important

1. Global Burden of Disease:
An Introduction
Kirk R. Smith
Professor of Global Environmental Health
Designing Strategies for Neglected Disease Research
Jan 20, 2009
Law 284.26, Public Policy 290, 190

2. What is health?
• “Health is a state of complete physical,
mental and social well-being and not merely
the absence of disease or infirmity.”
– First of nine principles on first page of World
Health Organization Constitution adopted in NYC
in July 1946 by 61 nations
– “spiritual well-being” added in 1999 by World
Health Assembly, which at that time had 191 member
states
• http://www.ldb.org/iphw/whoconst.htm

3. How would this be operationalized
for the following common queries?
• What is the total impact of disease and injury in
the population? -- the overall target for public
health interventions?
– Which diseases are most important for which groups?
– Are things getting better or worse?
• How do we compare the impacts of different risk
factors and potential interventions that affect
different populations?
– For example, what is the burden of disease from
environmental factors?
– How does the impact of tobacco smoking compare to
that from air pollution?

4. Environmental Health Effects
• Example of results from outdoor air pollution studies
– Asthma attacks
– Missing workdays
– Missing school days
– Days with cough
– Emergency room visits
– Hospital admissions
– Physician visits
– Medication use
– Daily death rate
– Lung function
– Self-reported health status
– Etc.
• How can these be compared across time, cities, countries, age
groups, sectors (e.g., transport versus power plants), etc.?
• Let alone compared with the health impacts from completely
different risk factors, such as water pollution, lead exposure, high
cholesterol, unsafe sex, etc.?

5. Ultimate Measure of Ill-health?
• Death is most common
– Easy to determine
– Commonly tabulated
• Severe problems as a measure
– Everyone dies
– Health never achieved
– Age is clearly important
• Deaths + Illness = ?

6. Combined Measure
• What else to use?
– Money? Are you kidding?
– Is used in legal and other realms, but not
appropriate for public health
• Most fundamental deprivation is loss of time:
– Same potential life length shared by all humans
– The degree to which a person does not achieve this life
length is a measure of ill-health
– Can be used for disabilities, as well, but need to weight
relative severity of disabilities as well as tabulate their
duration

7. Health Adjusted Life Years
HALY
• Basically the number of fully healthy life
years lost to a particular disease or risk
factor.
• Considers the age at which the disease or
death occurs and the duration and severity
of any disability created.

8. Global Burden of Disease
Database
• Developed at Harvard University originally for the
World Bank
• Extended greatly in the mid-1990s and now
adopted by the World Health Organization
– Updated database published on web each year and
summarized in World Health Report
• Dozens of countries now have NBDs
• Even states (provinces) and cities have them,
including SF and LA

9. Need for a C4 Database in Health
(Which we have had in many other fields for long periods)
• Combined mortality and morbidity
• Complete
– Much of the world unrepresented in past databases
– Many important disabilities unaccounted
• Consistent definitions of disease states
• Coherent
– Deaths by disease need to add to total
• By age and sex
• Match with demographic stats
– No natural discipline, i.e. no import stats from the
afterlife tabulating how many died of what

10. Just having coherence in mortality is valuable
Global Deaths in 2002
0
5
10
15
20
25
0-4 5-14 15-29 30-44 45-59 60-69 70+
Age Groups
Million
Deaths
LDCs
MDCs
Total Population
LDCs – 4.78 billion
MDCs – 1.45 billion
Total Global Deaths in 2002: 57 million

11. Disability Adjusted Life Year
The DALY, a kind of HALY
• Principle #1: The only differences in the rating of
a death or disability should be due to age and sex,
not to income, culture, location, social class.
• Principle #2: Everyone in the world has right to
best life expectancy in world
• DALY = YLL + YLD
– Years of Lost Life (due to mortality)
– Years Lost to Disability (due to injury & illness)

