Methods and Tools for the Human Health Sector Kristie L. Ebi, Ph.D., MPH Washington, DC USA...

Methods and Tools for the Human Health Sector

Kristie L. Ebi, Ph.D., MPH

Washington, DC USA

[email protected]

V&A Assessment Hands-On Training Workshop

April 2005

mailto:[email protected]

Outline

1. Overview of the potential health impacts of climate variability & change

2. Health data to determine the current burden of climate-sensitive diseases

3. Methods and tools for V&A assessment in the health sector

4. Methods for determining a health adaptation baseline

Overview of the Potential Health Impacts of Climate

Variability & Change

Topics• Pathways for weather to affect health• Potential health impacts of climate change

– Extreme weather events• Temperature• Floods

– Vector-borne diseases– Diseases related to air pollution– Diarrheal diseases

Pathways for Weather to Affect Health: Example = Diarrheal Disease

TemperatureHumidityPrecipitation

Distal Causes Proximal Causes Infection Hazards Health Outcome

Living conditions(water supply andsanitation)

Food sources andhygiene practices

Survival/ replicationof pathogens in theenvironment

Contamination ofwater sources

Rate of personto person contact

Consumption ofcontaminated water

Consumption ofcontaminated food

Contact withinfected persons

Incidence of mortality andmorbidityattributableto diarrhea

Vulnerability(e.g. age andnutrition)

Contamination of food sources

WHO

Corvalan et al. 2003

Pathways from Driving Forces to Potential Health Impacts

Drivers of Health Issues• Population density• Urbanization• Public health infrastructure• Economic and technologic development• Environmental conditions• Populations at risk

– Poor– Children– Increasing population of elderly residents– Immunocompromised

Climate change may entailchanges in variance, as well as changes in mean

Temperature Extremes in

the Caribbean,1955-2000

Climate Variability & Change Impacts in the Caribbean

DATE COUNTRY EVENT DEATH ESTIMATED COSTS(US$ million, 1998)

1974 Honduras Hurricane Fifi 7,000 1,331

1982/3 Bolivia, Ecuador, Peru El Niño 0 5,661

1997/98

Bolivia, Colombia, Ecuador, Peru

El Niño 600 7,694

1998 Central America Hurricane Mitch 9,214 6,008

1998 Dominican Republic Hurricane Georges 235 2,193

Cuba Hurricane Georges 6 N/A

1999 Venezuela Landslide 25,000 N/A

Fuente: ECLAC, América Latina y El Caribe: El Impacto de los Desastres Naturales en el Desarrollo, 1972-1999, LC/MEX/L.402; OFDA, Venezuela- Floods, Fact Sheet #10, 1/12/ 2000.

2000 Flood in Mozambique• Heavy rains from Cyclones Connie and Eline in

February 2000 caused large scale flooding of the Limpopo, Incomati, Save, and Umbeluzi rivers– Environmental degradation and poor river system

management and protection contributed to the crisis

• 700 people died, 250,000 people were displaced and 950,000 required humanitarian assistance (of which 190,000 were children under the age of 5)– 14,800 people were rescued by helicopter

Health Impacts of Floods

Philip Wijmans, LWF/ACT Mozambique, March 2000

•Immediate deaths and injuries

•Non specific increases in mortality

•Infectious diseases – leptospirosis, hepatitis, diarrhoeal, respiratory, & vector-borne diseases

•Exposure to toxic substances

•Mental health effects

•Increased demands on health systems

20

30

40

50

60

70P

erc

en

t o

f m

ala

ria c

ases in

ho

sp

ital

-2

-1

0

1

2

3

4

5

Tem

pera

ture

an

om

alies

Jan 97May SepJan 98May SepJan 99May SepTime

Malaria cases Maximum temp Minimum Temp

Proportion of malaria cases andanomalies in maximum temperture: Kenya

Dr. Githeko, personal communication

Climate Change and Malaria Under Different Scenarios (2080)• Increase: East Africa, Central Asia, Russian Federation• Decrease: Central America, Amazon

[within current vector limits]

