Study of the Psalms Chapter 1 verse 1 by wanderean
MCDA
1. 6TH FUTURE TRENDS LATAM 2016
Multicriteria decision analysis:
Opportunities for changing the paradigm on healthcare
decision making to tackle ethical dilemmas
12 October 2016 / Panama City, Panama
Mireille Goetghebeur MEng PhD
Global Scientist, LASER Analytica, Montreal, Québec, Canada Adjunct Professor,
School of Public Health, University of Montreal, Quebec Canada / President,
EVIDEM Collaboration Research Associate, Research Center, Ste Justine
Hospital University Center, Montreal,Quebec Canada
2. • Multiple technologies, procedures & programs call for fair
decisions for prioritization.
• Fair decisions are value-laden and based on a complex
reasoning.
• There is a diversity of perspectives of what constitutes a fair
decision.
• Perception of decisions as fair is crucial and depends on
communication of the reasoning that has been performed
through a fair process.
• Pragmatic approaches to support reasoning underlying fair
decisions are lacking.
The need for a new paradigm in
healthcare decision making
3. Voting Question:
Do you think MCDA can help address legitimacy of healthcare
decisions?
1. Strongly disagree
2. Disagree
3. Neither agree nor disagree
4. Agree
5. Strongly agree
4. Open-ended question:
Could you elaborate why do you think MCDA can help address
legitimacy of healthcare decisions or not?
5. MCDA to support decision making?
• MCDA is a reductionist approach to a decision problem.
• We cannot reduce human reasoning to a series of steps driven by an
algorithm.
• We cannot make complex calculation that drives us even more away than
current approaches from our individual interpretive frames.
6. Voting Question:
Do you think MCDA should be used as an algorithm or as a
support to the decision making process (DMP)?
1. Absolutely as an algorithm
2. Mostly as an algorithm
3. Neither as an algorithm nor as a support to DMP
4. Mostly as a support to DMP
5. Absolutely as a support to DMP
7.
8. • Motivation
• Actual decision
• Evidence
• Values
• Substantive values (criteria)
• Procedural values (process)
• Communication
• Implementation
• Revision
Accountability for reasonableness (A4R)
Relevance
Publicity
Leadership
Revision
Daniels and Sabin. Philos Pub Health 1997; 26:305.; Baltusen et al. Cost Eff Resour Alloc 2006; 4:14; Goetghebeur et al. BMC Health Serv Res 2008; 8:270 ;
Battista IJTAHC 2010Clark and Weale. J Health Org Manag 2012; 26:293;
Anatomy of the natural decision process
9. General motivation is to achieve health which can be further defined into:
• Ethical imperatives (normative aspects) to:
• Prevent/alleviate suffering in individual
patients with meaningful healthcare.
• Prioritize those who are worst off while
providing the greatest benefit for the
greatest number
• Ensure long term sustainability.
• Wisdom to make decisions adapted to the context
(feasibility aspect).
Includes aspects of deontology, distributive justice, utilitarianism & virtue ethics (practical wisdom)
Anatomy of the natural decision in health
MOTIVATION
10. Natural decision relies on all types of
evidence
• Scientific
• Colloquial
• Imputed by logic
Anatomy of the natural decision in health
EVIDENCE
11. Substantive values can be made explicit by:
• Criteria derived from the motivation and its
underlying aspects.
• By identifying trade-offs between generic criteria
(weights), which represent the individual value
system.
Anatomy of the natural decision in health
12. Procedural values can be operationalized by
• A decision committee with members representing
the diverse perspectives of stakeholders in a given
society; representativeness ensures legitimacy of
the decision.
• A deliberation process with democratic participation
by which each member can share their individual
interpretive frames to reach an equilibrium.
Anatomy of the natural decision in health
13. Communication
• Increase acceptability of decisions by
communicating the reasoning (rooted in the
common goal) on which they are based.
Implementation
• Success of implementation is dependent on
acceptability.