12. Years of Lost Life: Examples
Age at Death Female Male
0 82.5 80.0
1 81.8 79.5
5 78.0 75.4
15 68.0 65.4
25 58.2 55.5
35 48.4 45.6
50 34.0 31.0
80 8.9 7.5
100 2.0 1.5
100

13. What is Meant by “Disability?”
• Impairment: Symptoms at organ level, e.g.,
broken leg
• Disability: Objective alteration of behavior or
performance at the individual level, e.g., cannot
walk
• “Handicap”: Changed interaction with others at
the social/environmental level, e.g., cannot work
• http://www.disabilityhelper.com/Disability-
Impairment-Handicap.htm

14. Schema for Assessing Non-fatal
Health Outcomes
Disease Impairment
Polio Paralyzed
legs
Brain Mild mental
injury retardation
Disability “Handicap”
Inability Unemployed
to walk
Difficulty Social
learning isolation

15. Whom do you ask to determine
disability weights?
• Patient
• Family
• Caregiver
• Health professional
• Public health experts
• Public at large
• Insurance companies and lawyers (court cases)
Used in GBD

16. Time
Accident
Reported
Disability
Weight
1.0
When do you ask?

17. Classes of Disability Weights,
with examples
1: 0-0.02 Vitiligo on face
2: 0.02-0.12 Diarrhea, sore throat
3: 0.12-0.24 Radius fracture in stiff cast
4: 0.24-0.36 Below the knee amputation
5: 0.36-0.5 Down syndrome, COPD
6: 0.5-0.7 Unipolar depression, tetanus
7: 0.7-1.00 Psychosis, quadriplegia

18. Top Ten Causes of Disability
in 15-44 year olds (2000)
Male Female
Unipolar depression 13.9 18.6
Alcohol disorders 10.1
Schizophrenia 5 4.8
Bipolar depression 5 4.4
Fe-Deficiency anaemia 4.2 5.4
Hearing loss 4.1 3.6
Road traffic 3.8
HIV/AIDS 3.2 2.5
Drug use 3
COPD 2.6
Obstructed labor 4
Chlamydia 3.3
Abortion 3.1
Panic disorder 2.8
Percent of Total YLDs

19. Sample DALY Calculations
Diseases A and B
• A. 100,000 children are stricken for 1 week
with a disability weighting of 0.3; 2% die
at 1 year old.
• B. 100,000 adults are stricken for 2 years
with a disability weighting of 0.6; 20% die
at 80 years old.
• A: YLL (= 2000 x 80) + YLD (=100k x
(7/365) x 0.3) = 160,000 + 575 = 160,600
• B: YLL (= 20,000 x 8) + YLD (=100k x 2 x
0.6) = 160,000 + 120,000 = 280,000

20. Global Burden of Disease Database
World Health Organization
Being completely updated
2007-2009

21. Occam's Razor
• “One should not increase, beyond what is
necessary, the number of entities required to
explain anything”
• Occam's razor is a logical principle attributed to
the 14th Century philosopher William of Occam
(or Ockham). The principle states that one should
not make more assumptions than the minimum
needed. This principle is often called the
Principle of Parsimony

22. The DALY Passes Occam’s razor criterion,
because it reveals something different from deaths
Deaths DALYs
1 – Cancer 12.4% 5.3% (4)
2 - Heart 12.3% 3.8% (7)
3 - Stroke 9.2% 3.1% (9)
4 - ARI 7.1% 6.6% (1)
5 - HIV 5.3% 6.1% (3)
7 – Perinatal
8 - Diarrhea
? - Depression
4.4%
3.8%
0.03%
6.2% (2)
4.2% (6)
5.3 % (5)

23. Examples of Using a C4 database:
World DALYS Lost (2000)
Pop
million
Deaths
million
Deaths
per
1000
DALYs
million
DALYs
Per
Death
LDCs 4693 43.3 9.2 1256 29.0
MDCs 1352 12.4 9.2 216 17.4
World 6045 55.7 9.2 1472 26.4

