Learning from international experience

1
Learning from international experience?
Evaluating and comparing health and health care
in Europe and beyond
Visby, 3 July 2006 Ellen Nolte London School of
Hygiene Tropical Medicine
2
Context

• Health systems are complex
• Most countries have multiple systems
• US has Medicare, Medicaid, Veterans Affairs,
  Bureau of Indian Services, Armed Forces (and that
  is only the public sector!)
• Wide variation of health care systems across OECD
  countries
• finance, organisation, outcomes
• Each system is influenced heavily by its
  environment (political, cultural, economic)

3
Context

• Common challenges
• Rising costs / need for cost containment
• Demographic changes
• Technological advances
• Increasing public expectations
• Common goals
• Ensuring accessible health care of high quality
  that is responsive, affordable and financially
  sustainable

4
International comparisons of health care systems

• Cross country learning potential
• Experience of countries to provide an
  experimental laboratory for others
• Consider alternative options
• Mutual learning
• Cross-fertilisation
• Transference of models and ideas
• Confirming the positive/negative

5
International comparisons of health care systems

• OECD
• Measuring health care, 1960-1983 expenditure,
  costs and performance (1985)
• OECD Health Data Set (from 1993)
• WHO
• Health System Performance Assessment Framework
  (HSPA 1998)
• World Health Report 2000 (WHR 2000)

6
International comparisons of health care systems

• Commonwealth Fund International Working Group on
  Quality Indicators (1999)
• Nordic Indicator Group Project
• OECD Health Care Quality Indicator Project (2001)
• European Union
• Benchmarking regional health management (BEN)
  (within EU Health Monitoring programme) (12/2001)
  
• DG SANCO Working Party on Health Systems (set up
  11/2003)

7
Challenges

• Definitions vary and contexts differ Are we
  comparing like with like?
• Availability and comparability of data
• Appropriateness of available data are we
  measuring what is important, not just what is
  available?
• Timeliness of comparison
• Comparing health systems or health care systems?

8
One hospital .
Marienhospital Letmathe, Germany

• Features
• 120 beds
• 3000 inpatients/a
• 4600 outpatients/a
• 3 specialist departments
• Staff
• 21 physicians
• 48 nurses

Source Märkische Kliniken GmbH, Qualitätsbericht
2005
9
and another
Karolinska University Hospital, Stockholm

• Features
• gt1,600 beds
• 1.3 mill patient visits/a
• 7 surgeries per hour
• Staff
• 2,400 physicians
• 5,000 nurses
• 1,200 biomedical/lab tech

Source www.karolinska.se (accessed 30 June 06)
10
Measuring hospital capacity
If counting these
11
How can we compare health systems?

• Descriptive studies
• systematic, structured descriptions can provide
  basis for subsequent analysis
• use of structure identifies areas that are
  unclear or poorly thought out
• Quantitative studies
• shifting from studies of process (determinants of
  health care expenditure) to outcome (health
  system performance)
• Focussed analytic studies
• what are the advantages and disadvantages of
  different ways of funding a health care system?

12
Descriptive studies
13
Quantitative studies

• Largely evolved from the health economics
  perspective
• Use of production function approach that
  describes the production of health in terms of a
  function of possible explanatory variables
  (e.g. OECD, Starfield and colleagues)
• Epidemiological approaches the concept of
  avoidable mortality
• Tracer approach

14
Florence NightingaleThe concept of avoidable
mortality
15
Avoidable Mortality (I)

• Rutstein et al. unnecessary, untimely deaths
  (1976)
• Conditions from which, in the presence of
  effective and timely medical care, premature
  death should not occur
• early detection, e.g. cervical cancer
• medical treatment, e.g. hypertension
• surgery, e.g. appendicitis

16
Avoidable Mortality (2)

• Treatable (amenable) mortality
• Deaths from causes sensitive to health care
  (primary hospital care, collective health
  interventions eg screening)
• selected cancers (breast, colorectal, testes,
  cervix), diabetes lt50, hypertension/stroke,
  surgical conditions, maternal mortality,
  perinatal conditions etc.
• Preventable mortality
• Deaths from causes sensitive to public health
  policies
• Lung cancer, liver cirrhosis, transport injuries

17
Avoidable mortality in Sweden (1)
women
men
All causes
Other causes
Avoidable causes
Source Nolte, unpublished
18
Avoidable mortality in Sweden (2)
women
men
Change to ICD 10
Treatable causes
Preventable causes
Of total mortality Treatable mortality 15
(men) to 27 (women) Preventable mortality
10 (men women)
Source Nolte, unpublished
19
Avoidable mortality in selected countries, 2000
women
men
Treatable causes
Liver cirrhosis
Transport injuries
Lung cancer
Source Nolte, unpublished
20
Age- cause-specific contributions to differences
in male life expectancy(0-75) between Sweden
USA, 2000
treatable
Source Nolte, unpublished
21
Age- cause-specific contributions to differences
in female life expectancy(0-75) between Sweden
USA, 2000
treatable
Source Nolte, unpublished
22
Age-standardised death rates(0-74) from treatable
causes, 1990/91 2000/02
men
Source Newey, Nolte et al. 2004
23
Age-standardised death rates(0-74) from treatable
causes, 1990/91 2000/02
women
Source Newey, Nolte et al. 2004
24
Rankings of health systems
Source Nolte McKee 2003
25
How do countries compare?

• Different models of health care provision
• Differences at different levels

• Approach probe disorders or tracer
  conditions that capture certain elements of the
  health care system
• Discrete and identifiable health problem
• Evidence of effective, well-defined health care
  intervention
• Natural history of condition varies with
  utilisation and effectiveness of health care
• Sufficiently common

26
Diabetes as tracer condition

• Deaths (lt45) considered avoidable by timely and
  effective health care
• Optimal management requires
• co-ordinated inputs from range of health
  professionals incl. primary care specialists
• access to essential medicines monitoring
  equipment
• active participation of informed patients
• Can provide important insights into primary and
  specialist care, and into systems for
  communicating among them

27
Diabetes Mortality-incidence ratio
Source Nolte et al. 2006
28
Comparative policy analyses
29
Comparative policy analyses

• What policies that are already in operation work
  in the exporting setting?
• What are the contextual factors that are
  necessary for it to work in that setting?
• Do those factors exist in the setting into which
  the policy is being imported, and in what ways
  does the policy need to be modified?
• Once imported, does the policy work as intended?

30
Does it make a difference?

• Preliminary evaluation of dissemination OBS
  activities
• (Interviews with (i) key informants in selected
  countries and (ii) WHO/OBS directors/staff (2004)
  n 80)
• Added value
• Evidence generation through primary research in
  the region
• Analysis and synthesis of published evidence
• Provision of conceptual frameworks and consistent
  methodologies for comparative research and
  analysis
• Cross-(country) learning from comparative
  research, empirical evidence, and practical
  experience

31
Conclusion

• Increasing interest in international comparisons
• Comparisons as a tool to learn from the many
  experiences of others
• To optimise benefit there is a need
• to overcome the temptation of drawing simplistic
  conclusions from comparisons (nationally and
  internationally)
• for a nuanced approach that is timely and is
  based on a detailed understanding of the nature
  of systems and sub-systems and the settings in
  which they are embedded
• for information intelligence, i.e. understanding
  the underlying data and their limitations