Research challenges: scale, complexity, uncertainty, policy applicability

1
Research challenges scale, complexity,
uncertainty, policy applicability
Short course. Climate change and health. Canberra
2003

• Alistair Woodward
• Tuesday 30th September 2003
• Wellington School of Medicine and Health Sciences
• University of Otago

2
PREDICTED CLIMATE CHANGE UNDER THREE SCENARIOS
(UKMO)
Business as Usual S 750 S 550
4
Temp increase (o C)
3
2
1
1900
2000
2100
2200
Time
3
. outbreak investigation
4
child health study
5
Clinical
Trial with extended followup (eg MRFIT)
6

...
...
... ... Co
hort Study World Record (British Doctors Study)
7

...
...
... ...
...
...
...
... ...
...
...
...
... ... (M
edium term climate scenarios - 2150)
8
Some responses

• Long-run data sets
• Advanced times series methods
• Analogues
• Modelling

9
ENSO and cholera a nonstationary link related
to climate change? Rodo X et al.
PNAS20029912901-6 Data cholera in
Bengal/Bangla Desh and the Southern Oscillation
Index, 1890 - 2000 Methods - times series with
adaptations to detect transient
couplings Findings 1. Increased amplitude of
ENSO since 1975 2. Strong relation between
cholera and SOI in last two decades, not apparent
in earlier series
10
Changes in risk of malaria due to various
environmental changes caused by increased demand
for food and energy
CO2 emissions
Demand for Food
FOOD PRODUCTION
Forest Clearing
Migration of Populations
Urban Crowding
Changes in Mosquito Habitat
Risk of acquiring malaria
severity
Malnutrition
REGIONAL POPULATION GROWTH
Increased Geographic Range of Mosquito and
Malarial Parasite
CLIMATE CHANGE
CO2 emissions
ENERGY USE
11
???
Increase in malaria
Local warming
12
Migration Drug resistance
???
Increase in malaria
Local warming
13
Ineffective vector control
Increase in malaria
Local warming
Mosquitoes
Inadequate health care
14
Some responses

• Measure and control confounders (eg choose
  diseases not sensitive to health care)
• Describe heterogeneity (eg region-specific
  analyses)
• Use complexity-rich methods (eg empirical
  (statistical) models)

15
Dengue

• Dengue fever - the worlds most important viral
  vector-borne disease. No vaccine, no treatment.
• Current global distribution correlated with
  vapour pressure and temperature potential
  distribution under climate change projected to
  2080 (Hales et al, Lancet 2002360830-4)

16
Model of baseline transmission (1961-1990 climate)
17
Model of future transmission (2080s climate)
18
Sources of uncertainty for assessing
CC-attributable burden-of-disease
Emissions
GHG concentrations
Change in climate
Climate/weather-health relationships
Impact model
Socio-economic scenarios modulating effects
19
Climate change and mosquito-borne disease IPCC
tend to increase in range and incidence actual
occurrence strongly influenced by local
conditions
20
Climate change and mosquito-borne disease Reiter
Climate change and mosquito-borne disease. Env
Health Perspect 2001 109 141-161 Climate change
and malaria temperatures without fevers? Science
2000 289 1697-8 From ague to West Nile.
Scientific American 2000 283(6) 10
21
Climate change and mosquito-borne disease Reiter
climate has rarely been the principal
determinant in the past human activities and
their impact on local ecology have generally been
much more significant. It is therefore
inappropriate to use climate-based models to
predict future prevalence
22
Climate change and mosquito-borne disease where
there is not disagreement

• Climate change is happening
• Vectors are temperature-sensitive
• The etiology of these diseases is complex social
  and economic factors are critical
• Climate has not been the major determinant of
  distribution of disease over the last few hundred
  years

23
Climate change and mosquito-borne disease some
major differences

• Disease potential v. disease prediction

It is inappropriate to use climate based models
to predict future prevalence. (Reiter)
The objective is not to propose testable
long-range hypotheses (in a once-only global
experiment). It is to provide indicative
forecasts of a critically important consequence,
to guide pre-emptive policy-making. (McMichael)
24
Climate change and mosquito-borne disease some
major differences

• Assumptions about a future world

if the present warming trend continues, human
strategies to avoid these temperatures are likely
to become more prevalent (Reiter)
25
A1. World Markets
B1. Global Sustainability
Globalisation
High economic growth 2100 population 7
billion High mitigation, low adaptation Temp
(2050s) 0.8o C Rainfall 7 winter, -1 summer
Very high economic growth 2100 population 7
billion Medium mitigation, high adaptation Temp
(2050s) 1.6o C Rainfall 11 winter, -7 summer
Markets, Consumerism
Community, conservation
Moderate economic growth 2100 population 15
billion No mitigation, low adaptation Temp
(2050s) 2.2o C Rainfall 14 winter, -10
summer
Low economic growth 2100 population 10
billion Variable mitigation and adaptation Temp
(2050s) 1.6o C Rainfall 11 winter, -7 summer
A2. Provincial Enterprise
B2. Local Stewardship
Regionalisation
26
Climate change and mosquito-borne disease some
major differences

• Assumptions about a future world

re-establishment of the disease in Italy is
unlikely unless living standards deteriorate
drastically.(Reiter)
malaria could become established again in
Europe under the prolonged pressures of climatic
and other changes if a strong public health
infrastructure is not maintained (IPCC)
27
Different world views?

• IPCC
• A cautious approach that does not take historic
  social advances for granted, and sees humans as
  relatively minor players on a very large
  ecological stage

• Reiter
• Places greater weight on the human capacity to
  shape and control environments, with confidence
  that past achievements will be sustained and
  extended

28
Contributions to uncertainty

• Language
• the level of CO2 in the Earths atmosphere has
  risen by 30 in the last 100 years (IPCC)
• a measurable increase in atmospheric CO2 from
  around 0.029 in 1890 to 0.037 today (Reiter)

29
Climate change and mosquito-borne diseases where
is the uncertainty?

• The most important disagreements are not to do
  with data or methods of analysis
• Conclusions differ because the problem is framed
  in different ways
• The causes of uncertainty lie not only in the
  quality of the science but also with basic
  assumptions and values

30
Policy applicability

• Relevant outcomes
• Local data
• Interventions with short-term as well as
  long-term benefits

31
High risk areas for Dengue fever (Hotspots
analysis which considers vector receptivity,
human population, introduction risk). Vector
Aedes albopictus. IPCC Emission scenario 1. (de
Wet et al, 2001)
32
(No Transcript)
33
1934 4000 vehicles and 2000 pedestrians per
day 2001 34,000 vehicles and 350 pedestrians
34
Transport sector fastest growing contributor to
New Zealand greenhouse emissions
35
Opportunities for multiple benefits - local
pollution due to vehicle emissions
MfE, 2001
36
Research challenges scale, complexity,
uncertainty, policy applicability
Short course. Climate change and health. Canberra
2003

• Chapter 4. Climate change and human health.
  (McMichael et al). World Health Organisation 2003