41research only or anonymous um an
1:02:44anonymously by experts and of course
1:02:48apply GDP for Identify drawings data and
1:02:52sudonize made data.
1:02:56Um so uh provide participants with
1:02:59access to the final archive for
1:03:02verification or flow and ensure cultural
1:03:05sensitivity when publishing
1:03:08uh floor or publishing.
1:03:13So one more I think let's see
1:03:26any questions
1:03:30thank you very much very useful