24. Impact of Development on
Women and Children
W & C share
of Pop
W & C share
of DALYs
South Asia 66% 71%
Western Europe 60% 49%
World 65% 67%
Children under 15 years in 2000

25. 2000
0 5 10 15
Road Traf f ic
Child Cluster
TB
Diarrhea
P
erinatal
COP
D
HIV
ARI
Stroke
Cancer
Heart (Ischaem
ic)
Pe rce nt of Total
World Deaths in 2000

26. Child Cluster Diseases: the
World’s Largest Scandal
• 1.4 million children
• Rates in LDCs are
thousands of times those
in MDCs (Africa = 4700x
that of W. Europe)
• Vaccine coverage in
Africa went from 60% in
1990 to 46% in 1999
• Has stayed at 70% in
South Asia for many
years
Measles 777000
Tetanus 309000
Pertussis 296000
Diphtheria 3400
Poliomyelitis 675
Total Global Deaths in <5y

27. Relative Risks between Poor
Africa and USA
• Chance of woman dying in childbirth: 400
times greater
• Child dying of diarrhea: 400 times
• Of pneumonia: 500 times
• Of measles: 4000 times
• Similar in South Asia (India, Bangladesh,
etc)

28. 2000
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
Congenital
COPD
Maternal
TB
Malaria
Road Traffic
Stroke
Malnutrition
Child Clus ter
Heart (Is chaem ic)
Diarrhea
Cancer
Depres s ion
HIV
Perinatal
ARI
Percent of Total
World DALYs
in 2000
The major disease targets for public
health interventions in the world
today
Almost all
Women &
Children

29. 2000
0.0 5.0 10.0 15.0 20.0 25.0
Dementia
Diabetes
ARI
COPD
Stroke
Heart (Ischaemic)
Cancer
Percent of Total
2000
2000
North America - Deaths

30. 2000 North America - DALYs
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
Osteoarthritis
Hearing loss
COPD
Diabetes
Dementia
Road Traffic
Stroke
Alcohol
Heart (Ischaemic)
Depression
Cancer
Percent of Total
2000
2000
The major disease targets for
public health interventions
in the USA

31. 222
178
118
107
92
90
86
86
75
67
67
59
59
52
47
39
0 50 100 150 200 250
Very poor Africa
Poor Africa
Poor Eastern Med
Poor South Asia
All Men
Former USSR
Poor Latin America
All Women
Middle income South Asia
Middle income Latin America
Eastern Europe
China +
Middle income Eastern Med
North America
Western Europe
Japan, Australia +
DALD/capita
Disability Adjusted
Lost Days = DALY x 365
2000
Global

32. 0
100
200
300
400
500
600
Very poor Africa 548 57 174 264 191 193 218 212
Poor South Asia 307 44 75 80 114 157 169 159
Poor Latin America 198 34 74 75 93 120 140 160
Middle Income East Asia 146 24 44 43 67 107 139 158
North America 35 14 51 43 58 87 110 111
Western Europe 29 11 39 33 48 79 98 110
Japan/Australia 32 13 31 27 41 59 75 91
0-4 5-14 15-29 30-44 45-59 60-69 70-79 80+
DALDs
Per Capita
by Age
Group
Selected
World
Regions
2000
Annual loss per person

33. Time
The Classic Epidemiological Transition
Non-
Communicable
Diseases
Infectious Diseases

34. Cancer
CVD

35. Disease Categories
• I - Traditional, Communicable
– Infectious, maternal, perinatal, nutritional
• II - Modern, Non-communicable
– Cancer, heart, neuro-psychiatric, chronic lung, diabetes,
congenital
• III - Injuries, Non-Transitional
– Unintentional
• Motor vehicle, poisoning, falls, fire, drowning
– Intentional
• Suicide, violence, war

36. Classic Epi Transition
• I. Infectious diseases decline during
development
• II. Chronic disease rise during development
• III. Injuries show no pattern during
development and are thus “non-transitional”