C hange o f consecutive m onths

> +2

+2

-2

< -2

A1

B2

A2

B1 Van Lieshout et al. 2004

China Haze 10 January 2003

NASA

Effect of Temperature Variation on Diarrheal Incidence in Lima, Peru

Daily Temperature

Daily Diarrhea Admissions

Diarrhea increases by 8% for each 1 ºC increase in temperature

Checkley et al. 2000

Number of Cholera cases in Uganda 1997-2002

0

10000

20000

30000

40000

50000

1996 1997 1998 1999 2000 2001 2002 2003

Time in years

Num

ber o

f cas

es

El Nino starts El Nino stops

Dr. Githeko, personal communication

Resources• McMichael AJ, Campbell-Lendrum DH, Corvalan

CF, Ebi KL, Githeko A, Scheraga JD, Woodward A (eds.). Climate Change and Human Health: Risks and Responses. WHO, Geneva, 2003.– Summary pdf available at

http://www.who.int/globalchange/publications/cchhsummary/

• Kovats RD, Ebi KL, Menne B. Methods of Assessing Human Health Vulnerability and Public Health Adaptation to Climate Change. WHO/Health Canada/UNEP, 2003.– Pdf available at http://www.who.dk/document/E81923.pdf

Health Data to Determine the Current Burden of Climate-

Sensitive Diseases

Questions to be Addressed

• What climate-sensitive diseases are important in your country or region?– What is the current burden of these diseases?

• What factors other than climate should be considered?– Water, sanitation, etc.

• Where are data available?• Are health services able to satisfy current

demands?

Health Data Sources

• World Health Report provides regional level data for all major diseases– http://www.who.int/whr/en– Annual data in Statistical Annex

• WHO databases– Malnutrition http://www.who.int/nutgrowth/db– Water and sanitation

http://www.who.int/entity/water_sanitation_health/database/en

• Ministry of Health– Disease surveillance/reporting branch

http://www.who.int/whr/en

http://www.who.int/nutgrowth/db





Health Data Sources - Other

• UNICEF at http://www.unicef.org

• CRED-EMDAT provides data on disasters– http://www.em-dat.net

• Mission hospitals

• Government district hospitals

http://www.unicef.org/



http://www.em-dat.net/

Mozambique

• Total population = 18,863,000• Annual population growth rate = 2.4%• Life expectancy at birth = 45 years• Under age 5 mortality rate = 158/1000

– 72% of 1-year-olds immunized with 3 doses of DTP

• 5.8% of gross domestic product spent on health

World Health Report 2005

WHO Region Afr-E (Countries with High Child & Very High Adult Mortality)

World Health Report 2004

Population 360,965,000

Total deaths 6,007,000

HIV/AIDS 1,616,000

Diarrheal diseases 356,000

Malaria 579,000

Protein-energy malnutrition

54,000

Seychelles National Communication

Methods and Tools for V&A Assessment in the Health

Sector

Methods and Tools

• Qualitative assessments• Methods of assessing human health

vulnerability to climate change• MARA/ARMA -- climate suitability for stable

malaria transmission• WHO Global Burden of Disease Comparative

Risk Assessment– Environmental Burden of Disease

• Other models

Qualitative Assessments

• Available data allows for qualitative assessment of vulnerability

• For example, given current burden of diarrheal diseases and projected changes in precipitation, will vulnerability likely remain the same, increase, or decrease?

Methods of Assessing Human Health Vulnerability

and Public Health Adaptation to Climate

Change

Kovats et al. 2003

Methods for:

• Estimating the current distribution and burden of climate-sensitive diseases

• Estimating future health impacts attributable to climate change

• Identifying current and future adaptation options to reduce the burden of disease

Kovats et al. 2003

Estimate Potential Future Health Impacts

• Requires using climate scenarios• Can use top-down or bottom-up approaches

– Models can be complex spatial models or be based on a simple exposure-response relationship

• Should include projections of how other relevant factors may change

• Uncertainty must be addressed explicitly

Kovats et al. 2003

Case Study: Risk of Vector-Borne Diseases in Portugal

• 4 qualitative scenarios developed of changes in climate and in vector populations– Vector not present– Focal distribution of vector– Widespread distribution of vector– Change from focal to potentially regional

distribution

• Expert judgment determined likely risk under each scenario for 5 vector-borne diseases

Kovats et al. 2003

Sources of Uncertainty

• Data– Missing data or errors in data

• Models– Uncertainty regarding predictability of the system– Uncertainty introduced by simplifying relationships

• Other– Inappropriate spatial or temporal data– Inappropriate assumptions– Uncertainty about predictive ability of scenarios

Kovats et al. 2003

Estimating the Global Health Impacts of Climate ChangeCampbell-Lendrum et al. 2003 (pdf available)

• What will be the total potential health impact caused by climate change (2000 to 2030)?

