Getting Health Reform Right From Diagnosis to Health Sector Reform

1
Getting Health Reform RightFrom Diagnosis to
Health Sector Reform

• Shahram Yazdani

2

• Once health sector reformers have decided which
  performance problems to focus onthat is, once
  they have decided on their strategic
  prioritiesthey have to determine causes of the
  poor performance they are concerned with.
• This is the process we call diagnosis. Then
  they have to decide what to do about the
  situationthe process we call policy
  development.

3

• With respect to diagnosis, the fundamental
  strategy is, Work backwards.
• Keep asking why until you have discovered the
  causes of the poor performance you want to
  improve.
• The goal of this process, called a diagnostic
  journey, is to construct a diagnostic tree
  that links aspects of poor performance to those
  causal factors that might be changed or modified
  by policy interventions.

4
Simple guidelines

• Process matters
• Imitate but adapt
• Use evidence
• Use control knob framework

5

• Identifying potentially effective changes in one
  or more control knobs is the end of the
  diagnostic task and the beginning of the policy
  development task.
• Analysts first work backward to causes. Next,
  they have to look to policies to change those
  causes, and then work forward from causes to
  forecasts of improved health sector performance
  as a result of various policy changes.

6
The Genesis of Major Health Sector Reform Efforts

• Process of health system diagnosis and policy
  development as if it focused on one specific
  performance problem at a time is likely to
  provoke piecemeal, programmatic initiatives
• But as we have noted before, countries often
  confront a number of inter-linked performance
  problems.
• The diagnosis for each specific problem may well
  reveal that certain common features of the health
  care system are important causal factors.
• To deal effectively with these multiple problems,
  multiple interventions may be required. This
  kind of analysis, then, is one path that can lead
  a country to major health sector reform efforts.

7
The Genesis of Major Health Sector Reform Efforts

• There is also a matter of coherence here.
  Multiple programmatic efforts run the risk of
  being uncoordinated and even inconsistent.
• Someone may urge decentralization of primary care
  to the provincial level, even as someone else
  urges carrying out a centralized national
  vaccination program.
• Dealing with such conflicts in an integrated way
  may require a coordinated look at and a plan for
  the revision of the system as a whole.

8
The Genesis of Major Health Sector Reform Efforts

• There is also the issue of the limited political
  and technical resources typically available to
  reformers.
• Tackling a large number of performance problems
  in isolation from each other can place far too
  great a demand on those resources.
• Focusing instead on certain common and more
  fundamental problems can, potentially, be a
  cost-effective use of limited reformist
  energiesassuming, of course, that such changes
  are possible.

9
The Genesis of Major Health Sector Reform Efforts

• An integrated and extensive reform effort may
  have political value, in that it often appears
  more dramatic and attention-getting than a series
  of narrow programmatic initiatives.
• By promising more to the population, broad reform
  can seem like an effective political
  strategyespecially when the population is
  dissatisfied with the current system

10
Developing a Health System Diagnostic Tree

• There are likely to be several stages or series
  of causes
• Each effect may well be produced by more than one
  cause
• Each cause is likely to have multiple effects
• Causes and effects may interact and reinforce
  each other in various ways
• Not every cause will be manipulable by public
  policy
• Change may well require acting on more than one
  cause at a timei.e., changing more than one
  control knob.

11
Developing a Health System Diagnostic Tree

• Health system diagnostic tree is similar to, but
  different from decision tree, which is used for
  analyzing decisions under uncertain
  circumstances.
• At each branching point in a decision tree, one
  of a set of mutually exclusive events occurs.
• In a health system diagnostic tree, however, an
  analyst represents various possible causes for a
  particular situation at each branching point, and
  it may well be the case that several of the
  causes are in fact operating at the same time.
• Unlike the decision tree, there is no presumption
  that paths are mutually exclusive.

12
Developing a Health System Diagnostic Tree

• As we have done this analysis, quite a number of
  the intermediary performance characteristics have
  made an appearance various aspects of
  efficiency, access and quality have played a role
  in all the potentially causal chains.

13
(No Transcript)
14
Linking Diagnosis to the Control Knobs

• Financing If we do have a quantity of service
  problem, can we tap new or additional sources of
  funding to expand the budget for this service?
  Alternatively, can we alter how we spend existing
  funds to increase allocative efficiency?
• Payment If we discover we have a problem of
  motivation, whether at the level of the
  organization or the individual provider, can we
  change how we pay, or the rates we pay, to
  improve both the quantity and quality of
  services?
• Organization If we are not organized to
  facilitate economies of scale and coordination,
  if patients experience cultural barriers or have
  preferences for certain types of providers, what
  changes could we make in who offers various
  maternal health services to overcome these
  problems? If the issue is technical efficiency
  and/or clinical or service quality, can we change
  how providers are organized at the microlevel in
  ways that would increase their performance?
• Regulation Are there regulatory devices we can
  deploy to increase the quality and quantity of
  services or to increase their utilization?
• Behavior If there are important causes on the
  demand side, could we seek to influence mothers
  to increase their use of the services that are
  available?

15
The Process of Policy Development participation
of stakeholders

• First, participation by those affected allows
  policymakers to hear and take into account
  interest groups concerns
• Second, participation can increase the acceptance
  of a plan, because it increases the psychological
  and philosophical legitimacy of the process that
  developed it
• Third, participation can serve to educate
  interest groups about the concerns of, and
  pressures exerted on reformers by, other
  constituencies.
• Fourth, participation educates the participants
  about the detailed content of the proposals that
  ultimately emerge

16
The Process of Policy Development generating
policy alternatives

• People in general have a difficult time thinking
  outside the box of their familiar concepts and
  paradigms.
• Imitate but adapt.
• Imitation can take various forms.
• Ideas from health sector initiatives in other
  countries
• Ideas developed in other policy areas that can be
  adapted to the health context.
• General theoretical arguments can also provide a
  source of new ideas.

17
Screening Tests for Policy Interventions

• Implementability
• Political Feasibility
• Controllability

18
Finding and Using Evidence for Diagnosis and
Therapy

• Evidence-based health policy requiresevidence!
• Yet many lower and middle-income countries lack
  adequate evidence.
• Even when data and information are available,
  they are often not well utilized.
• Most desirable kind of evidences are
  well-designed studies that produce reliable
  statistical results linking policy to
  performance. However, such studies are rare

19
Finding and Using Evidence for Diagnosis and
Therapy

• Advices for Health sector reformers
• Know the literature
• Get advice
• Rapid assessment
• Support good research

20
Summary Advices

• Work backwards up the causal chain until you can
  identify potentially manipulable variables
• Do not jump to quick conclusions
• Be scientific, not judgmental
• Use numbers