Accountability for
Reasonableness in Priority
Setting:
Theory and Applications
Norman Daniels, Harvard School of Public Health
Rotterdam, Netherlands, December 9, 2010
Overview
•
•
•
•
The Problem: In designing and modifying benefit packages and
setting other priorities for allocation in health we encounter
conflicting needs, limited resources, and reasonable ethical
disagreements about what priorities to set (conflicting ethical goals)
Existing economic tools aim at maximization, efficiency, generally
ignore fairness and other ethical concerns
Procedural justice--why and what
•
•
Provide legitimacy and possibly fairness in the face of
disagreement
Accountability for Reasonableness--a proposal
Some applications
Bottom Line
•
Fair, deliberative (and participative) process is needed to assure
legitimacy and fairness given reasonable disagreement about
design of benefit package, setting priorities
•
•
•
inclusions/exclusions and other resource allocation decisions
create winners and losers with different claims
absence of consensus on principles means we need procedural
justice
(Just Health, Cambridge 2008; (with Jim Sabin) Setting Limits
Fairly, Oxford 2002,8
Conflicting ethical goals
•
•
•
•
•
Two goals of health policy: improving population health and
distributing health fairly under resource limits
Sometimes goals coincide but often in tension
Competing public and provider demands on benefit increments-aging population, epidemiological transition, new technologies,
access to information
Decisions create winners and vocal losers
Reasonable ethical disagreement
Limits of Economic Tools
•
CER
•
•
•
US: limited questions, political opposition
Germany (IqWIG)--limited use, mainly for price negotiation
CEA
•
•
UK (NICE) and some Commonwealth and European
Distributive insensitivity--(partial response: flexibility in threshold
in NICE)
CEA vs. Fairness (equity)
(why we should go beyond CEA)
CEA
FAIRNESS
BO vs FC
BO
Weighted chances
Priority to worst off
None
Some--varies
Aggregation
Any
Some
Unsolved rationing problems
(and CEA cont’d)
•
•
•
•
extreme positions described by principle (CEA takes one), but not
endorsed by many
in between positions not described by principle, more acceptable to
people
reasonable disagreement about trade-offs
(modified) form of pure procedural justice needed (no prior
agreement on principle)--outcomes of a fair process count as fair
(here only defensibly fair), enhances legitimacy if broad buy-in to
process as fair
Example
•
•
•
•
•
Consider a group of patients with worst prognosis without treatment -the “worst off”
We can give a) no priority to them, b) maximum priority to them, or c)
something in between
If treatment has little effect, many want to give no priority (or get
bottomless pit); if treatment has big effect, many want to give more
priority
CEA favors no priority, contrary to public concerns that we give some
priority
Reasonable disagreement about how much to give
Procedural justice: why
•
If we have no prior agreement on principle regarding what is a fair or
just allocation of resources, we may accept the outcome of a fair
process as fair and legitimate (pure procedural justice)*
•
•
•
presupposes we can better agree on fair process than on
principle
where there is disagreement on values, respect for the
disagreement is important, and showing that respect in the
process is key to establishing legitimacy for the decision
* distinguish types of procedural justice
Types of Procedural Justice
•
Impure (have principle, need process to apply it)
•
•
•
criminal trial
cake at birthday party
Pure (no prior agreement on principle, outcome of fair process is
fair)
•
•
example: gambling--unconstrained, not defeasible
A4R: constrained, defeasible
Accountability for
Reasonableness (A4R)
•
•
•
•
Publicity (transparency including reasons)
Relevant reasons (as judged by appropriate stakeholders)
Revisability (in light of new evidence, arguments, appeals)
Enforceability (assurance that other conditions are met)
A4R: Remarks (1)
•
•
•
Strong publicity requirement compared to market rationale, but
middle ground on explicit vs implicit debate
Relevant reasons: consensus on what reasons are relevant, but
different weights allowed--losers cannot say decision based on
wrong or irrelevant reasons
Due process-- included in revisability
A4R: remarks (2)
•
•
•
Process at various levels with adaptation of it to each level
Conditions necessary (but not clearly sufficient) to assure legitimacy,
fairness
Considerable room to adapt to political culture (one size does not fit
all)--realizable to different degrees (more less transparent, revisable,
etc)
A4R yields:
•
Presumption of similar treatment for similar cases
•
•
•
Commitment to coherent use of reasons
“Similarity” defined by reference to reasons and principles
Rebuttal
•
•
•
•
•
Show relevant difference in cases
Show rationale for revising principle
Public record of commitments - behavior matches pronouncements
Similar to case law
resolving moral disagreement by deliberation, not mere vote
Some Implications
• Implicit vs explicit rationing debate
• Divergent results of fair process
• Consumer voice--stakeholder participation
• Social learning curve
• Contribution to democratic deliberation
Stakeholder participation
•
•
•
•
•
not for enhanced democracy, but for improved deliberation--broader
consideration of arguments, more transparency
can enhance buy-in because more voices included (so voices to be
broadly relevant)
especially important in publicly administered or heavily publicly
regulated private schemes
risks: pure lobbying, domination of debate, power intrudes and
controls; avoid tokenism
selection is key: relevant to level of decision
International Experience with
A4R
•
•
•
•
US-- little uptake (land of death panels)
WHO equity guidelines for 3 by 5, Canada, Norway, Sweden-adopting features
UK--NICE
Mexico--catastrophic insurance, IMMS formulary (proposal to be
described)
Figura 1. Propuesta para la Priorización de Intervenciones Alto Costo
1. Priorización numérica cuantitativa
(1) ECONÓMICO
(1) CLÍNICO
2. Opinión cualitativa sobre
priorización numérica clínica y
económica
(2) ÉTICO
4. Analiza y emite opinión
sobre los resultados de los
grupos clínico, económico,
ético y aceptabilidad social y
presenta al CSG
(3) Comisión para
Definir Tratamientos
y Medicamentos
4. Consulta a representantes de la
sociedad civil
(5) Consejo de
salubridad General
6. Recibe opinión y
recomendaciones sobre
priorización del CSG
(4) ACEPTABILIDAD
SOCIAL
(6) Comisión Nacional de
Protección Social en Salud
5. Emite opinión final sobre la
priorización de intervenciones de
alto costo y envía a CNPSS
Three further apps
•
•
•
Reducing health inequalities, since we encounter unsolved rationing
problems in trying
Supplementing human rights based approach to health
other policy disagreements --e.g., identified vs statistical victims
Further research
•
•
Is it “better”--for what? how to measure?
“better decision-making”-- better in what way?
•
•
•
how should stakeholder participation be managed?
Impact on legitimacy-- should be measurable, but difficult
Impact on fairness-- may itself be contested
Summary
•
•
•
•
•
Benefit design, other priority setting creates winners and losers,
faces reasonable ethical disagreement
To enhance legitimacy and assure fairness in benefit package
design, there is need for fair deliberative process (A4R)
Process adapted to various levels in system, to different political
cultures
Difficult element is stakeholder participation
Varied international uptake in A4R--need better evidence it “works”
Descargar

Norman Daniels - Erasmus Observatory on Health Law