• How much of this could be avoided by reducing the risk factor (i.e. stabilizing greenhouse gas (GHG) emissions)?

Comparative Risk Assessment

2020s

2050s

2080s

Greenhouse gas emissions scenarios

Global climate modelling:

Generates series of maps of predicted future climate

Health impact model: Estimates the change in relative risk of specific diseases

Campbell-Lendrum et al. 2003

Time

2080s2050s2020s

Criteria for Selection of Health Outcomes

• Sensitive to climate variation

• Important global health burden

• Quantitative model available at the global scale– Malnutrition (prevalence)– Diarrhoeal disease (incidence)– VBD – dengue and Falciparum malaria – Inland and coastal floods (mortality)– Heat and cold related CVD mortality


Exposure: Alternative Future Projections of GHG Emissions

• Unmitigated current GHG emissions trends

• Stabilization at 750 ppm CO2-equivalent

• Stabilization at 550 ppm CO2-equivalent

• 1961-1990 levels of GHGs with associated climate

Source: UK Hadley Centre models


Relative Risk of Deaths and Injuries in Inland Floods in 2030, by Region

0

1

2

3

4

5

6

7

8

Afr

D

Afr

E

Am

r A

Am

r B

Am

r D

Em

r B

Em

r D

Eur

A

Eur

B

Eur

C

Sea

r B

Sea

r D

Wpr

A

Wpr

B

Rel

ativ

e R

isk

s550

s750

UE

Climate scenarios, as function of GHG emissions

Relative Risk of Diarrheoa in 2030, by Region

0.94

0.96

0.98

1

1.02

1.04

1.06

1.08

1.1

Afr

D

Afr

E

Am

r A

Am

r B

Am

r D

Em

r B

Em

r D

Eur

A

Eur

B

Eur

C

Sea

r B

Sea

r D

Wpr

A

Wpr

B

Rel

ativ

e R

isk

s550

s750

UE

Floods

Malaria

Diarrhea

Malnutrition

020406080100120 2 4 6 8 10

DALYs (millions)Deaths (thousands)

2000

2020

Estimated Death and DALYs Attributable to Climate Change


Conclusions• Climate change may already be causing a

significant burden in developing countries

• Unmitigated climate change is likely to cause significant public health impacts out to 2030– Largest impacts from diarrhea, malnutrition,

and vector-borne diseases

• Uncertainties include:– Uncertainties in projections– Effectiveness of interventions– Changes in non-climatic factors


Environmental Burden of Disease

• Introduction and Methods: Assessing the Environmental Burden of Disease at National and Local Levels by A Pruss-Ustun, C Mathers, C Corvalan, and A Woodward [pdf available at http://www.who.int/peh/burden/burdenindex.html]

• Climate change document will be published soon

The website [http://www.mara.org.za] contains prevalence and population data, and regional and county-level maps

Climate and Stable Malaria Transmission

• Climate suitability is a primary determinant of whether the conditions in a particular location are suitable for stable malaria transmission

• A change in temperature may lengthen or shorten the season in which mosquitoes or parasites can survive

• Changes in precipitation or temperature may result in conditions during the season of transmission that are conducive to increased or decreased parasite and vector populations

Climate and Stable Malaria Transmission (continued)

• Changes in precipitation or temperature may cause previously inhospitable altitudes or ecosystems to become conducive to transmission. Higher altitudes that were formerly too cold or desert fringes that were previously too dry for mosquito populations to develop may be rendered hospitable by small changes in temperature or precipitation.