Revision
• Revision is performed based on an
assessment of whether the initial motivation of
the decision is fulfilled in real life, on the long
run (new evidence, different reasoning).
Anatomy of the natural decision in health
14. • A bottom up approach based on the natural decision
process (not a specific methodology) calls for the
integration of several domains of research such as
ethics (substantive & procedural), human rights,
evidence-based medicine, health economics, health
technology assessment, decision analytics and
communication.
• MCDA adapted to healthcare might provide a mean to
develop such an approach.
MCDA Framework that support the natural
decision process?
15.
16. 1. Goal
2. Criteria
3. Weights
4. Evidence
5. Scores
6. Visualization & uncertainty
7. Ranking and deliberation
Methodological choices for a MCDA framework to support evaluation and
prioritization at system level need to be made with the natural decision in mind
Adapted from Thokala et al. 2016
MCDA Steps
17. The framework needs to be rooted in the common goal and its
underlying ethical imperatives, so the motivation can be
acceptable by all, so:
• Derive criteria from a major ethical position (e.g.,
utilitarianism)? This will not address the point that
several aspects of the major ethical positions are
considered in a natural decision process.
• Derive criteria from aspects of major ethical positions
derived from the common goal?
• This will be helpful if each criteria are justified by at
least one ethical foundation, allowing to tackle, by
design, the ethical dilemmas
From Reflection
18. For an application at the system level, the framework
need to rank interventions based on their value towards
the common goal, so:
• Value measurement based on everything but
cost? This will not address the sustainability
imperative, which is inherent to the common
goal.
• Value measurement integrating all the criteria?
• This will be helpful if it criteria are defined to fulfil
the principles of non-redundancy, independence,
operationalizability and completeness of MCDA.
From Reflection
ETC…
19. The framework cannot be reductionist but it has to be
pragmatic, so:
• Few criteria in framework to keep it simple? This
will not address the point that many criteria will
anyway be considered.
• Generic criteria to structure high level value
system which can be further defined with sub-
criteria? This will be helpful if it does not
constrain reasoning and support individual
interpretive frame.
From Reflection
20. Many jurisdictions use cost-effectiveness but this is a
composite measure combing data from several concepts
(efficacy, safety, cost of intervention, other types of cost, etc.)
which does not allow for interpretation of these concepts, so:
• Keep it in the framework? But does not address the
point that effectiveness and other concepts will be
anyway considered separately, which will create
distortion in reasoning & double counting.
• Make it optional? Will be helpful to transition from
current paradigm but recommend to remove it to
support the interpretation of distinct concepts to
clarify reasoning.
From Reflection
21. The framework cannot be reductionist but it has to be
pragmatic, so:
• Scoring based on numbers? This will not address
the point that “interpretation of data that takes
place during appraisals requires judgement” Sir
Rawlins, NICE 2013.
• Interpretive scoring scales? This will be helpful as
it can capture the interpretation of numbers, but
we need agreement upon what constitutes low
and high of such scale and thus criteria may be
considered quantitatively; if not, no scoring but
rather qualitative consideration
From Reflection
22. • Provide a basis for fair decision rooted in the common goal and
operationalization of all its ethical aspects (pervasive ethics).
• 2006: Initial framework
• 2009: Collaboration:
• Share the framework: independent not-for-profit organization, free
membership,
• Enrich it collaboratively: open source philosophy (volunteering, feedback
from users,)
• Protect it: int’l board of directors, cannot be commercialized, rules &
regulations,
• Sustainable: volunteering, current funding based on user fees as
applicable, public funds
To Action
www.evidem.org
23. 23 www.evidem.org
Initial framework the natural decision
process; Pr M Goetghebeur PhD, U Of
Montreal, LSER; Dr M Wagner PhD
LASER, Canada.
MCDA Pr. R Baltusen PhD, Radboud U,
Netherlands.
HTA Pr. R Battista MD PhD, Quebec Research
Funding Agency, Canada.