37. Empirical Test of the Epi Tranistion
• Does it hold up to examination using the
first C4 database?
• Classic epidemiologic transition only deals
with mortality, thus here termed the
“Mortality Transition”
• “Epidemiologic Transition” here applied to
same evaluation using DALYs

38. Epidemiological Transition - Age Adjusted
98
9
32
46
168
112
87
120
111
129
29
12
25
28
37
0 50 100 150 200 250 300 350 400
World
High
Upper Middle
Lower Middle
Low
Income
Group
DALYs per thousand
I - "Infectious" II - "Chronic" III - Injuries

39. Epi Transition: Updated
• In terms of actual age-adjusted impact on
populations, all classes of disease decline during
development
– I. Declines dramatically at every level
– II. Declines slowly, but with little decline seen across
middle income regions
– III. Declines in a similar way to II and thus is not “non-
transitional”
• Better to be rich for all major types of ill-health,
although there are exceptions for individual
diseases

40. Comparison of GBD Estimates
for 2005 with GBD for 1990
• Population: 5.3/6.4 billion (+21%)
• Deaths: 50/64 million (+28%)
• DALYs: +7%
• DALYS/capita: -11%
• I = 44/38.5%;
• II = 41/48.9%;
• III = 15/12.5%
WHO Databases

41. Changes in Important Diseases: 1990-2005
What is happening with each?
• Diarrhea: 7.3/3.9% (-42% in absolute terms)
• ARI: 8.5/5.9% (-25%)
• Malaria: 2.3/2.3% (-6%)
• Lung Cancer: 0.65/0.8% (+32%)
• TB: 2.8/2.1% (-18%)
• HIV: 0.8/5.6% (7.4 times as much)
• Depression: 4.7/5.8 (+29%)
WHO Databases

42. Can we reach public health?
• Is there a absolute value of health (lost
DALYs) beyond which society does not
have an obligation to exceed?
• Is there a cost per unit improvement in
health ($ per DALY) above which society
does not benefit from further expenditure?

43. World Health
Reports – 2002,
2001
4.9 million deaths/y
Global Burden of Disease from Top 10 Risk Factors
plus selected other risk factors
0% 2% 4% 6% 8% 10%
Climate change
Urban outdoor air pollution
Lead (Pb) pollution
Physical inactivity
Road traffic accidents*
Occupational hazarads (5 kinds)
Overweight
Indoor smoke from solid fuels
Lack of Malaria control*
Cholesterol
Child cluster vaccination*
Unsafe water/sanitation
Alcohol
Tobacco
Blood pressure
Unsafe sex
Underweight
Percent of All DALYs in 2000
Environmental Risk Factors

44. Entry into GBD databases
• Best single modern book covering the GBD and CRA ideas, methods, and
results, but without full detail and sophistication/complexity: Global
Burden of Disease and Risk Factors, (Lopez, Mathers, Ezzati, Jamison,
Murray) Oxford University and World Bank Presses, 2006. 475
pp. Fully downloadable at http://www.dcp2.org/pubs/GBD which also
has links to data used in the book.
• Best single page to find GBD data divided by world regions defined in
several ways (WHO regions, World Bank regions, income groups etc.) for
2004.
http://www.who.int/healthinfo/global_burden_disease/2004_report_updat
e/en/index.html
• For projections to 2030 and links to dozens of other publications, see
http://www.who.int/healthinfo/global_burden_disease/en/index.html
• The full set of background materials and pubs of the previous (2004)
Comparative Risk Assessment (CRA) covering 26 major risk factors,
environmental and other:
http://www.who.int/healthinfo/global_burden_disease/cra/en/index.html
• Full databases for the previous CRA study:
http://www.who.int/healthinfo/global_burden_disease/risk_factors/en/inde
x.html
• Description of the GBD/CRA 2005 Revisions now underway:
http://www.who.int/healthinfo/global_burden_disease/GBD_2005_study/
en/index.html

45. Thank you.
Kirk R. Smith
krksmith@berkeley.edu
http://ehs.sph.berkeley.edu/krsmith/