MARA/ARMA Model

• Biological model that defines a set of decision rules based on minimum and mean temperature constraints on the development of the Plasmodium falciparum parasite and the Anopheles vector, and on precipitation constraints on the survival and breeding capacity of the mosquito

• CD-ROM $5 or can download components from website

Relationship Between Temperature and Daily Survivorship of Anopheles

0.000.100.200.300.400.500.600.700.800.901.00

Mean Temperature (°C)

Prop

ortio

n of

Mos

quito

es

Surv

ivin

g O

ne D

ay

Proportion of Vectors Surviving Time Required for Parasite Development

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40


Pro

port

ion

Sur

vivi

ng

Relationship Between Temperature and Time Required for Parasite Development

0

20

40

60

80

100

120

Mean Temperature (°C )

Days

Mozambique – Endemic Malaria Season Length

Mozambique – Endemic Malaria Prevalence

Mozambique – Endemic Malaria Prevalence by Age

Climate Suitability for Stable Malaria Transmission in Zimbabwe Under

Different Climate Change ScenariosEbi et al. Climatic Change

Objective: to look at the range of responses in the climatic suitability for stable falciparum malaria transmission under different climate change scenarios in Zimbabwe

Malaria in Zimbabwe

• Patterns of stable transmission follow pattern of precipitation and elevation (which in turn influences temperature)

• >9,500 deaths and 6.4 million cases between 1989-1996

• Recent high-altitude outbreaks

Cases by Month

Source:

South African Malaria Research Programme

Ebi et al. Climatic Change

Methods

• Baseline climatology determined

• COSMIC was used to generate Zimbabwe-specific scenarios of climate change; changes were added to baseline climatology

• Outputs from COSMIC were used as inputs for the MARA/ARMA (Mapping Malaria Risk in Africa) model of climate suitability for stable Plasmodium falciparum malaria transmission


Data Inputs

• Climate data– Mean 60 year climatology of Zimbabwe on

a 0.05° lat/long grid (1920-1980)– Monthly minimum and maximum

temperature and total precipitation

• COSMIC output– Projected mean monthly temperature and

precipitation (1990-2100)


Climate in Zimbabwe• Rainy warm austral summer October – April• Dry and cold May-September• Heterogeneous elevation-dictated temperature

range• Strong interannual and decadal variability in

precipitation • Decrease in precipitation in the last 100 years

(about 1% per decade) • Temperature changes 1933-1993

– Increase in maximum temperatures +0.6°C– Decrease in minimum temperatures –0.2 °C


GCMs

• Canadian Centre for Climate Research (CCC)• United Kingdom Meteorological Office

(UKMO)• Goddard Institute for Space Studies (GISS)• Henderson-Sellers model using the CCM1 at

NCAR (HEND)


Scenarios

• Climate sensitivity– High = 4.5ºC– Low = 1.4ºC

• Equivalent carbon dioxide (ECD) analogues to the 350 ppmv and 750 ppmv greenhouse gas emission stabilization scenarios of the IPCC SAR


Assumptions

• No change in the monthly range in minimum and maximum temperatures

• Permanent water bodies do not meet the precipitation requirements

• Climate did not change between the baseline (1920-1980) and 1990


Fuzzy Logic Value

• Fuzzy logic boundaries established for minimum, mean temperature and precipitation

• 0 = unsuitable

• 1 = suitable for seasonal endemic malaria


Assignment of Fuzzy Logic Values to Climate Variables

Fuzzy Logic Value for Mean Temperature

0

0.2

0.4

0.6

0.8

1

1.2

17.5

19.5

21.5

23.5

25.5

27.5

29.5

31.5

33.5

35.5

37.5

39.5


Fu

zzy

Val

ue

Fuzzy Logic Value for Precipitation

0

0.2

0.4

0.6

0.8

1

1.2

0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84

Precipitation (mm)

Fu

zzy

Val

ue

Fuzzy Logic Value for Minimum Temperature

0

0.2

0.4

0.6

0.8

1

1.2

3.5

3.7

3.9

4.1

4.3

4.5

4.7

4.9

5.1

5.3

5.5

5.7

5.9

6.1

6.3

6.5

Minimum Temperature (°C)

Fu

zzy

Val

ue

Climate Suitability Criteria

• Fuzzy values assigned to each grid• For each month, determined the lowest fuzzy

value for precipitation and mean temperature

• Determined moving 5-month minimum fuzzy values

• Compared these with the fuzzy value for the lowest monthly average of daily minimum temperature

• Assigned the lowest fuzzy valueEbi et al. Climatic Change

UKMO• S750 ECD stabilization scenario with 4.5°C

climate sensitivity• Model output

– Precipitation• Rainy season (ONDJFMA) increase in precipitation

of 8.5% from 1990 to 2100

– Temperature• Annual mean temperature increase by 3.5°C from

1990 to 2100, with October temperatures increasing more than July temperatures.