Health economics: Pr. P Kind, U of York, UK.
Policy decision makers Dr M Tringali MD
PhD, Health Directorate, Milan, Italy; Pr. J
Miot PhD, U of Witwatersrand, South Africa;
Dr H Castro MD PhD, Ministry of Health,
Colombia
Clinical decision makers, Pr. C Deal MD PhD,
Univ. Hospital Center, Montreal; Pr. J Dolan, MD
U of Rochester, NY, USA
Global health: A Velasquez, MSc WHO, Switzerland
Ethics: Pr. N Daniels PhD, Harvard U, Boston, MA, USA
A platform for an international reflection
24. • Success:
• 300 members across the decision continuum, >40 countries,
• Translated in >10 languages
• V3.2 in Dec 2016 (10th release)
• Implemented in Europe, Americas and Asia
• Challenges
• Misunderstanding of its intention / design
• Misuse
To Action
NB: users create their team and seek resources/funding/expertise
to adapt & apply framework,
www.evidem.org
25. 1-GENERIC GOAL : HEALTH
Tool EVIDEM v3.1
Conceptual approach and
operationalization
Goal is further defined in 4 substantive aspects rooted in decision
ethics:
• Patient: imperative to prevent/alleviate suffering (aspect of
deontology).
• Population: prioritize those who are worst off (aspect of
distributive justice) and greatest good to greatest number
(aspect of utilitarianism).
• Sustainability : ensure sustainable healthcare system (aspect of
utilitarianism).
• Context awareness: practical wisdom (aspect of virtue ethics).
Four aspects are further defined in 20 generic criteria abiding with
MCDA principles
Evidence and Values Impact On Decision Making
Output : Generic criteria operationalizing the motivation of healthcare
decisions and its underlying ethical imperatives
26. Interactive exercise
As a policy decisionmaker, which criteria would you keep into your framework to support
deliberation and to rank interventions according to their holisitic value in your context?
27. 2- CRITERIA 3-WEIGTHS 4-EVIDENCE 5-SCORES & INSIGTHS
Quantitative Minimize mental
distance
Scientific and colloquial Interpretive scales Narratives
Disease severity Direct rating scale
Point allocation
etc
Turner syndrome: Female
specific generic disorder
characterized by reduced
life expectancy(details)
3 Very severe
2
1
0 Not severe
Several of my
patients have
experienced etc…
Etc
Qualitative NA Scientific and colloquial Non-numerical
impact
System capacity Risk of inappropriate use
of growth hormone for
Turner syndrome due to
….(details)
negative
neutral
positive
In my hospital,
specific
constraints due
to etc…
Etc
Tool EVIDEM v3.1
Evidence synthesis
& quality
Tool EVIDEM v3.1
Weighing methods
Tool EVIDEM v3.1
Assessment package
Evidence and Values Impact On Decision Making
Output : Pragmatic multicriteria evidence matrix to support
reasoning & deliberation
28. INSIGTHS
Several of my patients
have experienced etc…
In my hospital, specific
constraints due to etc…
6 - Visualisation of Reasoning
QUANTITATIVE CRITERIA - VALUE OF INTERVENTION A
Criteria contribution to value & insights
Tool EVIDEM :
Calculator,
visualisation &
presentation
Output: Face validity of reasoning & uncertainty at group level.
29. INSIGTHS
Several of my patients
have experienced etc…
In my hospital, specific
constraints due to etc…
QUALITATIVE CRITERIA - IMPACT ON VALUE
Impact of criteria & insights
Tool EVIDEM :
Calculator,
visualisation &
presentation
6 - Visualisation of Reasoning
Output: Face validity of reasoning & uncertainty at group level.
30. Output: Final deliberation & decision based on a group reasoning on value and
opportunity cost (financial impact)
Output: Management of opportunity costs guided by identification of interventions with
“best value towards the goal”
7 - Ranking & Deliberation
31. Lombardie, Italy: Reimbursement decisions
• Adaptation of EVIDEM combined with
EUNetHTA core model.