Baseline


2025


2050


2075


2100


Conclusions

• Assuming no future human-imposed constraints on malaria transmission, changes in temperature and precipitation could alter the geographic distribution of stable malaria transmission in Zimbabwe

• Among all scenarios, the highlands become more suitable for transmission

• The lowveld and areas currently limited by precipitation show varying degrees of change

• The results illustrate the importance of using several climate scenarios


Other Models

• MIASMA– Global malaria model

• CiMSiM and DENSim for dengue– Weather and habitat-driven entomological

simulation model that links with a simulation model of human population dynamics to project disease outbreaks

– http://daac.gsfc.nasa.gov/IDP/models/index.html

Sudan National Communication

• Using an Excel spreadsheet, modeled malaria based on relationships described in MIASMA

• Calculated monthly changes in transmission potential for the Kordofan Region for the years 2030-2060, relative to the period 1961-1990 using the IPCC IS92A scenario, simulation results of HADCM2, GFDL, and BMRC, and MAGICC/SCENGEN

Sudan – Projected Increase in Transmission Potential of Malaria in 2030

Sudan – Projected Increase in Transmission Potential of Malaria in 2060

Sudan – Malaria Projections

• Malaria in Kordofan Region could increase significantly during the winter months in the absence of effective adaptation measures– The transmission potential during these months is

75% higher than without climate change

• Under HADCM2, the transmission potential in 2060 is more than double baseline

• Transmission potential is projected to decrease during May-August due to increased temperature

Methods for Determining a Health Adaptation Baseline

Questions for Designing Adaptation Policies & Measures• Adaptation to what?• Is additional intervention needed?• What are the future projections for the outcome?

Who is vulnerable?– On scale relevant for adaptation

• Who adapts? How does adaptation occur?• When should interventions be implemented?• How good or likely is the adaptation?

Current and Future Adaptation Options

• What is being done now to reduce the burden of disease? How effective are these policies and measures?

• What measures should begin to be implemented to increase the range of possible future interventions?

• When and where should new policies be implemented?– Identify strengths and weaknesses, as well as threats

and opportunities to implementation

Kovats et al. 2003

Public Health Adaptation to Climate Change

• Existing risks– Modifying existing prevention strategies

– Reinstitute effective prevention programs that have been neglected or abandoned

– Apply win/win or no-regrets strategies

• New risks

Policy Analysis of Flooding Adaptation Strategies, Policies and Measures in

the UK

Theoretical Range of Choice

Technically feasibility demonstrated?

Economically feasible?

Socially and Legally Acceptable?

Effective to address health outcome?

Closed/Open (Practical Range of Choice)

Land use planning to reduce risk exposure

Yes at County and District levels only

Yes Yes Yes Open

Engineering works to reduce risk exposure

Yes Yes Yes Yes Open

Insurance Generally not available

Closed

Emergency relief

Yes Yes Yes Yes Open

Burton and Ebi, in preparation

Practical Range of Choice

Size of Events/ Exposure Intensity

Technically viable?

Economically possible (includes needed infrastructure available)?

Institutional support and human capital available?

Compatible with current policies?

Policy change needed?

Trans-boundary issue?

Land use planning to reduce risk exposure

Yes Yes Over 400 local planning authorities; little central coordination

Variable Variable No

Engineering works to reduce risk exposure

Yes Grant aid to supplement local resources for flood defense is provided only for capital schemes

Through Environment Agency and County Councils

Variable Variable No

Emergency relief

Yes Yes County and District Councils; emergency services; local and regional health authorities

Yes No No

Burton and Ebi, in preparation

Thank You

Methods and Tools for the Human Health Sector Kristie L. Ebi, Ph.D., MPH Washington, DC USA...

Documents

Transcript of Methods and Tools for the Human Health Sector Kristie L. Ebi, Ph.D., MPH Washington, DC USA...