• All EVIDEM modules used.
WHO: List of priority devices
• Adaptation of EVIDEM to devices.
• Qualitative MCDA.
• Tools transferable to member states.
Examples of applications of EVIDEM
32. Lead: Michele Tringali - Radaelli et al IJTAHC 2014;30(1); Tringali. HTAi Oslo 2015; http://vts-hta.asl.pavia.it
List of health technologies appraised for reimbursement in 2013-2014
Lombardy (Italy), Health Directorate
• In place since 2012.
• Web based system to
support deliberation.
• Transparent & efficient
process.
• Enhanced
communication and
acceptability of
decisions by
stakeholders.
33. Ranges of scores: High: XXX, Medium: XX, Low: X, Very
Low: 0, Not applicable: NA, Unknown: ?
• Working tools: collect data from experts & justify decision to include/exclude devices
A: Function/Intervention External beam radiation therapy
B: Contentious Option YES
C: Specific Medical Devices
Linear accelerator LNAC (at least 3D
conformal therapy,6MeV)
D: Contentious Option YES
E:ValueCriteriaandscores(seedefinitionsand
guidanceforscoringinAppendix)
Completeasapplicable
Effectiveness XXX
Safety XX
Patient perspective XXX
Therapeutic benefit* XXX
Multi-disease applicability X
Multi-cancer applicability XXX
Ease of use XX
Ease of training XX
Remote communities NA
Affordability –device XX
Affordability –
maintenance & replacement
X
Healthcare resources consequences XX
Quality of evidence XXX
Notes
Specialized Human resources needed,
5% of cancer care of the overall budget
(Radiotherapy)
A: Unit Radiotherapy
B: Basic Service or subunit External Radiotherapy
C: Category Therapeutic
D: Basic Function or Intervention External beam Radiation therapy
E: Expected Outcomes Eliminate malignancies by radiation.
F: Type of Cancer Breast, Cervical, Colorectal, Prostate, Leukemia, Lung.
G: Contentious Option (function) YES
H: Specific Basic Medical devices
Linear accelerator LNAC (at least 3D conformal
therapy,6MeV)
Notes NA
I: Contentious Option (device) YES
J:Key&contextualconsiderations(see
definitions&guidanceforbelow)
Completeasapplicable
Interdependencies
Simulation and planning process, Mould make process,
Computerized treatment planning systems
HR requirements
1. Physicist
2. Radiation oncologist
3. Radiation technologist
Infrastructure requirements
Adequate Network Infrastructure and storage capacity,
space for adequate furniture for moulds storage, Closed-
circuit television system (and voice and audio), Oxygen
Supply
Key Associated devices
Immobilization and patient positioning systems,
Computerized treatment planning systems
QA & management
i.e. Dosimeter, Phantom for daily mechanical and light
field checks on teletherapy unit, Radiation survey meter.
Refer to Radiotherapy General Medical Devices (Quality
assurance equipment)
Other NA
K: General devices General Radiotherapy Devices
L: Level of Health Care DH, RH
Lead: Adriana Velazquez World Health Organization. WHO list of priority medical devices for cancer management. 2016.
WHO List of Priority DevicesWorkingtools:exampleexternalbeamtherapy
34. Canadian Minister of Health Jane
Philpott, seen at Parliament on Sept. 29,
2016
“We need to find ways to put health care
on the road to long-term sustainability”
35. Researchers &
Developers
Regulators
HTA
HC systems Payers
Hospitals
Clinicians
Patients
& carers
• Integrate the diversity of perspectives to tackle ethical
dilemmas across the decision continuum.
• Reach an equilibrium on what defines interventions with
“best value towards the goal”.
A Road Map to Achieve the Goal
of Healthcare Collectively
36. What are the major opportunities and challenges for MCDA in
Latin America?
Opportunities:
Challenges:
Open-ended question: