Estimates of DALE for 191 countries:
methods and results.
Colin D Mathers
Ritu Sadana
Joshua A Salomon
Christopher JL Murray
Alan D Lopez
Global Programme on Evidence for Health Policy Working Paper No. 16
World Health Organization
June 2000
1. Introduction
The annual assessment of world health is a key component of the global public policy process
to improve health levels and reduce health inequalities. Current estimates of death and
disability in countries disaggregated by age, sex and cause, are useful for several public health
purposes, ranging from the monitoring of new epidemics to progress in reducing old ones for
which disease control programmes are in place in countries. To adequately describe health
patterns in almost 200 countries according to age, sex and cause, a vast array of estimates
need to be generated. It then becomes extremely difficult to ascertain the main findings of
such a review unless the data are summarised in some fashion.
In this year´s World Health Report [1], the primary summary measure of population health
used is Disability-Adjusted Life Expectancy, or DALE. DALE measures the equivalent
number of years of life expected to be lived in full health, or healthy life expectancy. This
technical paper provides details of the methods and data sources used to prepare the DALE
estimates for the 191 member countries of WHO. In constructing the estimates, we sought to
address some of the methodological challenges regarding comparability of the health status
data collected [2].
The Global Burden of Disease project developed two summary measures, the DisabilityAdjusted Life Year (DALY) and Disability-Adjusted Life Expectancy (DALE), to provide a
comprehensive assessment of the global burden of disease and injury [3, 4], to inform global
priority setting for health research [5], and to report on trends in population health across the
world [1, 6]. Both these summary measures of population health (SMPH) combine
information on the impact of premature death and of disability and other non-fatal health
outcomes. The burden of disease methodology provides a way to link information at the
population level on disease causes and occurrence to information on both short-term and longterm health outcomes, including impairments, functional limitations (disability), restrictions
in participation in usual roles (handicap), and death.
DALYs are a gap measure; they measure the gap between a population's actual health and
some defined goal, while DALE belongs to the family of health expectancies, summarizing
the expected number of years to be lived in what might be termed the equivalent of "full
health". Both DALE and DALYs require a number of social value choices relating among
other things, to the valuation of time spent in states of health worse than ideal health, the
definition of an implied norm for population health, and the weighting of years of life lived at
different ages.
Murray and Lopez [7, 34] published disability-adjusted life expectancy (DALE) estimates for
the eight regions of the world based on the estimates of severity-weighted disability
prevalence developed for the non-fatal component of disease and injury burden. As a
summary measure of the burden of disability from all causes in a population, DALE has two
advantages over other summary measures. The first is that it is relatively easy to explain the
concept of an equivalent “healthy” life expectancy to a non-technical audience. The
increasing popularity of health expectancy indicators among policy makers has been
documented (van de Water et al. 1996; Barendregt et al. 1998) [8, 9]. The second is that
DALE is measured in units (expected years of life) that are meaningful to and within the
common experience of non-technical audiences (unlike other indicators such as health gaps,
mortality rates or incidence rates).
1
Improving the overall level of population health has been identified by WHO as one of the
five intrinsic goals of health systems (see Section 1.3 below), and DALE provides the best
available SMPH for measuring the overall level of health for populations in a way that is
appropriately sensitive to probabilities of survival and death and to the prevalence and
severity of health states among the population.
1.1 Background
In the last two decades, considerable international effort has been put into the development of
summary measures of population health (SMPH) that integrate information of mortality and
non-fatal health outcomes and international policy interest in such indicators is increasing
[10]. Efforts to develop summary measures of population health have a long history [11–17].
In the past decade, there has been a markedly increased interest in the development,
calculation and use of summary measures.
The concept of combining population health state prevalence data with mortality data in a
lifetable to generate estimates of expected years of life in various health states (health
expectancies) was first proposed in the 1960s [11, 13] and Disability-free Life Expectancy
was calculated for a number of countries during the 1980s. An informal international research
network, the Network on Health Expectancy (Réseau Espérance de Vie en Santé or REVES)
was established in 1989 with objectives including the harmonisation of calculation methods
and identification of the conditions necessary for comparison of health expectancy estimates,
both across populations and over time [18–22].
During the 1990s, Disability-Free Life Expectancy (DFLE) and related measures were
calculated for many countries [23–26]. In 1993, OECD included disability-free life
expectancy among the health indicators reported in its health database [27] and by 1999 the
number of countries for which some estimates of disability-free life were available had grown
to 29 [28].
However, DFLE and related measures incorporate a dichotomous weighting scheme, i.e., that
does not account for varying levels of severity (see Section 1.2 below for more detail on this).
The threshold definition of disability, therefore, has a dramatic effect on the results [29].
Wilkins and Adams [30] suggested a more sensitive weighting scheme based on the severity
of functional limitations, leading to the disability-adjusted life expectancy (DALE) approach.
DALE are described in more detail in the following section.
Another type of summary measure, Disability-Adjusted Life Years (DALYs) has been used in
the Global Burden of Disease Study [7, 31–36] and in a number of National Burden of
Disease Studies [37–46]. Reflecting this rising interest in the academic and policy
communities, the United States’ Institute of Medicine convened a panel on summary measures
and published a report that included recommendations to enhance public discussion of the
ethical assumptions and value judgements, establish standards, and invest in education and
training to promote use of summary measures. [10]. More recently, WHO convened a
conference of experts across a range of disciplines including descriptive epidemiology, public
health, health economics and philosophy and ethics to discuss issues around the conceptual,
technical and ethical basis for summary measures of population health. A book addressing
these issues based on the papers presented at Marrakech is in preparation [47].
Interest in summary measures relates to a range of potential uses. Murray, Salomon and
Mathers [48] identified eight of these:
1) Comparing the health of one population to the health of another population.
2
2) Comparing the health of the same population at different points in time.
3) Identifying and quantifying overall health inequalities within populations.
4) Providing appropriate and balanced attention to the effects of non-fatal health
outcomes on overall population health.
5) Informing debates on priorities for service delivery and planning.
6) Informing debates on priorities for research and development in the health sector.
7) Improving professional training curricula in public health.
8) Analyzing the benefits of health interventions for use in cost-effectiveness analyses.
Broad interest and use of summary measures in the policy arena demonstrates the recognition
of their value at the practical level for many of these purposes. The World Health Report 2000
has used DALE as a summary measure of the level of population health in member countries
in order to provide a comparative assessment of levels of health (use 1 above), and as a
component of the composite health goal performance measure (see section 1.3 below). Over
time, successive reporting on DALE will provide evidence of progress towards achieving
global goals for improving health (use 2 above).
1.2 Relationship of DALE to other SMPH
The Global Burden of Disease Study (GBD) developed a new SMPH, the Disability-Adjusted
Life Expectancy (DALY). For a review of the development of DALYs and recent advances in
disease burden measurement, see Murray and Lopez [3, 49]. The GBD also used DisabilityAdjusted Life Expectancy (DALE) as a simple summary for comparative purposes across
populations. This section explains the relationship between these two summary measures, the
comparative advantages of each, and why WHO is using both types of indicators in its annual
assessment of the health situation of member countries and in the assessment of health system
performance.
Health expectancies and health gaps
On the basis of a simple survivorship curve, SMPH can be divided broadly into two families:
health expectancies and health gaps. The bold curve in Figure 1 is an example of a
survivorship curve S(x) for a hypothetical population. The survivorship curve indicates, for
each age x along the x-axis, the proportion of an initial birth cohort that will remain alive at
that age. The area under the survivorship function is divided into two components, A which is
time lived in full health and B which is time lived at each age in a health state less than full
health. The familiar measure of life expectancy at birth is simply equal to A+B (the total area
under the survivorship curve. A health expectancy is generally of the form:
Health expectancy = A + f(B)
(1)
where f(.) is a function that weights time spent in B by the severity of the health states that B
represents. When a set of health state valuations are used to weight time spent in health states
worse than ideal health, the health expectancy is referred to as a health-adjusted or disabilityadjusted life expectancy (DALE). Another type of health expectancy is exemplified by
disability-free life expectancy in which time spent in any health state categorized as disabled
3
is assigned arbitrarily a weight of zero, and time spent in any state categorized as not disabled
is assigned a weight of one (i.e., equivalent to full health).
4
Figure 1. Survivorship function for a population
Survivors (%)
100
90
C
80
B
70
60
A
50
(Full health)
40
30
20
10
0
0
10
20
30
40
50
60
70
80
90
100
Age (years)
In contrast to health expectancies, health gaps quantify the difference between the actual
health of a population and some stated norm or goal for population health. The health goal
implied by Figure 1 is for everyone in the entire population to live in ideal health until the age
indicated by the vertical line enclosing area C at the right1. In the specific example shown, the
normative goal has been set as survival in full health until age 100. By selecting a normative
goal for population health, the gap between this normative goal and current survival, area C,
quantifies premature mortality. A health gap is generally of the form:
Health gap = C + g(B)
(2)
where g(.) is a function that weights time spent in B by the severity of the health states that B
represents. Note that because health gaps measure a negative entity, namely the gap between
current conditions and some established norm for the population, the weighting of time spent
in B is on a reversed scale as compared to the weighting of time spent in B for a health
expectancy. More precisely, full health is 1 in a health expectancy, whereas death or a state
equivalent to death is 1 in a health gap. Because health gaps measure the distance between
current health conditions and a population norm for health, they are clearly a normative
measure.
Years of life lost measures are all measures of a mortality gap, or the area between the
survivorship function and some implied target survivorship function (area C in Figure 1).
Mortality gap measures were first suggested by Dempsey (1947) [50] and potential years of
life lost has been extensively used as a population health indicator since its first calculation by
Romeder and McWhinnie [51]. Murray [3] and others [16, 52] have since proposed and
calculated a variety of health gaps.
1 Figure 1 graphically illustrates the magnitude of both health expectancies and health gaps only when
a population has a stable distribution with a zero population growth rate. In practice, health
expectancies are not sensitive to differences in the age structure of different populations. Health gaps
are usually reported in absolute terms so that health gaps are sensitive to variations in the age
distribution of different populations although age independent forms of health gaps can be
formulated.
5
Health state expectancies and disability-adjusted life expectancies
We can categorise health expectancies into two main classes: those that use dichotomous
health state weights and those that use health state valuations for an exhaustive set of health
states. Examples of the first class include:
Disability-free life expectancy: This health expectancy gives a weight of 1 to states of health
with no disability (above an explicit or implicit threshold) and a weight of 0 to states of health
with any level of disability above the threshold. Other examples of this type of health
expectancy include active life expectancy, independent life expectancy and dementia-free life
expectancy.
Life expectancy with disability: This is an example of a health expectancy which gives 0
weight to all states of health apart from one specified state of less than full health (in this case,
disability above a certain threshold of severity). If health state 3 in Figure 2 is ‘moderate
disability’ then the area under the survival curve corresponding to health state 3 represents life
expectancy with moderate disability. Other examples of this type of health expectancy include
handicap expectancy, severe handicap expectancy and unhealthy life expectancy.
Examples of the second type of indicator include:
Health-adjusted life expectancies: These have been calculated for Canada and Australia using
population survey data on the prevalence of disability at four levels of severity together with
more or less arbitrary severity weights [53–55]. More recently, Canada has produced the first
estimates of health-adjusted life expectancy based on population prevalence data for health
states together with measured utility weights [56].
Disability-adjusted life expectancy: This was calculated for the Global Burden of Disease
Study using disability weights reflecting social preferences for seven severity levels of
disability [7]. DALE has also been calculated for Australia using prevalence data from the
Australian Burden of Disease Study [46] and preference weights derived from the Global
Burden of Disease Study and from a Dutch study using similar valuation methods [57].
Although health states form a continuum, in practice they are generally conceptualised and
measured as a set of mutually exclusive and exhaustive discrete states ordered on one or more
dimensions. If we enumerate health states using a discrete index h, then we can calculate
disability-adjusted life expectancy as:
L
DALE x = å ò wh (u ) * S h (u ) du
h
(3)
x
where u represents age and the integral is over ages from x onwards. If the weight wh for state
h is independent of age u, then
L
ö
æ
ç
DALE x = å wh * ò S h (u ) du ÷ = å wh *HE hx
÷ h
ç
h
x
ø
è
(4)
where HEhx is the health state expectancy at age x for years lived in state h.
In terms of the four health states illustrated in Figure 2, if HE1,0 to HE4,0 are the health state
expectancies at birth for each of the four states, and we give age-independent weights w2, w3,
w4 (less than 1) to the three states of less than full health, then the disability-adjusted life
expectancy at birth and total life expectancy at birth are given by:
DALE 0 = HE1,0 + w2 * HE 2, 0 + w3 * HE3,0 + w4 * HE 4, 0
6
(5)
(6)
LE 0 = HE1,0 + HE 2, 0 + HE3, 0 + HE 4,0
Figure 2. Survivorship functions for four health states
Survivors (%)
100
Health state 2
90
80
Health state 3
70
60
Health state 4
50
40
30
20
10
0
0
10
20
30
40
50
60
70
80
90
100
Age (years)
Terminology
In the mid-1990s, REVES developed a set of recommendations for terminology that was
widely adopted [58]. With the development of health gaps measures in the 1990s, there has
been some shift in the use of these terms, and health expectancy is now used to denote the
general class of summary measures that relate to the area under the survival curve. We use the
revised terminology proposed by Mathers [59]:
Health expectancy (HE): Generic term for summary measures of population health that
estimate the expectation of years of life lived in various health states.
Health state expectancy: Generic term for health expectancies which measure the expectation
of years lived in a single specified health state (eg. Disability-free).
Disability-adjusted life expectancy (DALE): General term for health expectancies which
estimate the expectation of equivalent years of good health based on
an exhaustive set of health states and weights defined in terms of
health state valuations. Health-adjusted life expectancy (HALE) is a
synonym for DALE.
Healthy life expectancy: Used as a simple synonym for DALE.
Critical appraisal of health expectancies
Murray, Salomon and Mathers [48] proposed a set of desirable properties for evaluating
summary measures of population health (SMPH) based on common sense notions of
population health of the following type:
7
If two populations are identical in every way except that infant mortality
is higher in one, then we expect that everybody would agree that the
population with the lower infant mortality is healthier.
They suggested a minimal set of desirable properties for summary measures that will be used
to compare the health of populations:
1. If age-specific mortality decreases in any age-group, everything else being the same, then
a summary measure should improve (i.e. a health gap should decrease and a health
expectancy should increase)2.
2. If age-specific prevalence of some health state worse than ideal health increases,
everything else being the same, a summary measure should get worse.
3. If age-specific incidence of some health state worse than ideal health increases,
everything else being the same, a summary measure should get worse.
4. If age-specific remission for some health state worse than ideal health increases,
everything else being the same, a summary measure should improve.
5. If the severity of a given health state worsens, everything else being the same, then a
summary measure should get worse.
Mathers [59] has assessed health expectancies against these five criteria. All health
expectancies meet criterion 1. Health expectancies based on prevalence data (for example,
those calculated using Sullivan’s method) meet criteria 1 and 2 but fail criteria 3 and 4 (until
prevalence rates change to reflect the change in transition rates). Health expectancies based on
transition rates (for example, those calculated using the multistate life table method) meet
criteria 1, 3 and 4 but fail criterion 2. Disability-adjusted life expectancies (DALE) meet
criterion 5, whereas health expectancies using dichotomous health state weights (eg.
disability-free life expectancy) do not. Table 1 summarises these conclusions.
Table 1. SMPH criteria met by various forms of health expectancies
Health state expectancies
Dichotomous weights
(eg. DFLE)
Disability -adjusted life expectancies
Polytomous weights
(eg. DALE)
Prevalence-based measures
1, 2
1, 2, 5
Transition-rate based measures
1, 3, 4
1, 3, 4, 5
Health state expectancies such as DFLE give an implicit value of zero (equivalent to the
valuation of death) for disability above a certain threshold, below this threshold the valuation
is 1. This means that the summary indicator is not sensitive to changes in the severity
distribution of disability within a population (criterion 5). The overall DFLE value for a
population is largely determined by the prevalence of the milder levels of disability and
comparability between populations or over time is highly sensitive to the performance of the
disability instrument in classifying people around the threshold. For this reason, Murray,
Salomon and Mathers [48] concluded that health state expectancies are not appropriate for use
2 This criterion could be weakened to say that if age-specific mortality decreases in any age-group,
everything else being the same, then a summary measure should improve or stay the same. The weaker
version would allow for deaths beyond some critical age to leave a summary measure unchanged.
Measures such as potential years of life lost would then fulfil the weak criterion.
8
as SMPH, and that DALE is the most appropriate form of health expectancy for use as an
SMPH.
Murray, Salomon and Mathers [48] proposed two other desirable attributes of summary
measures that are to be used to inform policy discussions. These are not attributes based on
arguments about whether a population is healthier than another but rather on practical
considerations:
1. Summary measures should be comprehensible and feasible to calculate for many
populations. Comprehensibility and complexity are different. Life expectancy at birth is a
complex abstract measure but is easy to understand. Health expectancies are popular
because they are also easily understood.
2. Summary measures should be linear aggregates of the summary measures calculated for
any arbitrary partitioning of sub-groups. Many decision-makers, and very often the public,
desire information that is characterized by this type of additive decomposition. In other
words, they would like to be able to answer what fraction of the summary measure is
related to health events in the poor, in the uninsured, in the elderly, in children, and so on.
Additive decomposition is also often appealing for cause attribution.
Most health expectancies satisfy the first attribute. However, they cannot be additively
decomposed in respect of causes or population sub-groups. Disability-adjusted life
expectancies are additively decomposable into health expectancies for specified levels of
disability severity (see above). This form of decomposition may be useful in understanding
which levels of disability severity are contributing most to changes in population health.
Health state expectancies should thus be understood as a decomposition of a DALE summary
measure than as SMPH in themselves. This interpretation is consistent with the usual ways in
which families of health state expectancies are presented for a population [60, 61].
In general, health gaps can be decomposed into the contribution of various causes in a more
intuitive and easily communicated fashion than health expectancies. DALYs are additive
across causes to give the total health gap for a population. Disability-adjusted life expectancy
and a health gap measure such as the DALY thus fulfill different needs for SMPH to
summarise and report on trends and achievements in population health across countries.
Wolfson [62] outlined a vision of a coherent and integrated statistical framework, with
summary measures of population health status at the apex of a hierarchy of related measures,
rather than a piecemeal set of unconnected measures. The macro measures at the apex of the
system, such as DALE and DALYs, provide a broad population-based overview of trends and
causes (Figure 3). DALE would be used for monitoring overall progress in improving the
level and distribution of health, and DALYs would be used for quantifying the causes of
health losses, for identifying the potential for health gain and for linking health interventions
to changes in population health.
As shown in Figure 3, such a system should include the capability to ‘drill down’ below the
summary measures to component parts such as incidence rates, prevalence rates, severity
distributions, case fatality rates, etc. It should also allow us to ‘drill down’ below whole of
population level to examine inequalities in health and to estimate the impacts of a given
intervention on various sub-groups.
9
Figure 3. The pyramid of population health measures
Health states and valuations
Prevalence
Duration
Incidence, remission
Case fatality
Mortality
Ethnic,
etc.
Structural
Behaviour, environme
Socioeconomic
groups
Physiological and
pathophysiological
Geographic
Diseases
Injuries
Age and sex
Impairments
Disabilities
1.3 Role as a measure of health system goal performance
The World Health Report [1] has carried out an assessment of the performance of health
systems of member countries in achieving three main (intrinsic) goals for the health system:
health, responsiveness and fairness in financing. WHO’s work on operationalizing the
measurement of goal attainment is focussed on measuring these three goals as well as relating
goal attainment to resource use in order to evaluate performance and efficiency [1, 63].
In operationalizing the assessment of level of health for member countries, WHO has chosen
to use DALE as the summary measure for the reasons outlined above. In the World Health
Report 2000, health inequality is assessed for member countries in terms of child mortality
inequality. In future assessments, it is planned to move to the use of more comprehensive
indicators based on inequality in DALE within the population [64–66].
A third related concept is efficiency or composite goal performance. Efficiency is how well
we achieve the socially desired mix of the five components of the three goals compared to the
available resources. DALE is one of the measures used in the development of the composite
measure of health system goal attainment and in the analysis of health system performance.
Composite goal performance and individual goal performance are discussed in more detail in
the World Health Report [1] and in related technical papers.
10
2. Previous approaches to the calculation of DALE
A key step in the construction of a health expectancy or a health gap is comparing time lived
in a health state worse than full health with time lived in full health (in health expectancies)
and with time lost due to premature mortality, compared to some normative goal (in a health
gap). Two sets of issues are common to both health expectancies and health gaps: the
conceptual framework and measurement strategy to describe health states and the conceptual
framework and measurement strategy to value time spent in health states. We first review
general issues in measuring and valuing health states in Section 2.1. We then review the two
main approaches which have been used for calculating DALE based on whether the health
state prevalences are derived from population health/disability surveys (described in Section
2.2) or from a disease-specific approach based in a full burden of disease analysis (described
in Section 2.3).
2.1 Measurement and valuation of health states
The literature on both description and valuation of health states is vast and rapidly expanding
[67–70]. Murray [3] provides a more detailed discussion with regards to the original GBD
approach.
Defining and measuring health status
Until recently, most health expectancies have been defined in terms of disability (functional
limitations), or handicap (role limitations, dependence, restrictions in participation). In its
early meetings, the Network on Health Expectancy (REVES) agreed that the WHO
International Classification of Impairments, Disabilities, and Handicaps or ICIDH [71] should
provide the conceptual framework for the development of health expectancy indicators based
on impairment, disability and handicap states. The original ICIDH framework recognised four
dimension - diseases or disorders, impairments, disabilities and handicaps. In the context of a
health condition (disease or disorder), impairment corresponds to any loss or abnormality of
psychological, physiological or anatomical structure or function; disability corresponds to any
restriction or lack (resulting from an impairment) of ability to perform an activity in the
manner or within the range considered normal for a human being; and handicap or social
disadvantage for a given individual results from an impairment or a disability that limits or
prevents the fulfilment of a normal role (depending on age, sex and social and cultural
factors). Handicap is characterised by discordance between the activity and status of the
individual and the expectations of his or her social environment [71].
The beta draft revision of the ICIDH-2 [72] replaces the concept of handicap by the concept of
social participation and includes limitations in performing more complex activities (formerly
handicaps) as types of activity (the concept replacing disability). Impairments are renamed
functional abilities. The ICIDH-2 classifies domains of functioning rather than persons and is
a classification scheme rather than a health state measurement instrument. It thus cannot be
used directly to classify persons according to health state for constructing summary population
measures of health. Robine and Jagger [25] have reviewed the ICIDH and other models of the
disablement process and note that there is considerable confusion and disagreement over the
boundaries between impairment and disability, and disability and handicap, particularly in
relation to where functional limitations and complex activity restrictions fall.
11
A wide range of instruments have been developed in various languages to use individual
responses to measure various dimensions or domains of health states. Some of the more
widely used instruments are summarised in Table 2. Some instruments sacrifice the capacity
to discriminate between health states by restricting the number of questions or items in the
survey and restricting the number of response categories in order to increase measured
reliability – for example, this is the strategy used in the Euroqol EQ5D, which includes five
Table 2. Health domains included in 12 generic health status measurement instruments.
Health Domains
QWB
McM
SIP
QLI
NHP
FSQ
CP
Duke
SF-36
(multi-dimensional profile)
1970
1976
1976
1981
1981
1986
1987
1990
1992
ü
ü
ü
General Health
ü
ü
Physical Health
ü
ü
ü
Work
ü
Home
ü
Recreation
ü
Ambulation
ü
Eating
ü
ü
ü
ü
ü
ü
ü
ü
ü
ü
Pain/discomfort
ü
Self Care
ü
ü
Sleep/Rest
ü
ü
ü
ü
ü
ü
ü
ü
ü
ü
ü
ü
ü
Communication
ü
Interaction
ü
ü
ü
ü
ü
ü
ü
Support
ü
ü
ü
Mental Health
ü
ü
ü
ü
Activities/roles
ü
Alertness
ü
Anxiety/Depression
ü
ü
Cognition
ü
ü
ü
ü
ü
ü
ü
Self-esteem
Handicap/Participation
ü
ü
ü
Social Health
Outlook
ü
ü
ü
Energy/vitality
Emotional status
ü
ü
Change in Health
Activities/roles
ü
ü
Perceived Health
Mobility/fitness
1999
WHO
DASII
1999
EQ6D
ü
Overall Well-Being
Activities/roles
WHO
QOL
1996
ü
ü
ü
Environmental Context
Source: Sadana, 2000. QWB:Quality of Well-Being Scale, McM: McMaster Health Index, SIP: Sickness Impact Profile,
QLI: Quality of Life Index, NHP: Nottingham Health Profile, FSQ: Functional Status Questionnaire, CP: COOP Charts for
Primary Care Practice, Duke: Duke Health Profile; SF-36:Short-Form 36 Health Survey, EQ6D: EuroQol 6 domain Quality
of Life Scale, WHOQOL: WHO Quality of Life Bref Field Trial Version, WHODAS II: WHO Disability Assessment
Schedule. Souvce: [73]
12
domains with three level categories on each [74]. Other instruments such as SF-36 have many
more items and more response categories per item. Increased discriminatory power often
comes at the price of increased complexity, which may have important implications for
valuation to time spent in a health state.
A fundamental problem with current self-reported instruments is a lack of cross-cultural
comparability (including comparisons of the same community over periods long enough that
cultural health norms may have changed). This is not simply a question of language and the
interpretation of the meaning of different categorical responses in different languages. The
endpoints of scales for a given domain such as best or worst mobility may also have very
different meanings across different cultures or across socio-economic groups within a society.
This is discussed further in Section 2.2 and is examined in detail by Sadana et al [2] who have
analysed over 60 representative health surveys.
The challenge for developing a profile for standardising health state descriptions is to include
all domains considered to be important in terms of societal health goals and in terms of health
state valuation and to develop methods for measuring each of these domains for individuals in
ways that maximise comparability across population groups. WHO is proposing as part of the
Global Burden of Disease 2000 project to develop a standardized description of health states
covering a broad range of health states for use in population surveys and health state
valuation. The challenge in seeking a standardized description involves trade-offs between
completeness of description and parsimony. Other desirable properties include cross-cultural
validity, and usability by younger and older adults with widely varying education levels and
cultural backgrounds.
Valuing health states
In order to use time as a common currency for years of life lived in various states of health
and for time lost due to premature mortality, we must numerically value time lived in nonfatal health states. The health state valuations (or disability weights) used in DALY and
DALE calculations represent societal preferences for different health states. They range from
0 representing a state of good or ideal health (preferred to all other states) to 1 representing
states equivalent to being dead. These weights do not represent the lived experience of any
disability or health state, or imply any societal value of the person in a disability or health
state. Rather they quantify societal preferences for health states in relation to the societal
‘ideal’ of good health.
At the time that the GBD was underway, and even today, there is no body of empirical
measurement of health state descriptions and valuations that can be used (a) to describe the
average health state in multiple domains associated with different diseases, injuries and risk
factors and (b) to value these average health states. As an effort to provide a practical interim
solution to these major data deficiencies, the GBD used a multiple methods (ordinal rankings,
two forms of person trade-off, time trade-off and visual analogue) approach with small groups
of public health professionals to measure values for approximately 20 indicator health states
ranging from mild to severe [3]. A deliberative approach was used with small groups in order
to ensure that the people involved understood and were aware of the implications of their
choices. Final weights for conditions were based on the person trade-off estimates in order to
reflect social rather than individual preferences for health states. Other health conditions were
valued by ordinal ranking against the indicator conditions.
There is a growing consensus among health economists that health state preferences should
reflect the preferences of the general population when they are to be used as part of a process
of broad health policy assessment, priority setting or resource allocation [75–76]. Health
13
experts were used in the GBD valuation exercise for convenience reasons due to the practical
difficulties in ensuring that lay persons fully understood the impact and severity distribution
of the conditions being valued. The Disability Weights Project for Diseases in the Netherlands
[57] attempted to address this problem by defining the distribution of health states associated
with a health condition by using the modified EuroQol health profile to describe the health
states. Few differences were seen in the average PTO preferences assigned by a lay panel
(people with an academic background but no medical knowledge) compared with those of two
panels of physicians. The Dutch study concluded that it makes little difference whether the
valuation panel is composed of health care experts or lay people, as long as accurate
functional health state descriptions are included in the specifications of the health problems
being valued.
Since the development of the original protocol for health state value measurement in the
GBD, a series of convenience samples of international public health practitioners has been
organized, and a number of modifications and refinements of the original protocol have been
examined. In eleven different groups, valuations for 15 to 22 states – with a set of 14 states
common to all exercises – have been measured using a multi-method approach with internal
consistency checks and group discussions. The study locations have included the United
States, Mexico, Brazil, the Maghreb countries (Morocco, Algeria and Tunisia), Japan,
Australia, the Netherlands, and four multi-national groups of health care practitioners.
Murray and Lopez [49] compared the median health state valuations for each state across ten
of these groups. Overall, the intraclass correlation coefficient for the ten studies was 0.954,
indicating that this measurement approach yields similar values in groups from very different
communities. The work completed by Ustun and colleagues in 14 countries [77], which
measures rank correlations for a set of 17 health conditions, provides further evidence that
valuations of health states appear to be quite stable across diverse settings. As further large
scale empirical studies are undertaken in different countries, it is likely that some important
variations in average health state valuations will be found, particularly with respect to the
contribution of selected domains such as sexual function or pain (see for example [78]).
Nevertheless, it is unlikely that the magnitude of this variation will have major implications
for summary measures of population health.
More recently, Mahapatra [79] has carried out in India the first large scale population survey
which has adapted the GBD methods to obtain health state valuations from the general
population. The valuations from this survey indicate similar ranks for the states included in
the GBD studies, with somewhat higher disability weights overall. The most likely
explanation for the higher weights is the use of a visual analog scale to obtain the valuations,
which has produced higher disability weights than other methods in previous empirical
studies. Further work is underway on developing and refining instruments for data collection
in the general community, as well as on understanding how responses to different valuation
questions relate to strength of preference for different health states.
One of the main objectives for the ongoing WHO work on a standardized description of
health states for use in population surveys is to facilitate reliable and valid measurements of
valuations of time spent in health states in populations across the world. If large-scale
empirical assessment in many different countries to inform health state valuations for the
calculation of DALE for member states are to be achieved, instruments that are reliable and
valid for populations with widely varying educational attainment need to be developed.
14
2.2 Estimating DALE from health survey data
To date, few health expectancy calculations have been carried out based on health state
profiles that address more than one domain of health. Health-adjusted life expectancies
(HALE) have been calculated for Canada and Australia using population survey data on the
prevalence of disability at four levels of severity together with more or less arbitrary severity
weights [53–55]. More recently, Canada has produced the first estimates of health-adjusted
life expectancy based on population prevalence data for health states together with measured
utility weights [62]. This is the only example of a DALE which is based on a true multidomain health status instrument together with measured population preference weights (using
a standard gamble, non-deliberative approach).
There are two main problems with the use of self-report health status survey data to estimate
DALE:
·
the problems of comparability of self-reported health status or disability across
populations and across time.
· the estimation of disability weights for the corresponding health states.
The first of these problems is illustrated by Australian estimates of disability-free and
handicap-free life expectancy from 1981 to 1998 based on disability prevalence data from the
population surveys of disability (Figure 4). The prevalence of handicap increased substantially
between 1981 and 1988, from 9.4 to 13.7 per cent for males and from 8.7 to 12.2 per cent for
females. It is highly likely that a substantial part of these increases is due to changes in
community awareness and perceptions of handicaps, changes in income support programs,
availability of aids, and increasing levels of diagnosis of some health problems [61]. In
contrast, the prevalence of severe handicap remained largely unchanged over the period 1981
to 1993, then jumped substantially between 1993 and 1998. The latter change is thought to be
at least partly due to changes in survey methodology, although the actual questions used were
largely unchanged [80].
Figure 4: Trends in disability-free life expectancy (DFLE), handicap-free life expectancy
(HFLE),
severe handicap-free life expectancy (SHFLE) and total life expectancy (LE), by sex,
Australia 1981 to 1998
Males
80.00
Expectancy at birth (years)
Expectancy at birth (years)
80.00
LE
70.00
SHFLE
60.00
HFLE
Females
LE
SHFLE
70.00
HFLE
60.00
DFLE
DFLE
50.00
1980
50.00
1985
1990
1995
2000
1980
Year
1985
1990
Year
15
1995
2000
There are similar problems in cross-national comparability of self-report health status data.
During the last two decades, REVES and international agencies (including WHO, OECD and
Eurostat) have put considerable effort into promoting the development and use of
standardized health status and disability instruments in order to improve the cross-national
comparability of population health data. As a result of these efforts, there are now a number of
multi-country surveys that have used strictly comparable instruments and survey methods.
Analyses of these surveys have shown that substantial problems with comparability of selfreport health data remain [2]. This is illustrated in Figure 5, which shows results from the
second wave of the EC Health Panel Survey for 13 European countries [81]. This figure
shows the distribution of perceived health status (on a five point scale running from very good
to very bad) for people aged less than 60 who reported no chronic conditions and no
disability. Among this “healthy” group, there are very substantial variations in the prevalence
of both good and poor health states, which are unlikely to reflect real differences in population
health. Uncritical use of such self-report data would result in up to two-fold variations in the
average prevalence of (severity-weighted) disability across European countries.
Figure 5: Distribution of perceived health status among people aged 0-59 years who
have no chronic conditions, are not hampered in daily activities and have not cut down
activities due to health problems, EC Health Panel Survey, Wave 2, 1995
Austria
Portugal
Spain
Greece
Italy
Very good
Ireland
Good
Fair
UK
Bad
Very bad
France
Luxembourg
Belgium
Netherlands
Denmark
Germany
0
0.2
0.4
0.6
Frequency
16
0.8
1
Sadana et al [2] document in more detail how the problems in the comparability of self-report
health data relate not only to differences in survey design and methods, but much more
fundamentally to unmeasured differences in expectations and norms for health. Recent
analyses of surveys containing both self-report and objective measurements of health status
have documented systematic biases in self-report data according to age, sex, socioeconomic
disadvantage, and other measures of social disadvantage within populations [2].
There are related problems in estimating disability weights for health states measured in selfreport health surveys. Even where these surveys collect information on severity of disability,
severity is not generally measured in a form which can be easily translated into disability
weights reflecting health state preferences.
For the foreseeable future, this means that summary measures of population health for
comparative purposes must make use of survey results on self-reported health status
instruments with great care, and then only if supported by many other condition-specific
epidemiological datasets. In order to decompose summary measures into component causes
(diseases, injuries or risk factors), such condition-specific data sets will also be needed. This
is discussed further in the following section.
2.3 The disease-specific approach
Burden of disease analysis uses a disease-specific approach to estimate the disability and loss
of healthy years of life associated with a comprehensive and exhaustive set of health
conditions. In particular, DALYs are calculated as the sum of years of life lost due to
mortality (YLL) and equivalent years of healthy life “lost” due to disability (YLD). YLD for a
particular health condition (disease or injury) are calculated by estimating the number of new
cases (incidence) of the condition occuring in the time period of interest. For each new case,
the number of years of healthy life lost is obtained by multiplying the average duration of the
condition (to remission or death) by a severity weight that quantifies the equivalent loss of
healthy years of life due to living with the health condition [3].
Burden of disease analysis involves making YLD estimates for a comprehensive set of at least
100 disease and injury categories involving analysis of many more disease stages, severity
levels and sequelae. For some conditions, numbers of incident cases are available directly
from disease registers or epidemiological studies but for most conditions, only prevalence data
are available. In these cases, a software program called DISMOD© is used to model incidence
and duration from estimates of prevalence, remission, case fatality and background mortality
[3]. Many different sources of information are used to calculate YLD. An iterative process and
extensive consultation with relevant experts is required to ensure consistency of
epidemiological estimates.
Murray and Lopez [7] presented estimates of DFLE and DALE for each region of the world
using Sullivan’s method and the severity-weighted prevalence of disability derived from the
YLD estimates in the Global Burden of Disease Study. For these calculations, severityweighted disability estimates were not discounted or age weighted. Murray and Lopez
calculated disability prevalences for seven disability classes with an adjustment to allow for
independent co-disability between different disability classes. The expected years of healthy
life lost ranged from 8% in the Established Market Economies (life expectancy at birth of
around 77 years) to 15% in sub-Saharan Africa (life expectancy at birth of around 50 years).
The Australian Burden of Disease Study [46] also estimated DALE for Australia using
Sullivan’s method. Rather than estimate prevalence of seven disability classes, this study
estimated undiscounted prevalence YLD per 1,000 population as a direct measure of severity17
weighted disability prevalence and adjusted for comorbidity between disease and injury causal
groups rather than for co-disability. Total DALE at birth were 68.7 years for males and 73.6
years for females in Australia for 1996, similar to the values for the EME estimated in the
GBD. It was estimated that approximately 9% of total life expectancy at birth was lost due to
disability for both males and females in Australia, again similar to the 8% lost in the EME.
The disease-specific approach to the calculation of DALE used in the GBD and the Australian
studies has a number of advantages over the health survey approach:
·
it ensures consistency with the health gap measure (DALYs) of the burden of disease
·
it ensures inclusion of all causes of disability (including those resulting in forms of
disability poorly reported in health surveys eg. substance abuse, intellectual disability)
· it avoids problems of self-report biases.
However, there are currently two major limitations with this approach:
·
problems with comorbidities, and
· the data demands for calculating YLD for a comprehensive set of conditions.
Comorbidity refers to the not uncommon situation where a person has two or more health
problems that result in disability (either dependently or independently of each other). It makes
little sense to simply add the independently determined disability weights for conditions that
are found to coexist as this can lead to the illogical possibility of having a combined weight of
more than one (i.e. more disabling than death), particularly in the case of two heavily
weighted conditions. Both the GBD and the Australian studies made adjustments for
comorbidity assuming that conditions occurred independently (ie. the probability of having 2
conditions was the product of the average probabilities for having each condition) and
adjusted the disability weights for comorbid conditions assuming a multiplicative model [7,
46].
Further work is needed to determine whether such simple comorbidity models are adequate
for characterising the burden of disease and the distribution of disability by severity in a
population. Substantial effort will be required to improve on the estimation of the prevalence
of non-independent comorbidity for future iterations of the GBD.
Mathers et al [46] have compared disability and handicap prevalence data derived from a
national population survey in Australia with weighted disability prevalences derived from the
estimates of total YLD in the Australian Burden of Disease project. The total prevalent YLD
per 100 population can be thought of as a severity-weighted disability prevalence measured as
a percentage of the population of that age. The disability survey data were used to estimate
weighted disability prevalence rates (%) by age and sex for the Australian population in 1998.
Weights for six disability and handicap severity levels were chosen to line up as closely as
possible with appropriate preference weight ranges for the Dutch disability weights defined in
terms of EuroQol health state descriptions. Results for males and females combined are
shown in Figure 6 and compared with the prevalence YLD from the Australian Burden of
Disease study. YLD associated with short-term conditions lasting less than six months (such
as colds and flu) have been excluded, since the survey definition of disability included only
chronic disability lasting six months or more. YLD associated with anxiety disorders and mild
to moderate (but not severe) depression have also been excluded, since the majority of
disability associated with these conditions is unlikely to have been captured by the ABS
Disability Survey.
The YLD-based prevalence estimates correspond quite closely to the survey-based prevalence
estimate at younger and middle ages and at ages 75 and over. For ages in the range 55–74
18
years, the YLD-based prevalence is significantly higher than the survey-based prevalence.
This may reflect the impact of chronic diseases prevalent at these ages that are not being
picked up by the Disability Survey screening questions. The Australian Burden of Disease
Study made some attempts to ensure consistency between the overall population prevalence
derived from survey data for some particular impairments (such as intellectual impairment,
renal failure, amputations) and for the total prevalence of these sequelae added across relevant
conditions for which YLD were estimated in the burden of disease study.
In the following Section, we describe a combined approach which uses available population
health survey data together with the disease-specific approach to develop consistent estimates
of the prevalence of disability for the calculation of DALE.
Figure 6: Comparison of severity-weighted prevalence of disability from
1998 ABS Disability Survey with prevalence YLD (per cent), by age, 1996
Severity-weighted prevalence (%)
30
1998 ABS Disability Survey
25
YLD for chronic conditions
20
15
10
5
0
0
20
40
Age (years)
19
60
80
3. Methods
3.1 Overview of approach
As discussed above, ideally we would use survey data for disability caused by specific
conditions to estimate YLD for those conditions, in a way that ensured consistency between
epidemiological estimates of disease incidence, prevalence, case fatality, progression to
disabling sequelae, and severity distributions of resulting health states. Because of the severe
problems found with the comparability of self-reported health status data in population
surveys [2], we have developed an analytical approach for this first analysis of DALE for all
WHO member countries, which combines the condition-specific approach based in burden of
disease analysis with the use of available representative population health survey data on the
population distribution of health states.
In brief, it involves the following steps:
1. The development for each country of age-sex specific weighted disability prevalence
estimates based on burden of disease analyses at country level which build on conditionspecific epidemiological information to the maximum extent possible.
2. The construction of latent health factor scores from representative population health
surveys.
3. The estimation of weighted disability prevalence from these latent health factor scores
using the disability estimates from step 1 as prior estimates. The rescaling of the factor
scores to improve comparability of survey data and adjust for self-report biases is based on
estimating a parsimonious set of self-report bias parameters which provide best fit
between factor scores and prior disability estimates .
4. The use of Sullivan’s method to calculate DALE from posterior disability estimates plus
country life tables. Where health survey data is not available, DALE are calculated using
the prior disability estimates from the country-level epidemiological analyses.
The advantages and limitations of this approach are discussed further in Section 5, which also
outlines the approach which will be taken over the next year to improve the estimation of
DALE for WHO member countries.
3.2 Life tables and cause of death distributions for countries
As a first step towards the estimation of DALE for WHO Member States, it is crucial to
develop for each country the best possible assessment for 1999 of overall mortality levels by
age and sex and the corresponding life table. New life tables were developed for all 191 WHO
Member States starting with a systematic review of all available evidence from surveys,
censuses, sample registration systems, population laboratories and vital registration on levels
and trends of child and adult mortality [82].
To aid in the cause of death and burden of disease analysis used as inputs to the DALE
estimation process, the six WHO regions of the world were divided into 5 mortality strata on
the basis of their level of child (5q0) and adult male mortality (45q15). The matrix defined by
the six WHO Regions and the 5 mortality strata leads to 14 subregions, since not every
mortality stratum is represented in every Region. These subregions are defined in Table 3.
20
Table 3. WHO Regions and mortality subregions used in epidemiological analyses and reporting for the
World Health Report 2000.
Region
Mortality subregion
Population
(millions)
WHO Member States
AFRO
D
High child
High adult
Algeria, Angola, Benin, Burkina Faso, Cameroon, Cape Verde, Chad, Comoros,
Equatorial Guinea, Gabon, Gambia, Ghana, Guinea, Guinea-Bissau, Liberia,
Madagascar, Mali, Mauritania, Mauritius, Niger, Nigeria, Sao Tome and
Principe, Senegal, Seychelles, Sierra Leone, Togo
286
AFRO
E
High child
Very high adult
Botswana, Burundi, Central African Republic, Congo, Côte d'Ivoire, Democratic
Republic of the Congo, Eritrea, Ethiopia, Kenya, Lesotho, Malawi, Mozambique,
Namibia, Rwanda, South Africa, Swaziland, Uganda, United Republic of
Tanzania, Zambia, Zimbabwe
330
AMRO
A
Very low child
Low adult
Canada, Cuba, United States of America
318
AMRO
B
Low child
Low adult
Antigua and Barbuda, Argentina, Bahamas, Barbados, Belize, Brazil, Chile,
Colombia, Costa Rica, Dominica, Dominican Republic, El Salvador, Grenada,
Guyana, Honduras, Jamaica, Mexico, Panama, Paraguay, Saint Kitts and
Nevis, Saint Lucia, Saint Vincent and the Grenadines, Suriname, Trinidad and
Tobago, Uruguay, Venezuela (Bolivarian Republic of)
425
AMRO
D
High child
High adult
Bolivia, Ecuador, Guatemala, Haiti, Nicaragua, Peru
EMRO
B
Low child
Low adult
Bahrain, Cyprus, Iran (Islamic Republic of), Jordan, Kuwait, Lebanon, Libyan
Arab Jamahiriya, Oman, Qatar, Saudi Arabia, Syrian Arab Republic, Tunisia,
United Arab Emirates
137
EMRO
D
High child
High adult
Afghanistan, Azerbaijan, Djibouti, Egypt, Iraq, Morocco, Pakistan, Somalia,
Sudan, Yemen
348
EURO
A
Very low child
Low adult
Andorra, Austria, Belgium, Croatia, Czech Republic, Denmark, Finland, France,
Germany, Greece, Iceland, Ireland, Israel, Italy, Luxembourg, Malta, Monaco,
Netherlands, Norway, Portugal, San Marino, Slovenia, Spain, Sweden,
Switzerland, United Kingdom
410
EURO
B
Low child
Low adult
Albania, Armenia, Bosnia and Herzegovina, Bulgaria, Georgia, Kyrgyzstan,
Poland, Romania, Slovakia, Tajikistan, The Former Yugoslav Republic of
Macedonia, Turkey, Turkmenistan, Uzbekistan, Yugoslavia
215
EURO
C
Low child
High adult
Belarus, Estonia, Hungary, Kazakhstan, Latvia, Lithuania, Republic of Moldova,
Russian Federation, Ukraine
246
SEARO
B
Low child
Low adult
Indonesia, Sri Lanka, Thailand
289
SEARO
D
High child
High adult
Bangladesh, Bhutan, Democratic People's Republic of Korea, India, Maldives,
Myanmar, Nepal
WPRO
A
Very low child
Low adult
Australia, Brunei Darussalam, Japan, New Zealand, Singapore
WPRO
B
Low child
Low adult
Cambodia, China, Cook Islands, Fiji, Kiribati, Lao People's Democratic
Republic, Malaysia, Marshall Islands, Micronesia (Federated States of),
Mongolia, Nauru, Niue, Palau, Papua New Guinea, Philippines, Republic of
Korea, Samoa, Solomon Islands, Tonga, Tuvalu, Vanuatu, Viet Nam
World
70
1,219
153
1,521
5,968
Because of increasing heterogeneity of patterns of adult and child mortality, WHO has
developed a system of two-parameter logit life tables for each of the 14 mortality subregions
used in the World Health Report [83]. This system of model life tables was used extensively
in the development of life tables for each Member State and in projecting life tables to 1999
when the most recent data available were from earlier years. Details on the data and methods
used for each country are given by Lopez et al [82]. In countries with a substantial HIV
21
epidemic, separate estimates were made of the numbers and distributions of deaths due to
HIV/AIDS and these deaths incorporated into the life table estimates [84].
The detailed distribution of causes of death by age and sex were estimated for each WHO
Member State based on data from national vital registration systems that capture 16.7 million
deaths annually. In addition, information from sample registration systems, population
laboratories and epidemiological analyses of specific conditions were used to produce better
estimates of the cause of death patterns. Cause of death patterns were carefully analysed to
take into account incomplete coverage of vital registration in countries and the likely
differences in cause of death patterns that would be expected in the low coverage areas of
countries with incomplete data. Techniques developed in the Global Burden of Disease Study
to undertake this analysis were further developed using a more extensive database and more
robust modelling techniques [85].
Special attention was paid to problems of misattribution or miscoding of causes of death in
cardiovascular disease, cancer injuries and general ill-defined categories. Deaths coded to illdefined cardiovascular categories were reclassified using a correction algorithm described by
Lozano et al [86]. A complete age-period-cohort model of cancer survival in each region was
used to identify cancer sites with significant undercoding of mortality in order to reclassify
cancer deaths coded to ill-defined categories [87].
3.3 Prior estimates for countries
The disease-specific approach described in Section 2.3 has been used to develop the best
possible initial (prior) estimates of weighted disability prevalence by age and sex for all 191
WHO member countries. These estimates are based on preliminary burden of disease analyses
at country level which build on condition-specific epidemiological information to the
maximum extent possible. This Section describes in more detail how they were developed.
Step 1. As part of its annual assessment of world health in the World Health Report, WHO is
updating and revising its estimates of disease burden for the 14 mortality subregions of the
world. This involves carrying out detailed and comprehensive reviews of the incidence,
prevalence, duration, and case fatality in all the regions of the world for each of 109 major
disease and injury causes of mortality and disability by age group and sex (see Table 4).
The ongoing revisions to the Global Burden of Disease analysis at WHO draw on a wide
range of data sources. Various methods have been developed to reconcile often fragmented
and partial estimates of epidemiological parameters that are available from different studies.
A specific software tool, DisMod, is developed to ensure that the results of these assessments
are internally consistent, and in particular, are consistent with cause of death distributions [7].
For a review of the development of DALYs and recent advances in disease burden
measurement, see Murray and Lopez [7, 49].
Annex Table 4 of the World Health Report (2000) summarises the disease burden estimates
for the 14 mortality subregions of the world in 1999. The ten leading causes of DALYs for
each of these regions are also shown in this report (Table 9).
22
Table 4. Disease and injury categories used for regional burden of disease analyses for the World Health
Report 2000.
I. Communicable, Maternal & Perinatal
A. Infectious and Parasitic
1. Tuberculosis
2. STD's excluding HIV
a. Syphilis
b. Chlamydia
c. Gonorrhea
d. Other STD's
3. HIV
4. Diarrhoeal Diseases
5. Childhood Cluster
a. Pertussis
b. Polio
c. Gonorrhea
d. Measles
e. Tetanus
6. Meningitis
7. Hepatitis
8. Malaria
9. Tropical Cluster
a. Trypanosomiasis
b. Chagas' Disease
c. Schistosomiasis
d. Leishmaniasis
e. Lymphatic Filariasis
f. Onchocerciasis
10. Leprosy
11. Dengue
12. Japanese Encephaliti
13. Trachoma
14. Intestinal Nematodes
a. Ascaris
b. Trichuris
c. Hookworm
15. Other Infectious
B. Respiratory Infections
1. ALRI
2. AURI
3. Otitis Media
C. Maternal Conditions
1. Hemorrhage
2. Sepsis
3. Hypertensive disorders of pregnancy
5. Obstructed Labor
6. Abortion
7. Other Maternal
D. Perinatal Conditions
E. Nutritional
1. Protein-Energy malnutrition
2. Iodine Deficiency
3. Vitamin A Deficiency
4. Anemia
5. Other Nutritional
II. Noncommunicable
A. Malignant Neoplasms
1. Mouth and Oropharynx
2. Esophagus
3. Stomach
4. Colon/Rectum
5. Liver
6. Pancreas
7. Trachea/Bronchus/Lung
8. Melanoma and other Skin
9. Breast
10. Cervix
11. Corpus Uteri
12. Ovary
13. Prostate
II. Noncommunicable (continued)
A. Malignant Neoplasms (continued)
14. Bladder
15. Lymphoma
16. Leukemia
17. Other Cancers
B. Other Neoplasms
C. Diabetes Mellitus
D. Nutritional/Endocrine
E. Neuro-psychiatric
1. Major Affective Disorder
2. Bipolar Affective Disorder
3. Psychoses
4. Epilepsy
5. Alcohol Dependence
6. Alzheimer's and other de
7. Parkinson's Disease
8. Multiple Sclerosis
9. Drug Dependence
10. PTSD
11. Obsessive Compulsive
12. Panic disorder
11. Other Neuro-psychiatric
F. Sense Organ
1. Glaucoma
2. Cataracts
3. Other Sense Organ
G. Cardiovascular
1. Rheumatic Heart Disease
2. Ischemic Heart Disease
3. Cerebrovascular Disease
4. Inflammatory Cardiac
5. Other
H. Respiratory
1. COPD
2. Asthma
3. Other Respiratory
I. Digestive
1. Peptic Ulcer Disease
2. Cirrhosis of the Liver
3. Appendicitis
3. Other Digestive
J. Genito-Urinary
1. Nephritis/Nephrosis
2. Benign Prostatic Hypertension
3. Other Genito-Urinary
K. Skin Disease
L. Musculo-Skeletal
1. Rheumatoid Arthritis
2. Osteoarthritis
3. Other Musculo-Skeletal
M. Congential Abnormalities
N. Oral Health
1. Dental Caries
2. Periodontal Disease
3. Edentulism
4. Other Oral Health
III. Injuries
A. Unintentional
1. Motor Vehicle Accidents
2. Poisoning
3. Falls
4. Fires
5. Drowning
6. Other Unintentional Injuries
B. Intentional
1. Self-inflicted
2. Homicide and Violence
3. War
23
Step 2. As described in Section 3.2, WHO has prepared estimates of numbers of deaths for
each of its 191 Member States according to sex, age group (0, 1-4, then 5-year age groups to
85+) and 130 disease and injury causes (covering all causes of disease and injury). These
estimates were used to calculate YLL by sex, age group and detailed cause for each Member
State.
Step 3. This country-level mortality data (Step 2), some country level epidemiological data
and regional burden of disease estimates (Step 1) were then used to develop country-level
estimates for YLD and total DALYs by sex, 5 year age group, and detailed cause as follows.
For specific disease and injury causes where mortality is responsible for a significant
proportion of the total burden (YLD/YLL ratio less than 5), regional estimates of YLD/YLL
ratios by age and sex together with country-level estimates of YLL were used to estimate
country-level YLD. This process ensures that country-specific knowledge on the
epidemiology of the disease (as reflected in the country-level mortality estimates of that
disease) is used to adjust the regional-level patterns of disability due to that cause.
For specific disease and injury causes where mortality is not responsible for a significant
proportion of the total burden (YLD/YLL ratio is 5 or higher), regional estimates of YLD
rates per 1,000 population by age and sex were used together with country-level population
distribution estimates and estimates of health expenditure per capita to make first estimates of
the resulting YLD for each country. For some diseases, notably cancers, major depression and
chronic respiratory conditions, available country-specific epidemiological estimates were also
examined.
In order to estimate disability prevalence at population level, it is also necessary to estimate
the YLD associated with residual categories of disease and injury such as ‘Other chronic
respiratory diseases’ or ‘Other malignant neoplasms’. We followed the procedure developed
by the Global Burden of Disease Study [7, page 211] to estimate YLD for all of these residual
categories.
Step 4. For each member country, we then used the incidence YLD by age, sex and detailed
cause (Step 3) to estimate undiscounted and un-age-weighted prevalence YLD by 5 year age
group, sex and detailed cause. The method for conversion of incidence YLD to prevalence
YLD used was dependent on the average duration of condition as follows:
Short duration (<5 years): Prevalent YLD are equal to incident YLD
Moderate duration (5 years to 50% of remaining life expectancy): We assume incident YLD
are evenly distributed across the age interval a to a+L, where a is average age
of onset and L is average duration.
Long duration (50% or more of remaining life expectancy): We construct a life table for years
lived with condition using the country life table and proportionately
increasing mortality rates at all ages to match remaining life expectancy to the
average duration of condition. We then use the Lx (years lived) column of the
resulting life table to distribute incident YLD across age groups.
Step 5. Adjustment for comorbidity. As discussed in Section 2.3, the total prevalent YLD per
100 population can be thought of as a severity-weighted disability prevalence measured as a
percentage of the population of that age. However, summation over all conditions of the
prevalence YLD calculated in Step 4 would result in overestimation of disability prevalence
because of comorbidity between conditions. We correct for independent comorbidity between
major condition groups (these approximately correspond to the Chapters of the International
Classification of Diseases) as follows:
24
PYLDs, x = 1 - Õ (1 - PYLDs, x, g )
(7)
g
where PYLDs,x,g is the prevalence YLD per 100 population for sex s, age x and cause g.
The resulting PYLD per 100 population for sex s, age x gives the severity-weighted
prevalence of disability by age and sex.
3.4 Factor analysis of health surveys
Population representative data obtained through national sample surveys which include an
assessment of general health status and physical and cognitive disability, have been collected
and critically reviewed in an accompanying Discussion Paper [2]. Table 5 lists these surveys
and summarises some of their characteristics. In order to estimate the prevalence of disability
(non-fatal health) by five year age groups and sex at the country level, Sadana et al. [2] outline
an analytic approach which addresses some of the methodological challenges regarding the
comparability of the health status data collected. These include differences in the range and
depth of questions and response scales [88] as well as differences in the interpretation of
responses given different norms and expectations for health by age, sex or other subpopulation groups. After conducting several validity and reliability checks, the analysis
confirmed a latent dimension of disability that is common across population survey data and
estimated the level of disability for each country and sex separately.
As shown in Figure 7, the cumulative distribution of disability prevalence by severity is
approximately exponential according to the detailed analyses carried out for the Global
Burden of Disease study [7]. The distributions of latent health factor scores derived from the
analysis of country health surveys were also generally exponential (see Figure 7 right-hand
graph). The distribution of disability by severity level (or disability weight) can thus be
approximately described by a two parameter exponential distribution as follows:
d ( x) =
a
b
e
-
x
b
(8)
where x is the disability weight (severity) measured on a scale where 1 represents good health
and 0 represents a state equivalent to death. The mean of this distribution is:
d =a ´b
(9)
The parameter a is readily interpreted as the proportion of the population with disability
(with non-zero disability weight) and b as the average disability weight among the people
with disability.
Figure 8 compares for selected countries the mean values by age and sex for the latent health
factor scores (expressed as a per cent on a scale where 0 represents good health and 100% the
worst possible health state) with the prior mean values of severity-weighted disability from
the condition-specific analyses. There are some countries where the latent health factor scores
correspond reasonably well with the prior disability estimates (eg. Both sexes in Ireland,
males in the USA). However, for many countries there are substantial differences in the latent
health factor scores for males and females (eg. Portugal, USA) or in age trends (eg. Brazil).
For many developing countries, very few people report disability in population surveys,
resulting in quite low mean values for the latent health factor scores compared to the prior
disability estimates. The graph for India in Figure 8 illustrates this situation.
25
Figure 7: Left: cumulative prevalence of disability by severity weight. Left: from GBD 1990. Right:
for males in four European countries (using disability weight based on latent health factor scores
unadjusted)
5-14
1.000
0.200
0.300
0.400
0.500
0.600
0.700
0.800
0.900
100
1.000
Germany
UK
Portugal
0.100
Cumulative prevalence
60+
10
45-59
0.010
Cumulative prevalence (%)
Italy
15-44
0-4
1
5-14
80
0.001
60
40
20
0
Disability w eight
1 – Disability weight
As documented in detail by Sadana et al [2], there are country-specific biases in self-reported
health data by sex and age that confound cross-national comparisons of health status or
disability. Yet these data do appear to contain information on the health of the population. The
task is to improve the comparison of health status across surveys and countries. Approaches
that have attempted to map similar questions from surveys across countries have not been
successful in achieving this aim, due to the substantial problems with self-report differences
between cultures [2].
In order to obtain strict comparability of health status across surveys, the latent health factor
derived from different surveys should map similar health states in different populations to the
same score. Even if there were no self-report biases of the type documented above, there is
still a question of whether the factor analyses of different surveys would result in a common
underlying factor which measures health states on the same ordinal scale. For example, it is
possible that surveys which contain questions that discriminate mild states of less than ideal
health better may spread out such states more on the latent factor scale of 0 to 1 than other
surveys.
It is clear that for the estimation of DALE on an internationally comparable basis, if we are to
use self-reported health survey information on population health status, we need to adjust the
self-report data for differences in survey content and self-reporting biases in a way that will
produce comparable estimates of severity-weighted disability prevalence.
26
Table 5. Sixty-four household interview surveys from 48 countries: survey characteristics by region
Region
Country
Year
Age
Range
1987
0+
Sample Size
M
F
Type of Survey
AFRO D
Ghana
6833
7162
Integrated Household Survey (LSMS type)
AFRO E
Côte d'Ivoire
1988
0+
4933
5174
Integrated Household Survey (LSMS type)
South Africa
1994
0+
21239
22703
Integrated Household Survey (LSMS type)
United Republic of Tanzania
1995
14+
2210
2068
Sub-national health survey
United States of America *
88-94
17+
9401
10649
National health survey (NHANES III)
United States of America
1994
18+
35914
41523
National health survey (NHIS-D Phase I)
AMRO A
AMRO B
Brazil
96-97
4+
9410
9999
Integrated Household Survey (LSMS type)
Guyana
1996
0+
3869
3924
Integrated Household Survey (LSMS type)
Jamaica
1996
0+
3422
3510
Integrated Household Survey (LSMS type)
Panama
1997
0+
10838
10599
Integrated Household Survey (LSMS type)
Paraguay
1996
4+
5543
5925
Integrated Household Survey (LSMS type)
AMRO D
Peru
1991
0+
5795
6051
Integrated Household Survey (LSMS type)
Peru *
1994
0+
9402
9883
Integrated Household Survey (LSMS type)
Bahrain
1991
60+
448
351
WHO Collaborating Study on Ageing
Jordan
1991
60+
545
652
WHO Collaborating Study on Ageing
Tunisia
1991
60+
658
578
WHO Collaborating Study on Ageing
EMRO B
EMRO D
Egypt
1991
60+
710
470
WHO Collaborating Study on Ageing
Morocco
90-91
0+
9444
10128
Integrated Household Survey (LSMS type)
Pakistan *
90-94
0+
10039
9792
National health survey (NHSP)
Pakistan
1991
0+
18731
17340
Integrated Household Survey (LSMS type)
Bulgaria
1995
0+
3350
3577
Integrated Household Survey (LSMS type)
Kyrgyzstan
1993
14+
2617
3030
Integrated Household Survey (LSMS type)
1998
14+
3750
4916
Longitudinal Integrated Household Survey
Austria
1995
15+
3567
3874
Longitudinal Integrated Household Survey
(European Community)
Belgium
1994
15+
3872
4249
1995*
15+
3666
4066
Longitudinal Integrated Household Survey
(European Community)
EURO B
EURO C
Russian Federation
EURO A
Denmark
1994a
15+
2855
3048
1995*
15+
2680
2824
Longitudinal Integrated Household Survey
(European Community)
Denmark
1994b
15+
2699
2913
National health survey (DHMS: SF-36)
France
1994
15+
6839
7494
1995*
15+
6368
6936
Longitudinal Integrated Household Survey
(European Community)
1994
15+
4150
4366
1995*
15+
3885
4073
Germany
Longitudinal Integrated Household Survey
(European Community)
(continued)
27
Table 5 (continued). Sixty-four household interview surveys from 48 countries: survey characteristics
by region
Region
Country
Year
Age
Range
Sample Size
M
F
Greece
1994
15+
1995*
15+
5878
6396
Ireland
1994
15+
4922
4982
1995*
15+
4263
4268
Italy
1994
15+
8660
9071
1995*
15+
8704
9079
Luxembourg
1994
15+
990
1056
1995*
15+
957
1011
Netherlands
1994
15+
4457
4950
1995*
15+
4299
4852
Portugal
1994
15+
5556
6065
1995*
15+
5691
6167
1994
15+
8625
9285
Type of Survey
EURO A
Spain
United Kingdom
5904
6589
1995*
15+
7837
8443
1994
15+
4986
5531
1995*
15+
3995
4396
Longitudinal Integrated Household Survey
(European Community)
Longitudinal Integrated Household Survey
(European Community)
Longitudinal Integrated Household Survey
(European Community)
Longitudinal Integrated Household Survey
(European Community)
Longitudinal Integrated Household Survey
(European Community)
Longitudinal Integrated Household Survey
(European Community)
Longitudinal Integrated Household Survey
(European Community)
Longitudinal Integrated Household Survey
(European Community)
SEARO B
Indonesia
1995
0+
Total 9901
National health survey (SKRT)
Indonesia *
93-94
0+
5509
5103
Longitudinal Integrated Household Survey
Indonesia
1990
60+
568
634
WHO Collaborating Study on Ageing
Sri Lanka
1990
60+
638
562
WHO Collaborating Study on Ageing
Thailand
1990
60+
598
601
WHO Collaborating Study on Ageing
1996
15+
5266
6311
Sub-national Integrated Household Survey
SEARO D
Bangladesh
DPR of Korea
1990
60+
585
596
WHO Collaborating Study on Ageing
India
95-96
0+
32353
6
30561
2
National Integrated Household Survey
(Round 57)
Myanmar
1990
60+
511
710
WHO Collaborating Study on Ageing
Nepal
1994
0+
9263
9592
Integrated Household Survey (LSMS type)
WPRO B
China
1993
0+
7836
7846
Longitudinal Integrated Household Survey
Fiji
1986
60+
360
321
WHO Collaborating Study on Ageing
Malaysia
1986
60+
389
589
WHO Collaborating Study on Ageing
Philippines
1986
60+
326
491
WHO Collaborating Study on Ageing
Republic of Korea
1986
60+
348
593
WHO Collaborating Study on Ageing
*If more than one survey from a country, survey with * is used as input to DALE estimations.
28
Figure 8. Comparison of latent health factor scores and prior disability prevalences from conditionspecific analyses, selected countries, 1999
Males: Latent health factor score
Males prior disability estimates
Females: Latent health factor score
Females prior disability estimates
Ireland
40
30
30
Mean value (%)
Mean value (%)
United Kingdom
40
20
10
0
20
10
0
15
25
35
45
55
Age group
65
75
15
25
35
40
40
30
30
Mean value (%)
Mean value (%)
65
75
65
75
65
75
Brazil
Portugal
20
10
0
20
10
0
15
25
35
45
55
Age group
65
75
15
25
35
45
55
Age group
India
USA
40
40
30
30
Mean value (%)
Mean value (%)
45
55
Age group
20
10
20
10
0
0
15
25
35
45
55
Age group
65
15
75
29
25
35
45
55
Age group
Figure 9. Estimation of true health status from self-reported health status in representative national
population surveys
Survey item 1
Country/culture
Sex
Perceived/
reported
health
status
Age
Survey item 2
Survey item 3
.
.
.
Socioeconomic status
True
health
status
Survey item n
Mortality
Healt determinants
Health/welfare
system
This estimation problem is illustrated in Figure 9. Determinants of “bias” in self-report health
data such as age, sex, socioeconomic status and other population-specific
(cultural/environmental) factors are also determinants of true health status. We made a
number of attempts to develop models to estimate true health status from observed health
status by making use of the self-reported health status of reference groups (such as young,
high socioeconomic status people) who could be assumed to have good true health status.
However, we found that the patterns of bias between social groups and across countries were
so heterogeneous that such “bootstrapping” calibration procedures using internal health
patterns within survey populations were not able to produce reliably comparable data across
surveys. We thus concluded that the inter-country biases in self-report health survey data
could only be adequately adjusted if external calibrators (of true health or of health
determinants) were used to make appropriate adjustments to self-report data.
3.5 Posterior estimates
In order to maximise comparability of the disability dimensions and to ensure that the
resulting scores appropriately reflected health state preferences, we have developed numerical
models for estimating and adjusting for age, sex and cross-country bias in reporting of health
states. These use prior estimates of disability distributions based on burden of disease analyses
at regional and country level (Section 3.3) in order to estimate a parsimonious set of
parameters to adjust for differences in survey content, in self-report responses and for the
mapping of the resulting latent health factor to health state preference weights. This procedure
30
enables us to incorporate all data from household surveys into our estimates of average levels
of severity-weighted disability for countries.
We briefly describe the model and the estimation procedure here. If f ¢ denotes the latent
health factor score derived from health survey data using confirmatory factor analysis (Sadana
et al 2000) (2), and f denotes the true underlying health state (in terms of a disability weight
ranging from 0 to 1), then we describe the relationship between these two factors in terms of
three parameters as follows (see also Figure 10):
æ f ¢ö
f1 + (1 - f1 )´ çç ÷÷
è f2 ø
b
if
f ¢ £ f2
if
f ¢ > f2
f =
(10)
1
Figure 10. Adjustment of latent health factor score for reporting biases and disability severity
100
b>1
80
True disability weight
True disability weight
100
60
40
f1
20
0
0
50
f2
b<1
80
60
40
f1
20
0
100
0
Latent health factor score
50
f2
100
Latent health factor score
The first parameter f1 specifies the disability weight for the worst health state observed in the
health survey, the second parameter f 2 specifies the latent health factor score above which all
health states are equivalent to good health f = 1 (see Figure 10). This latter parameter allows
for health surveys which include questions which discriminate between people in good health,
thus resulting in a range of factor scores which all equate to good health. This ceiling effect is
discussed in more detail by Sadana et al (2000) (2). The third parameter, b , specifies a power
transformation which allows the latent health factor score to decrease with increasing severity
faster or slower than the true disability weight as illustrated in Figure 10.
We used this model to rescale latent health factor scores for all the available health surveys in
a region as follows. Prior values for average disability severity were derived from the
condition-specific country-level analyses (described in Section 3.3) by sex for age groups 0-4,
5-14, 15-24, 25-34, 35-44, 45-54, 55-64, 65-74, 75+. Nonlinear optimization methods were
then used to estimate values for parameters f 2 and b which minimized the difference
between mean disability weight and the mean value (across all age groups combined) for the
rescaled factor score f . The parameter f1 was set at 0.1 in all models. Sensitivity analyses
31
found that the value assumed for this parameter in a range from around 0.0 to 0.2 did not
significantly affect the mean values for the rescaled factor scores.
We aimed to estimate common values f 2 and b for all countries in a region to the extent
possible, ie. to simultaneously fit equation (10) to data for a number of countries. We found
that this was rarely possible due to the heterogeneity of the self-report biases across countries,
and in some instances, we also needed to allow f 2 to differ for ages 65 and over, to allow for
age-related variations in reporting behaviour. Health survey data for a number of countries
was restricted to ages 60 and above (see Table 5). For these countries, the model parameters
were estimated by fitting the model to prior disability distributions for four 5-year age groups
(60-64, 65-69, 70-74, 75+).
Figure 11 compares the average values of the rescaled health factor (the posterior disability
prevalence) with the latent health factor scores by age and sex for selected countries. In some
countries, the posterior estimates are quite similar to the latent health factor scores (eg.
Ireland, Greece, males in the USA). However, for many countries the fitted bias parameters
result in quite substantial changes in the posterior estimates (eg. United Kingdom, Brazil,
females in Portugal).
Figure 12 compares the posterior (adjusted) disability prevalence estimates for seven age
groups (15-24, through to 75+) to the prior disability prevalence estimates (from the
condition-specific analyses) for 13 European countries. For these countries, the latent health
factor scores were derived from a single common health survey using the same instrument in
all 13 countries (see Table 5). Figure 13 similarly compares posterior and prior disability
prevalence estimates for nine age groups in 4 African countries.
Figure 14 illustrates the differences between the posterior disability prevalence estimates and
the (unadjusted) latent health factor scores derived from the initial factor analyses of the
country health surveys. The first 6 graphs are for selected countries in Europe and the
Americas. The health survey data results in moderate changes in the level and age trends of
average weighted disability prevalence for the European countries. For the USA, prior and
posterior estimates are quite close, apart from for the oldest male age group. For Brazil, the
use of the health survey data results in reductions in the prior disability estimates across all
ages for both sexes. The second six graphs (Figure 14 continued) are for selected countries in
Africa, Asia and the Western Pacific region. The health survey data resulted in increases
above estimated prior levels of disability at younger ages in both Ghana and Côte d’Ivoire.
The analysis of health survey data resulted in a reduction in estimated prior levels of disability
at older ages in the Republic of Korea (South Korea) and increases in the estimated prior
levels of disability at older ages in the Democratic People’s Republic of Korea (North Korea).
Health survey data were only available for ages 60 and over for Thailand and Myanmar.
Analysis of this data resulted in increases in prior estimates of disability for the older old (ages
75 and over) in both countries.
Using the methods outlined above, we obtained estimates of weighted disability prevalence by
sex and age for all WHO Member States. For countries where we were not able to obtain unit
record data for appropriate household surveys, we used the estimated prevalence of disability
by five year age groups and sex calculated by age, sex and cause at country level based on
regional estimates of the burden of disease for 1999 together with country-specific
epidemiological information.
Figure 15 shows the resulting distributions of weighted disability prevalence by age and sex
for the 14 WHO mortality subregions in 1999. These prevalences have been calculated by
32
summing disability prevalences weighted by population numbers across Member States in
each subregion.
33
Figure 11. Posterior (adjusted health survey) disability prevalence versus latent health factor score,
selected countries, 1999
Males: Latent health factor score
Males posterior estimates
Females: Latent health factor score
Females posterior estimates
Ireland
40
30
30
Average (%)
Average (%)
United Kingdom
40
20
10
20
10
0
0
15
25
35
45
55
Age group
65
75
15
25
40
30
30
20
10
65
75
65
75
65
75
20
10
0
0
15
25
35
45
55
Age group
65
75
15
25
USA
35
45
55
Age group
Brazil
35
35
30
30
25
25
Average (%)
Average (%)
45
55
Age group
Portugal
40
Average (%)
Average (%)
Greece
35
20
15
10
5
20
15
10
5
0
0
15
25
35
45
55
Age groups
65
75
15
34
25
35
45
55
Age groups
Figure 12. Comparison of posterior (adjusted) disability prevalence versus prior
disability prevalence for seven age groups by sex, 13 European countries, 1995
35
Modelled disability (%)
30
25
20
15
10
5
0
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
Prior disability (%)
Figure 13. Comparison of posterior (adjusted) disability prevalence versus prior
disability prevalence, for nine age groups by sex, 4 African countries countries
70
Modelled disability (%)
60
50
40
30
20
10
0
0.0
10.0
20.0
30.0
Prior disability (%)
35
40.0
50.0
60.0
Figure 14. Posterior (adjusted health survey) disability prevalence versus prior disability
prevalence,
selected countries, 1999
Males: Prior disability estimates
Males posterior estimates
Females: Prior disability estimates
Females posterior estimates
Netherlands
35
30
30
Wgt. disability prev. (%)
Wgt. disability prev. (%)
Germany
35
25
20
15
10
5
0
25
20
15
10
5
0
15
25
35
45
55
Age groups
65
75
15
25
35
30
30
25
20
15
10
5
0
75
65
75
65
75
25
20
15
10
5
0
15
25
35
45
55
Age groups
65
75
15
25
USA
35
45
55
Age groups
Brazil
35
35
30
30
Wgt. disability prev. (%)
Wgt. disability prev. (%)
65
Italy
35
Wgt. disability prev. (%)
Wgt. disability prev. (%)
United Kingdom
35
45
55
Age groups
25
20
15
10
5
0
25
20
15
10
5
0
15
25
35
45
55
Age groups
65
75
15
36
25
35
45
55
Age groups
Figure 14 (continued). Posterior (adjusted health survey) disability prevalence versus prior
disability prevalence, selected countries, 1999
Males: Prior disability estimates
Males posterior estimates
Females: Prior disability estimates
Females posterior estimates
Côte d'Ivoire
40
35
35
Wgt. disability prev. (%)
Wgt. disability prev. (%)
Ghana
40
30
25
20
15
10
5
0
30
25
20
15
10
5
0
0
5
15
25 35 45
Age groups
55
65
0
Republic of Korea
15
55
65
Wgt. disability prev. (%)
50
40
30
20
10
0
40
30
20
10
0
35
40
45
50 55 60
Age groups
65
70
75
35
40
45
Thailand
50 55 60
Age groups
65
70
75
65
70
75
Myanm ar
50
Wgt. disability prev. (%)
50
Wgt. disability prev. (%)
25 35 45
Age groups
Democratic People's Republic of Korea
50
Wgt. disability prev. (%)
5
40
30
20
10
0
40
30
20
10
0
35
40
45
50 55 60
Age groups
65
70
75
35
37
40
45
50 55 60
Age groups
3.6 Calculation of DALE
Sullivan’s method was used to compute DALE for each Member State from the country life
table and the severity-weighted prevalence estimates. Sullivan's method involves using the
observed prevalence of disability at each age in the current population (at a given point of
time) to divide the hypothetical years of life lived by a period life table cohort at different ages
into years with and without disability. The method is illustrated in Table 6.
Table 6. Illustration of Sullivan's method for the calculation of disability-free life expectancy.
Ordinary life table
Survivors
Age
0
lx
100000
Disability
Years
Years
Years Expectation prevalence
(%)
lived Lx
of life ex
with
without
496210
74.98
disability
4.5
22130
LED
disability disability
474080
DFLE
LE with Disability
16.60
-free LE
58.38
5
99134
495425
70.63
9.6
47506
447919
16.52
54.11
10
99045
495018
65.69
8.6
42568
452450
16.05
49.64
15
98940
493916
60.76
5.7
28100
465816
15.64
45.12
20
98572
491448
55.98
7.6
37433
454015
15.41
40.56
25
97997
488469
51.29
8.5
41623
446846
15.12
36.17
30
97383
485285
46.60
10.6
51280
434005
14.79
31.81
35
96722
481816
41.90
12.2
59013
422803
14.36
27.54
40
95988
477781
37.20
14.3
68247
409534
13.86
23.34
45
95079
472220
32.53
17.9
84507
387713
13.27
19.26
50
93701
463324
27.97
23.5
108766
354558
12.57
15.40
55
91452
448652
23.59
30.9
138780
309872
11.68
11.90
60
87702
424469
19.48
41.6
176738
247731
10.60
8.88
65
81656
386806
15.73
44.0
170265
216541
9.22
6.50
70
72512
332217
12.38
58.3
193526
138691
8.04
4.34
75
59796
259645
9.45
59.6
154714
104931
6.51
2.94
80
43550
173081
7.02
73.2
126672
46409
5.39
1.63
85
25802
132424
5.13
81.5
107916
24508
4.18
0.95
Notes: First four columns are from a standard life table for a population.
lx is the number of survivors at age x in the hypothetical life table cohort.
Lx is the number of years of life lived by the life table cohort between ages x and x+5.
prevx is the prevalence of disability between ages x and x+5 in the population
Years lived with disability YDx = Lx * prevx,
Years lived without disability YWDx = Lx * (1-prevx)
DFLEx = Sum of years lived without disability for ages x and above, divided by lx
DLEx = Sum of years lived with disability for ages x and above, divided by lx
DALE can be calculated using the same method as illustrated in Table 7 where disability
prevalence is replaced by severity-weighted disability prevalence.
38
Using standard notation for the country life table parameters, we calculated DALE at age x as
follows:
Dx
Severity-weighted prevalence of disability between ages x and x+5
YDx = Lx * Dx
Equivalent years of healthy life lost due to disability between
ages
x and x+5
YWDx = Lx * (1- Dx)
Equivalent years of healthy life lived between ages x and x+5
Lx is the total years lived by the life table population between
ages x and x+5
DALE at age x is the sum of YWDi from i = x to w (the last open-ended age interval in the life
table) divided by lx (survivors at age x):
æ w
ö
DALE x = ç å YWDi ÷ / l x
è i=x
ø
(11)
æ w
ö
DLE x = ç å YDi ÷ / l x = LE x - DALEx
è i=x
ø
(12)
DLEx,the equivalent years of healthy life lost due to disability, is the sum of YDi from i = x to
w divided by lx (survivors at age x).
3.7 Uncertainty analysis
Uncertainty intervals have been estimated for life expectancies and other life table parameters
for WHO member countries as described by Salomon and Murray [89]. To capture the
uncertainty due to sampling, indirect estimation techniques and projections, a total of 1000
life tables was developed for each Member State in order to quantify the uncertainty
distribution of key life table parameters. In countries with a substantial HIV epidemic, recent
estimates of the level and uncertainty range of the magnitude of HIV/AIDS deaths by age and
sex have been incorporated into the life table uncertainty analysis.
The degree of uncertainty in country-level weighted disability prevalences has also been
estimated for each country. This is mainly determined by levels of uncertainty in
(a) epidemiological estimates for prevalence, incidence and/or severity of disability associated
with specific conditions,
(b) estimation of prevalence YLD from incidence YLD, and
(c) the approximate nature of adjustments for comorbidity.
For all these reasons, the uncertainty distributions across different ages are likely to be highly
correlated for children, for adults, and at older ages. To be conservative in our estimation of
uncertainty, we assumed 100% correlation between uncertainty at each age within broad age
ranges 0-14, 15-29, 30-44, 45-59, 60-69 and 70+ (so that for a given sample of the disability
prevalence distribution, it is high at all ages or low at all ages within one of these ranges). The
uncertainty distributions of the DALE estimates for each Member State were quantified by
developing a total of 1000 DALE life tables for each Member State which simultaneously
sampled the uncertainty in the life tables and the disability prevalences. The techniques used
39
for estimating uncertainty in life expectancies, DALE and in country rankings are discussed
further in Salomon and Murray [89].
The uncertainty ranges for these quantities given in Annex Table B and shown in graphs and
tables in Section 4 give the 10th percentile and 90th percentile of the relevant uncertainty
distributions. The ranges thus define 80% uncertainty intervals around the estimates. Rank
uncertainty is not only a function of the uncertainty of the DALE measurement for each
country, but also the uncertainty of the measurement of adjacent countries in ranking table.
Figure 15. Weighted disability prevalence (%), by age, sex, and WHO mortality sub-region, 1999
50%
Males
Weighted disability prevalence (%)
Age
0-4
40%
5-14
15-44
45-59
30%
60+
20%
10%
0%
WprA
EurA
AmrA
EmrB WprB EurB
EurC SearB AmrB EmrD SearD AmrD
40
AfrD
AfrE
50%
Females
Weighted disability prevalence (%)
Age
0-4
40%
5-14
15-44
45-59
30%
60+
20%
10%
0%
WprA
EurA
AmrA
EmrB WprB
EurB
EurC SearB AmrB EmrD SearD AmrD
41
AfrD
AfrE
4. Results
Using the methods outlined in the previous Section, we have estimated healthy life
expectancy (DALE) for males and females in the 191 Member States of WHO for 1999, as
well as for 14 mortality sub-regions of the world, the 6 WHO regions and for the total global
population. These estimates of healthy life expectancy are based on country-specific estimates
of mortality, cause of death patterns, epidemiological analyses and health survey data where
available. We describe the results in this Section. Estimates and uncertainty intervals for
DALE at age 0 and 60 are given in full in Annex Table A for each Member State.
4.1 DALE for WHO regions and the world in 1999
Country-level estimates for mortality and disability were aggregated to estimate life
expectancy (LE) and healthy life expectancy (DALE) for each of the six WHO Regions and
for the world (Table 7). Regional healthy life expectancies at birth in 1999 ranged from a low
of 37 years for African males to a high of almost 70 years for females in the low mortality
countries of mainly Western Europe. This is an almost 2-fold difference in healthy life
expectancy between major regional populations of the world. Regional healthy life
expectancies at age 60 in 1999 ranged from a low of 8.4 years for African males to a high of
around 22 years for females in Europe and North America.
The difference between DALE and total life expectancy is DLE (expected years “lost” due to
disability), shown in Figure 16 as the light shaded areas. The equivalent healthy years “lost”
Table 7. Life expectancy (LE), healthy life expectancy (DALE), and years lost to disability as per cent of
total LE (DLE%), at birth and at age 60, by sex and total, WHO regions and world, 1999
Persons
WHO
Region
Males
Females
DALE
(years)
LE
(years)
DLE%
(%)
DALE
(years)
LE
(years)
DLE%
(%)
DALE
(years)
LE
(years)
DLE%
(%)
AFRO
37.5
46.3
18.9
37.3
45.6
18.1
37.8
47.0
19.7
AMRO
65.2
72.7
10.4
62.3
69.5
10.4
68.1
76.0
10.3
EMRO
54.5
61.9
12.0
54.4
61.2
11.1
54.6
62.6
12.8
EURO
66.5
72.9
8.8
63.1
69.1
8.6
69.8
76.7
9.0
SEARO
53.9
61.2
11.9
53.3
60.2
11.5
54.4
62.1
12.4
WPRO
63.3
70.4
10.2
61.9
68.5
9.7
64.7
72.4
10.7
World
56.8
64.5
11.9
55.8
62.5
10.7
57.8
66.4
12.9
At birth
At age 60
AFRO
9.0
14.9
39.5
8.4
14.4
42.0
9.6
15.3
37.1
AMRO
16.0
20.5
21.7
14.5
18.6
22.3
17.5
22.3
21.3
EMRO
10.6
14.5
27.2
10.6
14.4
26.6
10.6
14.6
27.7
EURO
15.8
19.6
19.4
14.0
17.5
19.7
17.6
21.8
19.2
SEARO
12.1
16.1
24.7
11.5
15.4
25.3
12.7
16.8
24.1
WPRO
13.7
18.0
23.9
12.6
16.5
23.7
14.9
19.6
24.0
World
13.5
18.0
24.7
12.6
16.6
23.6
14.4
19.4
25.6
42
due to disability range from 18.9% (of total life expectancy at birth) in Africa to 8.8% in the
European region. The equivalent healthy years “lost” due to disability at age 60 are a higher
percentage of remaining life expectancy, due to the higher prevalence of disability at older
ages. These range from around 40% in sub-Saharan Africa to around 20% in developed
countries.
Figure 16. Disability-adjusted life expectancy (DALE), healthy years lost due to disability (DLE)
and life expectancy (LE), at birth for total populations, WHO regions, 1999
80
75
70
DLE
LE
DALE
Expectancy (years)
65
Global LE
60
Global DALE
55
50
45
40
35
30
AFRO
SEARO
EMRO
WPRO
AMRO
EURO
WHO Region
4.2 DALE for the 14 mortality subregions of the world in 1999
When DALE is calculated for the 14 mortality subregions of the world, the range is even
greater (Table 8). Subregional healthy life expectancies at birth in 1999 ranged from a low of
35 years for males in the very high mortality subregion of Africa to a high of almost 77 years
for females in the low mortality countries of the Western Pacific region (these include Japan,
Australia, New Zealand and Singapore). Figure 17 compares DALE and total life expectancy
at birth across the 14 mortality subregions in 1999. Figure 18 similarly compares DALE and
total life expectancy at age 60 across the 14 mortality subregions in 1999.
The very low health expectancies of the African countries in both subregions D and E reflects
the high burden of HIV/AIDS, malaria, other communicable, maternal, perinatal and
nutritional conditions, and injuries.
Figure 19 summarises the relationship between life expectancy and DALE for the 14 mortality
subregions of the world, for both men and women. The gap between LE and DALE ranges
from between 8 and 9 years for the subregions in Africa to between 5 and 6 years for females
in developed countries. Despite the fact that people live longer in the richer, more
43
Table 8. Life expectancy (LE), healthy life expectancy (DALE), and years lost to disability as per cent of
total LE (DLE%), at birth and at age 60, by sex and total, by mortality subregion, 1999
Persons
WHO
Region
DALE
(years)
LE
(years)
Males
DLE%
(%)
DALE
(years)
LE
(years)
Females
DLE%
(%)
DALE
(years)
LE
(years)
DLE%
(%)
At birth
AFRO D
40.3
49.4
18.5
40.0
48.5
17.6
40.6
50.4
19.4
AFRO E
35.4
43.9
19.3
35.3
43.3
18.6
35.6
44.5
20.0
AMRO A
70.4
76.9
8.4
67.9
74.0
8.3
73.0
79.8
8.5
AMRO B
62.7
71.0
11.7
59.5
67.5
11.9
65.9
74.5
11.6
AMRO D
55.9
64.1
12.7
54.5
62.3
12.6
57.4
65.8
12.8
EMRO B
61.0
67.7
10.0
61.7
67.3
8.3
60.2
68.2
11.6
EMRO D
52.5
60.1
12.8
52.0
59.3
12.3
52.9
61.0
13.3
EURO A
71.8
77.7
7.6
69.0
74.5
7.4
74.7
81.0
7.7
EURO B
62.8
69.8
10.0
60.7
67.0
9.4
64.9
72.6
10.6
EURO C
61.6
68.4
9.9
56.6
62.9
10.0
66.6
73.9
9.8
SEARO B
59.8
67.8
11.8
58.5
66.1
11.6
61.1
69.4
11.9
SEARO D
52.8
60.0
12.0
52.4
59.2
11.5
53.1
60.7
12.5
WPRO A
74.4
80.6
7.7
71.7
77.3
7.2
77.1
83.9
8.2
WPRO B
61.9
69.3
10.6
60.8
67.5
10.0
63.1
71.0
11.1
World
56.8
64.5
11.9
55.8
62.5
10.7
57.8
66.4
12.9
9.6
15.6
38.3
9.1
15.1
40.1
10.2
16.0
36.6
At age 60
AFRO D
AFRO E
8.5
14.2
40.6
7.7
13.8
43.9
9.2
14.7
37.5
AMRO A
17.1
21.5
20.2
15.4
19.5
20.8
18.8
23.4
19.6
AMRO B
15.1
19.7
23.4
13.7
17.9
23.7
16.4
21.4
23.2
AMRO D
11.4
15.8
28.0
11.1
15.3
27.4
11.7
16.4
28.5
EMRO B
11.0
14.5
23.7
11.4
14.5
21.1
10.6
14.5
26.4
EMRO D
10.4
14.6
28.7
10.2
14.4
29.2
10.5
14.7
28.2
EURO A
17.7
21.4
17.2
15.8
19.1
17.3
19.6
23.7
17.2
EURO B
14.5
18.3
20.8
13.3
16.7
20.5
15.7
19.9
21.0
EURO C
13.2
17.2
23.3
11.0
14.7
25.2
15.3
19.7
21.9
SEARO B
15.3
20.1
23.9
15.3
20.1
23.9
15.4
20.2
23.9
SEARO D
11.5
15.4
24.9
10.9
14.6
25.7
12.2
16.1
24.1
WPRO A
19.6
23.7
17.2
17.6
21.1
16.5
21.6
26.3
17.7
WPRO B
12.8
17.1
25.5
11.8
15.8
25.3
13.7
18.4
25.6
World
13.5
18.0
24.7
12.6
16.6
23.6
14.4
19.4
25.6
44
Figure 17. Disability-adjusted life expectancy (DALE) and life expectancy (LE), at birth for total
populations, by mortality sub-region, 1999
WPR A
EUR A
AMR A
EUR B
AMR B
WPR B
EUR C
EMR B
SEAR B
AMR D
SEAR D
LE
EMR D
DALE
AFR D
AFR E
30
40
50
60
70
80
Expectation (Years)
Figure 18. Disability-adjusted life expectancy (DALE) and life expectancy (LE), at age 60 total
populations, by mortality sub-region, 1999
WPR A
EUR A
AMR A
SEAR B
AMR B
EUR B
EUR C
WPR B
SEAR D
AMR D
LE
EMR B
EMR D
DALE
AFR D
AFR E
0
10
20
Expectation (Years)
45
30
Figure 19. Life expectancy and disability-adjusted life expectancy at birth and at age 60,
for males and females, by WHO mortality subregion, 1999. The dotted line represents a
situation of no time lived with disability, so that life expectancy and DALE coincide.
20
75
Male
Male
70
Female
Female
65
DALE at age 60 (years)
DALE at birth (years)
80
60
55
50
45
15
10
40
35
5
30
40
45
50
55
60
65
70
75
80
10
85
15
20
25
Life expectancy at age 60 (years)
Life expectancy at birth (years)
developed countries, and have greater opportunity to acquire non-fatal disabilities in older
age, disability has a greater absolute (and relative) impact on healthy life expectancy in poorer
countries. Separating life expectancy into equivalent years of good health and years of lost
good health thus widens rather than narrows the difference in health status between the rich
and the poor countries.
The relative contributions of diseases and injuries to variations in DALE are best summarised
in terms of the loss of healthy life measured in DALYs. The World Health Report provides
detailed estimates of DALYs for over 100 disease and injury categories for the 14 mortality
subregions. The leading causes of DALYs worldwide are shown in Table 9 for males and
females separately. While the rankings are broadly similar for the two sexes, there are
important differences. Thus while perinatal conditions, HIV/AIDS and lower respiratory
infections are the three leading causes of DALYs their relative importance differs slightly for
males and females. More importantly, depression is the fourth leading cause of disease
burden for females but ranks ninth for males. Maternal conditions are the seventh leading
cause for females, causing almost 4% of their global disease burden in 1999. Road traffic
accidents are a leading cause of overall disease and injury burden for males (3.9%) but not for
females. In parts of South Asia, Eastern Europe and the Western Pacific, 20% or more of the
entire disease and injury burden is due to injuries alone.
There are marked contrast in epidemiological patterns between rich and poor regions of the
world (Table 9). Thus in the more developed countries, the share of disease burden due to
communicable, maternal, perinatal and nutritional conditions is typically around 5%,
compared with 70-75% in Africa. Specifically, the leading causes of disease burden in Africa
in 1999 were HIV/AIDS (20.0%). Malaria (10.0%) and acute lower respiratory infections
(8.4%), compared with ischaemic heart disease, depression, alcohol dependence and stroke in
the industrialized countries (Table 9).
46
Table 9. Top 10 causes of loss of healthy life expectancy (in DALYs) for the 14 mortality sub-regions, 1999
GLOBAL
DALYs (000)
%
96 682
89 819
89 508
72 063
59 030
58 981
6.7
6.2
6.2
5.0
4.1
4.1
1
2
3
4
5
6
Acute lower respiratory infections
HIV/AIDS
Perinatal conditions
Diarrhoeal diseases
Unipolar major depression
Ischaemic heart disease
7
Cerebrovascular disease
49 856
3.5
8
Malaria
44 998
3.1
9
Road traffic accidents
39 573
2.8
10
COPDa
38 156
2.7
1 438 154
100
DALYs (000)
%
18 600
15 778
14 858
12 351
11 867
8 762
4 954
3 180
3 158
3 011
11.7
10.0
9.4
7.8
7.5
5.5
3.1
2.0
2.0
1.9
HIV/AIDS
Malaria
Acute lower respiratory infections
Diarrhoeal diseases
Perinatal conditions
Measles
Maternal conditions
Tuberculosis
Congenital abnormalities
Road traffic accidents
158 439
100
All causes
DALYs (000)
%
3 298
2 623
1 863
1 628
1 612
1 509
1 410
1 267
1 141
1 087
8.5
6.8
4.8
4.2
4.2
3.9
3.7
3.3
3.0
2.8
Unipolar major depression
Alcohol dependence
Perinatal conditions
Nutritional/endocrine disorders
Homicide and violence
Ischaemic heart disease
Road traffic accidents
Acute lower respiratory infections
Osteoarthritis
Cerebrovascular disease
38 627
100
All causes
All causes
AFRO D
1
2
3
4
5
6
7
8
9
10
Malaria
HIV/AIDS
Acute lower respiratory infections
Perinatal conditions
Diarrhoeal diseases
Measles
Maternal conditions
Congenital abnormalities
Tuberculosis
Road traffic accidents
All causes
AMRO A
1
2
3
4
5
6
7
8
9
10
Ischaemic heart disease
Unipolar major depression
Alcohol dependence
Diabetes mellitus
Road traffic accidents
Trachea/bronchus/lung
Cerebrovascular disease
COPDa
Alzheimer and other dementias
Osteoarthritis
All causes
AMRO D
1
2
3
4
5
6
7
8
9
10
Perinatal conditions
Acute lower respiratory infections
Diarrhoeal diseases
Alcohol dependence
Unipolar major depression
Diabetes mellitus
HIV/AIDS
Nutritional/endocrine disorders
Tuberculosis
Cirrhosis of the liver
All causes
AFRO E
AMRO B
DALYs (000)
%
1 283
1 054
770
656
646
625
621
588
471
453
7.8
6.4
4.7
4.0
3.9
3.8
3.8
3.6
2.9
2.8
Ischaemic heart disease
Unipolar major depression
Perinatal conditions
Cerebrovascular disease
Diarrhoeal diseases
Acute lower respiratory infections
Road traffic accidents
Maternal conditions
Anaemias
Nutritional/endocrine disorders
EMRO B
16 346
100
All causes
DALYs (000)
%
58 671
18 238
16 871
12 454
11 746
8 701
7 571
5 563
3 257
3 207
27.3
8.5
7.8
5.8
5.5
4.0
3.5
2.6
1.5
1.5
214 921
100
DALYs (000)
%
4 240
3 762
3 707
3 386
3 000
2 915
2 901
2 315
2 224
2 208
6.0
5.3
5.2
4.8
4.2
4.1
4.1
3.3
3.1
3.1
70 969
100
DALYs (000)
%
1 484
1 312
1 134
1 041
977
921
881
704
607
492
7.1
6.3
5.4
5.0
4.7
4.4
4.2
3.4
2.9
2.4
20 895
100
(continued)
47
Table 9 (continued). Top 10 causes of loss of healthy life expectancy for the 14 mortality sub-regions, 1999
EMRO D
1
2
3
4
5
6
7
8
9
10
Perinatal conditions
Acute lower respiratory infections
Diarrhoeal diseases
Congenital abnormalities
Ischaemic heart disease
Unipolar major depression
Measles
Malaria
Road traffic accidents
Cerebrovascular disease
DALYs (000)
10 621
9 625
9 146
5 446
3 588
3 227
3 020
2 727
2 298
2 277
%
10.4
9.5
9.0
5.4
3.5
3.2
3.0
2.7
2.3
2.2
EURO A
Ischaemic heart disease
Unipolar major depression
Cerebrovascular disease
Alcohol dependence
Trachea/bronchus/lung
Alzheimer and other dementias
Osteoarthritis
COPDa
Road traffic accidents
Diabetes mellitus
101 688
EURO B
1
2
3
4
5
6
7
8
9
10
Ischaemic heart disease
Cerebrovascular disease
Unipolar major depression
Acute lower respiratory infections
Road traffic accidents
Perinatal conditions
Alcohol dependence
Osteoarthritis
Diabetes mellitus
Diarrhoeal diseases
%
8.2
7.5
5.4
3.9
3.7
3.5
3.3
3.2
2.4
2.4
EURO C
Ischaemic heart disease
Cerebrovascular disease
Unipolar major depression
Self-inflicted
Road traffic accidents
Osteoarthritis
Poisoning
COPDa
Homicide and violence
Alcohol dependence
36 484
SEARO B
1
2
3
4
5
6
7
8
9
10
Perinatal conditions
Road traffic accidents
Tuberculosis
Unipolar major depression
Acute lower respiratory infections
Cerebrovascular disease
Anaemias
Ischaemic heart disease
Falls
Self-inflicted
%
11.2
7.7
6.1
5.5
5.1
3.6
3.1
3.1
2.7
2.1
SEARO D
Acute lower respiratory infections
Diarrhoeal diseases
Perinatal conditions
Ischaemic heart disease
Falls
Unipolar major depression
Congenital abnormalities
Tuberculosis
Cerebrovascular disease
Anaemias
56 604
WPRO A
1
2
3
4
5
6
7
8
9
10
Unipolar major depression
Cerebrovascular disease
Ischaemic heart disease
Alcohol dependence
Osteoarthritis
Alzheimer and other dementias
Self-inflicted
Stomach
Trachea/bronchus/lung
Road traffic accidents
%
8.2
7.6
4.7
4.7
3.9
3.8
3.4
3.0
2.9
2.7
WPRO B
COPDa
Unipolar major depression
Cerebrovascular disease
Perinatal conditions
Acute lower respiratory infections
Road traffic accidents
Self-inflicted
Anaemias
Falls
Congenital abnormalities
15 235
(a)
DALYs (000)
7 113
5 584
2 297
2 200
1 589
1 481
1 396
1 323
1 240
1 140
%
14.0
11.0
4.5
4.3
3.1
2.9
2.7
2.6
2.4
2.2
DALYs (000)
33 746
28 960
26 353
20 133
13 742
12 165
10 889
10 648
8 639
8 462
%
9.5
8.1
7.4
5.7
3.9
3.4
3.1
3.0
2.4
2.4
355 876
DALYs (000)
1 255
1 160
721
714
599
580
513
462
448
414
9.7
6.9
5.8
5.2
3.6
3.5
3.1
2.9
2.9
2.4
50 868
DALYs (000)
6 362
4 377
3 453
3 104
2 904
2 041
1 749
1 732
1 515
1 207
4 757
3 376
2 857
2 562
1 774
1 737
1 508
1 441
1 423
1 176
%
48 999
DALYs (000)
2 986
2 727
1 964
1 435
1 355
1 269
1 192
1 151
876
875
DALYs (000)
DALYs (000)
22 593
17 132
15 982
12 914
10 798
8 615
8 407
7 764
7 316
6 955
252 204
Chronic obstructive pulmonary disease (chronic bronchitis and emphysema)
48
%
9.0
6.8
6.3
5.1
4.3
3.4
3.3
3.1
2.9
2.8
4.3 Global DALE and disability severity distribution, 1999
Overall, for the entire population of the world, average life expectancy at birth in 1999 was
64.5 years, an increase of almost 6 years over the last two decades. Global healthy life
expectancy at birth is 56.8 years, 7.7 years lower than total life expectancy at birth (see Table
7). At the global level, female healthy life expectancy is 57.8 years, 2.0 years higher than male
healthy life expectancy at 55.8 years.
Figure 20 shows global estimates of DALE, DLE and LE for 1999 at fifteen year age
intervals. Expectancies at ages greater than zero are plotted as bars based on the relevant age,
so that the total height of the bar represents expected age at death. Expected age at death rises
with achieved age, since the person has already survived the risk of death at earlier ages.
Although the primary focus of the DALE analyses for the World Health Report 2000 has been
on estimating severity-weighted disability prevalence and disability-adjusted life expectancy,
we have also made an estimate of the global pattern of disability prevalence in terms of the
seven disability severity classes used in the Global Burden of Disease 1990 Study [7, 34].
Figure 21 shows the estimated global distribution of disability by severity, in terms of the
corresponding survival curves for the global population. As discussed in Section 2.3, total
global healthy life expectancy (DALE) is the total area under this curve, weighted by the
complement (one minus) the average disability weight for each area.
Figure 20. Global disability-adjusted life expectancy (DALE), global healthy years lost due to disability
(DLE) and global life expectancy (LE), by age, 1999
Years lost due to disability (DLE)
Health expectancy (DALE)
90
Males
80
80
70
70
Expectation (years)
Expectation (years)
90
60
50
40
30
60
50
40
30
20
20
10
10
0
Females
0
0
15
30
45
60
75
0
Age (years)
15
30
45
Age (years)
49
60
75
Figure 21. Estimated global survival curves for seven classes of disability severity, 1999.
Disability severity classes and average weights as defined in Murray and Lopez (1996),
average expected years lived in each disability class shown in brackets.
100
90
VII (1.2)
80
VI (3.3)
V
Survivors (%)
70
III (6.5)
60
50
(2.1)
IV (3.4)
II (15.9)
40
I (13.8)
30
20
No disability
(18.1 years)
10
0
0
20
40
60
Age (years)
80
100
4.4 DALE estimates and rankings for WHO Member States, 1999
Japan leads the world with an average healthy life expectancy of 74.5 years at birth in 1999
(Table 10). Female healthy life expectancy in Japan was 77.2 years for females and 71.9 years
for males in 1999. After Japan, in second and third places, are Australia (73.2 years) and
France (73.1 years), followed by a number of other industrialized countries of Western
Europe. Note however, that there is a considerable range of uncertainty in the ranks for
countries other than Japan, with an 80% uncertainty range of around 6 to 10 ranks for many
countries. Canada is in twelfth place (72.0 years) with an uncertainty range of 8–14 in ranking
and the USA in 24th place (70.0 years with a ranking range of 22–27). The bottom ten
countries are all African countries with health expectancies in 1999 as low as 25.9 years in
Sierra Leone.
Table 10. DALE at birth (years), top 10 and bottom 10 countries, 1999
Rank
(a)
Uncertainty in ranka
Top 10 countries
Uncertainty in ranka
Bottom 10 countries
1
Japan
74.5
1
1
Sierra Leone
25.9
191
2
Australia
73.2
2 - 10
2
Niger
29.1
189 - 190
3
France
73.1
2-7
3
Malawi
29.4
188 - 190
4
Sweden
73.0
2-8
4
Zambia
30.3
188 - 189
5
Spain
72.8
3 - 11
5
Botswana
32.3
185 - 187
6
Italy
72.7
3 - 10
6
Uganda
32.7
183 - 187
7
Greece
72.5
5 - 11
7
Rwanda
32.8
181 - 187
8
Switzerland
72.5
4 - 11
8
Zimbabwe
32.9
181 - 187
9
Monaco
72.4
2 - 15
9
Mali
33.1
181 - 186
10
Andorra
72.3
6 - 15
10
Ethiopia
33.5
180 - 185
th
th
10 percentile and 90 percentile of rank for DALE based on estimated uncertainty in DALE for countries.
50
Figure 22. Disability-adjusted life expectancy at birth, 1999
Healthy Life Expectancy
Espérance de vie en santé
Esperanza de vida saludable
High expectancy / Espérance élevée / Alta esperanza
71.0 - 74.6
67.0 - 70.9
64.0 - 66.9
62.0 - 63.9
60.0 - 61.9
55.0 - 59.9
45.0 - 54.9
35.0 - 44.9
25.9 - 34.9
Low expectancy / Espérance faible / Baja esperanza
No Data / Pas de données / No hay datos
Measure: Disability adjusted life expectancy at birth, both sexes, estimates f or 1997
Mesure: Espérance de vie à la naissance corrigée de l'incapacité, population totale, estimations pour 1997
Medida: Esperanza de vida al nacer ajustada por incapacidad, ambos sexos, estimaciones para 1997
The boundaries and names shown and the designations used on this map do not imply the expression of any opinion whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area
Ó
or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines for which there may not yet be full agreement.
W HO 2000. All rights reserved
52
The worldwide pattern of health expectancies in 1999 is shown in Figure 22, highlighting the
enormous variation between developing countries and developed countries, as well as
between the lower and higher mortality regions of Europe. Japan leads the world with an
average healthy life expectancy of 74.5 years at birth in 1999. Female healthy life expectancy
in Japan was 77.2 years for females and 71.9 years for males in 1999. Total life expectancies
were 77.6 and 84.3 for males and females. The high DALE reflects the very high life
expectancy (low mortality) together with similar levels of disability to other low mortality
countries.
Australia ranks 2nd in the world with a DALE of 73.2. Note that the uncertainty ranges for
DALE estimates are quite large: the 80% range for Australia is 72.7–74.1, which translates to
an uncertainty range for its rank of 2–10. Uncertainty ranges for overall estimates of DALE at
birth are plotted for the 191 WHO Member States in Figure 23 against mean DALE. Figure 24
shows a similar plot of the uncertainty in DALE ranks for Member States.
Estimated healthy life expectancy for New Zealand is 69.2 years, 4 years lower than Australia.
The life expectancy difference is lower at 2.8 years. New Zealanders lived longer than
Australians until the 1970s, but during the 1980s New Zealanders fell behind Australians [90].
The mortality gap is compounded by a higher level of disability in New Zealand, reflecting
higher rates of cardiovascular diseases, diabetes and injuries and the higher proportion of
Indigenous people in the population.
The USA ranks 24th with a DALE of 70.0, compared to Canada (12th with 72.0) and Cuba
(33rd at 68.4). Other countries with reasonably high healthy life expectancies in the Americas
include Dominica (69.8 years), Chile (68.6 years), Uruguay (67.0 years), and Argentina and
Costa Rica at 66.7 years. Brazil is split, with a high healthy life expectancy in its southern
half, and a lower one in the north. The total average is a relatively low 59.1 years, at 55.2 for
males and 62.9 for females.
China has a healthy life expectancy above the global average, at 62.3 years, 63.3 years for
women and 61.2 for men. Other countries in the Asian region have lower DALE. Improving
health in Viet Nam has resulted in a healthy life expectancy of 58.2 years, while Thailand has
not improved significantly over the past decade, though it is still ahead of Viet Nam at 60.2
years. Healthy life expectancy in Myanmar is just 52 years, substantially behind its Southeast
Asian neighbors.
In Russia, healthy life expectancy is 66.4 for females, 3 years below the European average, but
just 56.1 years for males, 7 years below the European average. This is one of the widest sex
gaps in the world and reflects the sharp increase in adult male mortality in the early 1990s (see
Section 4.5 below). Similar rates exist for other countries of the former Soviet Union.
The bottom 10 countries for DALE are all in sub-Saharan Africa, where the HIV-AIDS
epidemic is rampant (Table 10). Life expectancy in several countries in southern Africa has
been reduced 15-20 years in comparison to life expectancy without HIV. Other African
countries have lost 5-10 years of life expectancy because of HIV [84]. AIDS is now the
leading cause of death in Sub-Saharan Africa, far surpassing the traditional deadly diseases of
malaria, tuberculosis, pneumonia and diarrheal disease. AIDS killed 2.2 million Africans in
1999, versus 300,000 AIDS deaths 10 years previously.
Figures 25 and 26 show worldwide patterns in DALE at birth for males and females
respectively. Annex Table A provides details of the estimates for each Member State.
53
Figure 23.Uncertainty in DALE, versus mean DALE at birth, 191 Member States, 1999
A ustralia
San M arino
Finland
P o rtugal
Cro atia
P anama
B o snia and
A zerbaijan
Co lo mbia
B elarus
Ho nduras
Jo rdan
M o ro cco
Tuvalu
Turkmenistan
Demo cratic P eo ple's
Co mo ro s
Cô te d'Ivo ire
Swaziland
So malia
Ethio pia
Sierra Leo ne
20
30
40
50
60
70
80
80% uncertainty interval fo r DA LE at birth (years)
Figure 24.Uncertainty in DALE ranking, versus mean rank, 191 Member States, 1999
Japan
San M arino
M alta
New Zealand
A rmenia
Republic o f Ko rea
B ahrain
P araguay
China
Russian Federatio n
Jo rdan
B razil
M arshall Islands
M o ngo lia
Yemen
Senegal
Chad
Leso tho
Liberia
Sierra Leo ne
0
20
40
60
80
100
120
80% uncertainty interval fo r DA LE rank
54
140
160
180
200
Figure 25. Disability-adjusted life expectancy at birth, males, 1999
(Years)
25.8 - 33.7
33.8 - 40.1
40.2 - 47.1
47.2 - 53.3
53.4 - 57.3
57.4 - 60.6
60.7 - 64.1
64.2 - 67.4
67.5 - 71.9
No Data
The boundaries and names shown and the designations used on this map do not imply the expression of any opinion whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area
Ó WHO 2000. All rights reserved
or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines for which there may not yet be full agreement.
55
Figure 26. Disability-adjusted life expectancy at birth, females, 1999
(Years)
26.0 - 34.5
34.6 - 41.3
41.4 - 49.7
49.8 - 56.7
56.8 - 61.1
61.2 - 64.8
64.9 - 68.4
68.5 - 72.6
72.7 - 77.2
No Data
The boundaries and names shown and the designations used on this map do not imply the expression of any opinion whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area
Ó WHO 2000. All rights reserved
or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines for which there may not yet be full agreement.
56
4.5 Expectation of years lived with disability
There is a reasonably close correlation across countries between total life expectancy and
DALE, both at birth (Figure 27) and at age 60 (Figure 28). The relationship between DALE
and LE is very similar for males and females. However, when we examine the relationship
between DLE (=LE – DALE) and LE (Figures 29 and 30), it is apparent that DLE at birth
declines as LE increases. Taking into account the uncertainty intervals for both LE and DALE,
these declines differ significantly from zero at the p=0.10 level for both males and females
with regression slopes of –0.053 and –0.035 respectively. The difference between the male
and female slopes is also statistically significant. The corresponding regression slopes at age
60 are –0.043 and +0.034 for males and females respectively. Only the female trend differs
significantly from zero, but the male and female trends also differ significantly at age 60. This
suggests that, for females but not males, there is some evidence of an expansion of morbidity
at older ages, probably reflecting the greater life expectancies of older females than males
together with the increasing prevalence of disability with age.
At least cross-sectionally across countries in 1999, we estimate that there is an absolute
compression of morbidity overall: DLE at birth decreases slightly in absolute terms (number
of years) as life expectancy increases. This translates into a substantial decline in DLE as a
proportion of total life expectancy as the latter increases (Figures 31 and 32).
DLE represents the equivalent healthy years of life lost through living with disability resulting
from diseases and injuries. Health expectancies are lower than life expectancies by amounts
ranging from around 9 years in Africa to 6 years in countries such as Japan and Australia.
These lost healthy years range from 20% of total life expectancy at birth in sub-Saharan Africa
down to 10% for the low mortality countries in the Western Pacific region, primarily Japan,
Australia, New Zealand and Singapore (Figures 33 and 34). Countries with longer life
expectancy also have fewer lost years of healthy life due to disability. Higher levels of
mortality are generally accompanied by more disability. Table 11 shows the top ten and
bottom ten countries in the world in terms of DLE/LE%— the equivalent “lost” years of
healthy life due to years lived with disability as a percent of total life expectancy at birth.
Table 11. DLE/LE% at birth, top 10 and bottom 10 countries, 1999
Rank
Top 10 countries
DLE/LE%
Bottom 10 countries
DLE/LE%
1
Greece
7.0
1
Niger
25
2
United Kingdom
7.1
2
Sierra Leone
25
3
Austria
7.4
3
Mali
22
4
Spain
7.5
4
Uganda
22
5
Italy
7.7
5
Malawi
22
6
Netherlands
7.7
6
Liberia
22
7
France
7.8
7
Rwanda
21
8
Japan
7.9
8
Zambia
21
9
Australia
8.0
9
Madagascar
21
10
Belgium
8.0
10
Burkina Faso
21
57
Figure 27. Disability-adjusted life expectancy (DALE) by
total life expectancy at birth, by sex, 191 countries, 1999
85
Males
75
Females
DALE (years)
65
55
45
35
25
30
35
40
45
50
55
60
65
70
75
80
85
Life expectancy (years)
Figure 28. Disability-adjusted life expectancy (DALE) by
total life expectancy at age 60, by sex, 191 countries, 1999
25
Males
Females
DALE (years)
20
15
10
5
0
0
5
10
15
20
Life expectancy (years)
58
25
30
Figure 29. Expected years lost due to disability (DLE) by
total life expectancy at birth, by sex, 191 countries, 1999
14
DLE at birth (years)
12
10
8
6
Males
4
Females
Male
2
Female
0
30
35
40
45
50
55
60
65
70
75
80
85
Life expectancy (years)
Figure 30. Expected years lost due to disability (DLE) by
total life expectancy at age 60, by sex, 191 countries, 1999
10
Males
9
Females
Males
8
Females
DLE at age 60 (years)
7
6
5
4
3
2
1
0
0
5
10
15
20
25
Life expectancy (years)
59
30
Figure 31. Per cent of total life expectancy at birth lost due to disability, by sex,
191 countries, 1999
30
Males
DLE/LE at birth (%)
25
Females
20
15
10
5
0
30
35
40
45
50
55
60
65
70
75
80
85
Life expectancy (years)
Figure 32. Per cent of total life expectancy at age 60 lost due to disability, by sex,
191 countries, 1999
60
55
50
DLE/LE at age 60 (%)
45
40
35
30
25
20
15
Males
10
Females
5
0
0
5
10
15
20
Life expectancy (years)
60
25
30
Figure 33. Percentage of total life expectancy at birth lost due to disability, males, 1999
(Years)
6.7 - 7.9
8.0 - 9.0
9.1 - 10.4
10.5 - 11.6
11.7 - 13.5
13.6 - 15.4
15.5 - 17.4
17.5 - 19.2
19.3 - 24.3
No Data
The boundaries and names shown and the designations used on this map do not imply the expression of any opinion whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area
Ó WHO 2000. All rights reserved
or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines for which there may not yet be full agreement.
61
Figure 34. Percentage of total life expectancy at birth lost due to disability, females, 1999
(Years)
7.4 - 8.7
8.8 - 9.9
10.0 - 11.0
11.1 - 12.5
12.6 - 14.2
14.3 - 16.2
16.3 - 18.8
18.9 - 22.0
22.1 - 26.7
No Data
The boundaries and names shown and the designations used on this map do not imply the expression of any opinion whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area
Ó WHO 2000. All rights reserved
or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines for which there may not yet be full agreement.
62
4.5 Male-female differences in healthy life expectancy
Health expectancies at birth are higher for females than males in most regions in the world
(Figure 35). In 1999, female-male difference was highest at around 10 years or even higher in
many countries of the former Soviet Union (including the Russian Federation, Latvia, Estonia,
Belarus and Kazakhstan) and lowest (males 1 to 2 years higher than females) in countries in
the Eastern Mediterranean region (including Turkey, Algeria, Iran and Saudi Arabia). For
other countries of the world, the female-male difference in healthy life expectancy generally
increases as average life expectancy increases (Figures 36).
Figure 35. Female - male difference in DALE
DALE
at birth, versus total life expectancy,
expectancy,
mortality sub-regions, 1999
Figure 36. Female - male difference in
at birth, versus total life
191 countries, 1999
15
Female - male difference in DALE at birth (years)
Female - male difference in DALE at birth (years)
15
EUR C
10
AMR B
5
0
EMR B
10
5
0
-5
-5
30
35
40
45
50
55
60
65
70
75
80
30
85
35
40
45
50
55
60
65
70
75
80
Life expectancy (years)
Life expectancy (years)
Russia has one of the widest sex gaps in healthy life expectancy in the world: 66.4 years for
females at birth but just 56.1 years for males (Figure 37). The most common explanation is
the high incidence of male alcohol abuse, which led to high rates of accidents, violence and
cardiovascular disease. From 1987 to 1994, the risk of premature death increased by 70% for
Russian males. Since 1994, life expectancy has been improving for males. Similar rates exist
for other major countries of the former Soviet Union. In Ukraine, female babies can expect to
live an equivalent of 67.5 years of healthy life versus 58.5 years for male babies. In Belarus,
67.2 for female babies and 56.2 for male babies.
There is generally a close correlation between the female-male difference in DALE at birth
and the female-male difference in total life expectancy at birth (Figure 38). There are a
number of countries with higher female than male life expectancies but higher male than
female DALE. These fall in the bottom right quadrant in Figure 38 and include mainly
countries in the Eastern Mediterranean region and North Africa.
63
85
64
Figure 37. Sex difference in disability adjusted life expectancy at birth, 1999
Years
-2.1 - -0.9
-0.8 - 0.3
0.4 - 1.3
1.4 - 2.5
2.6 - 3.7
3.8 - 5.0
5.1 - 6.4
6.5 - 7.7
7.8 - 11
No Data
The boundaries and names shown and the designations used on this map do not imply the expression of any opinion whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area
Ó
or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines for which there may not yet be full agreement.
WHO 2000. All rights reserved
65
In North Africa and the Middle East, males and females have similar levels of healthy life
expectancy, which is unusual. Also, the position of women in these societies is often not
good, Less care is given to female children, and they have a higher risk for reproductive
deaths than in other countries. In Saudi Arabia, the overall healthy life expectancy is 64.5
years -- 65.1 for male babies and 64.0 for female babies. In Bahrain, the overall healthy life
expectancy is 64.4, but 63.9 for male babies and 64.9 for female babies; Qatar, 63.5 overall,
and 64.2 for male babies, 62.8 for females; and Kuwait, 63.2 overall, with 63.0 for male
babies and 63.4 for female babies.
Figure 38. Female - male difference in DALE at birth, versus female - male
difference in DALE at birth, 191 countries, 1999
Female - male difference in DALE at birth (years)
15
10
5
0
-5
-5
0
5
10
Fem ale-m ale differences in life expectancy (years)
66
15
5. Discussion and conclusions
This paper has described the methods used to produce first estimates of healthy life
expectancy (DALE) for 191 countries in 1999. These estimates are based on a very substantial
information and analytic base for mortality rates and life expectancies, cause of death
distributions, internally consistent estimates of the incidence, prevalence and disability
distributions for 109 disease and injury causes by age group, sex and region of the world, and
an analysis of 60 representative health surveys across the world.
As with any innovative approach, methods and data sources can and will be refined and
improved. Careful scrutiny and use of the DALE results reported here will lead to
progressively better estimates. A wide range of people from WHO programs, from countries
and other agencies have already been involved and consulted in the development of these
initial estimates of DALE for WHO member countries. It is anticipated that continuing and
increasing collaboration and consultation will lead to progressive refinement and
improvement of the initial estimates presented here.
Over the next few years, WHO is planning a range of methodological and data collection
exercises to improve and extend the information base underpinning the estimates of healthy
life expectancies for Member States. These include:
§ Improved methods for calculating life tables for WHO member states and for estimating
cause of death distributions,
§ Complete revision of the Global Burden of Disease Study for 14 regions of the world for
the year 2000. To the extent possible, this will also provide country-level estimates of the
incidence, prevalence and severity of specific health conditions causing significant levels
of disability. It will involve widespread consultation with countries and with
epidemiologistsand other disease experts.
§ Improving the comparability and reliability of health survey data.
WHO is developing a standardized description of health states for use in population surveys
and health state valuation. The resulting instrument is being piloted in 10 countries over the
next year and will also be validated against objective measurements. The objective is to
develop an instrument with the maximum cross-cultural validity, and usability by younger and
older adults with widely varying education levels and cultural backgrounds, and to understand
better the systematic determinants of differences between self-report and underlying true
health.
Another objective for the ongoing WHO population survey work is to facilitate reliable and
valid measurements of valuations of time spent in health states in populations across the
world. The aim is to obtain large scale empirical assessment in many different countries to
inform health state valuations for the calculation of DALES for Member States.
Future burden of disease analyses should ensure to the extent possible, that there is
consistency between the condition-specific estimates for disabling sequelae and measured
population levels of these impairments and disabilities. Given the limitations of self-report
general disability data, both in terms of underreporting of certain types and causes of
disability, and in terms of the problems of ensuring comparability of self-report data across
populations, it is likely that the combination of condition-specific analyses with populationlevel data for specific disabilities and impairments will continue to be the analytic approach of
choice for some considerable time. The use of population-level impairment and disability data
67
will also assist in developing improved methods for addressing comorbidity. This approach is
also required for causal attribution, where it will be critical to link diseases, injuries and risk
factors to one or several average health states.
The new WHO framework for performance assessment also measures health inequality across
individuals. Eventually it will use inequalities in DALE. For the World Health Report 2000,
inequality has been measured by looking at differences in child survival, because that data is
available now. As adult data are added, it is planned to estimate inequalities in DALE for
Member States.
In conclusion, we believe that healthy life expectancy (DALE) is the most appropriate single
summary measure of population health for the measurement of level of health for Member
States. Unlike other forms of health expectancy, it takes into account differences in the
severity distribution of health states between populations. As WHO is concerned not just to
increase the numbers of people in full health, but also to reduce the severity of disability for
people in less than full health, DALE is the most appropriate summary measure for the
comparison of level of health of populations. Improvements in the estimation of DALE for
Member States will require further collaborative efforts between WHO and Member States to
improve population health survey data, to improve the completeness of death registration
systems, and to improve the estimation of burden of diseases and injuries at country level.
68
References
1.
2.
3.
World Health Organization (2000). World Health Report 2000. Geneva, World Health
Organization.
Sadana R. Mathers C, Lopez AD, Murray CJL (2000). Comparative analysis of more than 50
household surveys on health status. Geneva, World Health Organization, 2000 (GPE
Discussion Paper No. 15).
Murray CJL (1996). Rethinking DALYs. In: Murray CJL. and Lopez A. eds. The Global
Burden of Disease. Cambridge, Harvard University Press, 1996, chapter 1.
4.
World Bank (1993). World Development Report 1993. Investing in health. New York: Oxford
University Press for the World Bank.
5.
World Health Organization (1996). Investing in health research and development. Report of
the Ad Hoc Committee on Health Research Relating to Future Intervention Options. Geneva:
World Health Organization, 1996.
6.
World Health Organization (1999). World Health Report 1999. Geneva, World Health
Organization.
Murray CJL, Lopez, AD, eds (1996). The global burden of disease: a comprehensive
assessment of mortality and disability from diseases, injuries and risk factors in 1990 and
projected to 2020. Cambridge, Harvard University Press (Global Burden of disease and Injury
Series, Vol. 1).
Van de Water HP, Perenboom RJ, Boshuizen HC (1996). Policy relevance of the health
expectancy indicator: an inventory of European Union countries. Health Policy 36(2):117-29
Barendregt JJ, Bonneux L, van der Maas, PJ (1998). Health expectancy: from a population
health indicator to a tool for policy making. Journal of Aging & Health 10(2):242-258
Field MJ, Gold GM, eds. (1998). Summarizing Population Health: Directions for the
Development and Application of Population Metrics. Institute of Medicine, Washington, D.C.
National Academy Press.
Sanders BS (1964). Measuring community health levels. American Journal of Public Health,
54: 1063-1070.
7.
8.
9.
10.
11.
12.
Sullivan DF (1966). Conceptual problems in developing an index of health. US public Health
Service Publication Series No. 1000. Vital and Health Statistics Series 2. No. 17. National
Center for Health Statistics.
13.
Sullivan DF (1971). A single index of mortality and morbidity. HSMHA Health Reports, 86:
347-354.
14.
Piot M, Sundaresan TK (1967). A linear programme decision model for tuberculosis control.
Progress report on the first test-runs. Geneva, World health Organization, 1967,
WHO/TB/Tech.Information/67.55.
Fanshel S, Bush JW (1970). A health-status index and its application to health services
outcomes. Operations research, 18(6):1021-1066.
Ghana Health Assessment Project Team (1981). A quantitative method of assessing the health
impact of different diseases in less developed countries. International journal of epidemiology,
10(1):72-80.
15.
16.
17.
Preston SH (1993). Health indices as a guide to health sector planning: a demographic critique.
In Gribble JN, Preston SH, eds. The epidemiological transition. Policy and planning
implications for developing countries. Washington D.C., National Academic Press, 1993.
69
18.
Bone MR (1992). International efforts to measure health expectancy. Journal of Epidemiology
and Community Health, vol.46, pp.555-558.
19.
Mathers CD, Robine JM (1993). Health expectancy indicators: a review of the work of REVES
to date.. In: J.M. Robine, C.D. Mathers, M.R. Bone and I. Romieu (eds), Calculation of health
expectancies, harmonization, consensus achieved and future perspectives. Colloque INSERM
Vol. 226, John Libbey Eurotext and Les Editions INSERM, France.
Robine JM, Blanchet M, Dowd EJ (eds.) (1992). Health Expectancy. OPCS, HMSO, London.
20.
21.
22.
Robine JM, Mathers CD, Bone MR & Romieu I (eds) (1994). Calculation of health
expectancies, harmonization, consensus achieved and future perspectives. Colloque INSERM
Vol. 226, John Libbey Eurotext and Les Editions INSERM, France.
Mathers CD, McCallum J Robine JM (eds) (1994). Advances in health expectancies:
proceedings of the 7th meeting of the international network on health expectancy (REVES),
Australian Institute of Health and Welfare, Canberra, 1994.
23.
Robine JM, Mathers C, Brouard N (1996). Trends and differentials in disability-free life
expectancy: concepts, methods and findings. In: Caselli G, Lopez A (eds). Health and Mortality
Among Elderly Populations. Clarendon Press, Oxford, pp182-201.
24.
Crimmins, EM., Saito, Y, Ingengneri, D (1997). Trends in disability-free life expectancy in the
United States, 1970-90. Population and development review, 23(3): 555-572.
25.
Robine JM, Jagger C (1999). Developing Consistent Disability Measures to Address Public
Policy Needs for Older Populations. Background Paper prepared for OECD Meeting on
Implications of Disability for Ageing Populations: Monitoring Social Policy Challenges. Paris,
9-10 December 1999.
Saito Y, Crimmins EM, Hayward MD (1999). Health expectancy: an overview. NUPRI
Research paper Series No. 67, Nihon University Population research Institute, Tokyo.
OECD (1993). OECD Health Systems: Facts and Trends 1960-1991. OECD, Paris.
OECD (1999). Eco-santé. OECD Health Database 1999. OECD, Paris.
Romieu I and Robine JM (1994). World atlas of health expectancy calculations. In Mathers C,
McCallum J, Robine JM (eds.) Advances in health expectancies. Australian Institute of Health
and Welfare, Canberra.
Wilkins R, Adams O (1983). Health expectancy in Canada, late 1970’s: demographic, regional
and social dimensions. AJPH 73:1073-1080
Murray CJL, and Lopez A (1996b). Global Health Statistics. Cambridge, Harvard University
Press (Global Burden of disease and Injury Series, Vol. 2).
Murray CJL, and Lopez A (1996c). Evidence-based health policy—lessons from the Global
Burden of Disease Study. Science,, 274:740-743.
Murray CJL, and Lopez A (1997a). Mortality by cause for eight regions of the world: Global
Burden of Disease Study. Lancet, 349: 1269-1276.
Murray CJL, and Lopez A (1997b). Regional patterns of disability-free life expectancy and
disability-adjusted life expectancy: Global Burden of Disease Study. Lancet, 349: 1347-1352.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
Murray CJL, and Lopez A (1997c). Global mortality, disability, and the contribution of risk
factors: Global Burden of Disease Study. Lancet, 349: 1436-1442.
36.
Murray CJL, and Lopez A (1997d). Alternative projections of mortality and disability by cause
1990-2020: Global Burden of Disease Study. Lancet, 349: 1498-1504.
37.
Lozano R, Frenk J, Gonzalez MA (1994). El peso de la enfermedad en adultos mayores,
Mexico 1994. Salud Publica de Mexico, 38: 419-429.
38.
Lozano R, Murray CJL, Frenk J, and Bobadilla J (1995). Burden of Disease Assessment and
Health System Reform: Results of a study in Mexico. Journal for International Development,
7(3): 555-564.
70
39.
Fundación Mexicana Para La Salud (1995). Health and the Economy: Proposals for Progress in
the Mexican Health System. Mexico, Salud.
40.
República de Colombia Ministerio de Salud (1994). La Carga de la Enfermedad en Columbia.
Bogota, Ministerio de Salud, 1994.
Ruwaard D. and Kramers PGN (1998). Public health status and forecasts. The Hague: National
Institute of Public Health and Environmental Protection, 1998.
Bowie C, Beck S, Bevan G, Raftery J, Silvertion F, and Stevens J (1997). Estimating the
burden of disease in an English region. Journal of Public Health Medicine, 19:87-92.
Concha M, Aguilera, X, Albala C, et al (1996). Estudio ecarga de enfermedad informe final.
Estudio Prioridades de Inversio en salud Minsterio de Salud.
41.
42.
43.
44.
Murray CJL. Mahapatra P. Ashley R. Michaud C. George A. Horbon P. Akhavan D. et al.
(1997). The Health Sector in Mauritius: Resource Use, Intervention Cost and Options for
Efficiency Enhancement. Cambridge, Harvard Center for Population and Development Studies.
45.
Murray CJL. Michaud CM. McKenna MT. and Marks JS (1998). U. S. Patterns of Mortality
by County and Race: 1965-1994. Cambridge: Harvard Center for Population and Development
Studies and Centers for Disease Control, 1998.;
Mathers C, Vos T, Stevenson C (1999). The burden of disease and injury in Australia.
Australian Institute of Health and Welfare, Canberra: AIHW.
Murray CJL, Salomon J, Mathers CD, Lopez A, Lozano R (2000). Summary Measures of
Population Health. Geneva, World Health Organization (in preparation).
Murray CJL, Salomon J, Mathers C (1999). A critical review of summary measures of
population health. Geneva, World Health Organization, 2000 (GPE Discussion Paper No. 2).
Murray CJL, Lopez AD (2000). Progress and directions in refining the global burden of
disease approach: response to Williams. Health EconomicsDem 9: 69-82.
Dempsey M (1947). Decline in tuberculosis: the death rate fails to tell the entire story.
American review of tuberculosis, 56: 157-164.
Romeder JM. and McWhinnie JR (1977). Potential years of life lost between ages 1 and 70: an
indicator of premature mortality for health planning. International Journal of Epidemiology
6:143-151.
Hyder AA, Rotllanat G, Morrow RH (1998). Measuring the Burden of Disease: Healthy LifeYears. American Journal of Public Health, 88(2):196-202.
Wilkins R, Adams OB (1992). Quality-adjusted life expectancy: weighting of expected years in
each state of health. In Robine J-M, Blanchet M, Dowd JE (eds) Health expectancy, OPCS
studies on medical and population subjects no. 54, HMSO, London, 1992.
Wilkins R, Chen J, Ng E (1994). Changes in health expectancy in Canada from 1986 to 1991.
In: C.D. Mathers, J. McCallum, J.M. Robine (eds), Advances in health expectancies:
proceedings of the 7th meeting of the international network on health expectancy (REVES),
Australian Institute of Health and Welfare, Canberra 1994.
Mathers CD (1999a). Gains in health expectancy from the elimination of diseases among older
people. Disability and Rehabilitation 21(5-6): 211-221.
Wolfson MC (1996). Health-adjusted life expectancy. Health Reports 8(1): 41-45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
Stouthard M, Essink-Bot M, Bonsel G, Barendregt J & Kramers P (1997). Disability weights
for diseases in the Netherlands. Rotterdam: Department of Public Health, Erasmus University.
58.
Mathers CD, Robine JM, Wilkins R (1994). Health expectancy indicators: recommendations
for terminology.. In: C.D. Mathers, J. McCallum, J.M. Robine (eds), Advances in health
expectancies: proceedings of the 7th meeting of the international network on health expectancy
(REVES), Australian Institute of Health and Welfare, Canberra 1994.
71
59.
60.
61.
Mathers CD (1999b). Health expectancies: an overview and critical appraisal. Invited paper.
Global Conference on Summary Measures of Population Health, World Health Organisation,
Marrakech, 6-9 December, 1999.
Robine JM (1994). Disability-free life expectancy trends in France, international comparison.
In Mathers C, McCallum J, Robine JM (eds.) Advances in health expectancies. Australian
Institute of Health and Welfare, Canberra.
Mathers CD (1996). Trends In Health Expectancies In Australia 1981-1993. Journal of the
Australian Population Association 13(1): 1-16.
62.
Wolfson MC (1998). Measuring health—visions and practicalities. Joint ECE-WHO Meeting
on Health Statistics. Rome.
63.
Murray CJL, Frenk J (1999). A WHO Framework for Health System Performance Assessment.
Geneva, World Health Organization, 2000 (GPE Discussion Paper No.6).
Gakidou EE, Murray CJL, Frenk J (1999). A framework for measuring health inequality.
Discussion Paper. Global Program on Evidence for Health Policy, WHO, Geneva.
Gakidou EE, King G (2000). Using an extended beta-binomial model to estimate the
distribution of child mortality risk. Geneva, World Health Organization, 2000 (GPE
Discussion Paper No. 18).
Gakidou EE. Murray CJL (2000). Estimates of the distribution of child survival in 191
countries. Geneva, World Health Organization, 2000 (GPE Discussion Paper No. 19).
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
McDowell I, Newell C (1996). Measuring health. A guide to rating scales and questionnaires.
Second Edition. Oxford: Oxford University Press.
Krabbe PFM, Essink-Bot M, Bonsel GJ (1996). The comparability and reliability of five healthstate valuation methods. Social Science and Medicine 45:1641-1652.
Brazier J, Usherwood T, Harper R,Thomas K (1998). Deriving a preference-based single index
from the UK SF-36 Health Survey. Journal of Clinical Epidemiology 51(11): 1115-1128.
Nord E (1992). Methods for quality adjustment of life years, Social Science and Medicine
1992; 34: 559-569.
World Health Organization (1980). International Classification of Impairments, Disabilities,
and Handicaps. Geneva, World Health Organization, (Reprint 1993).
World Health Organization (1999). International Classification of Functioning and Disability Beta-2 Draft, Short Version, WHO, Geneva, July 1999.
Sadana R 2000. Measuring Reproductive Health: Conceptual and Operational Approaches,
Bulletin of the World Health Organization (forthcoming).
Van Agt HME. Essinck-Bot ML. Krabbe PFM. et al. (1994). Test-retest reliability of health
state valuations collected with the EuroQol questionnaire. Social Science and Medicine
39:1537-1544.
Gold MR, Siegel JE, Weinstein MC, & Russell LB (1996). Cost-effectiveness in health and
medicine. NY: Oxford University Press.
Ubel PA, Richardson J & Menzel P (1999). Societal value, the person trade-off, and the
dilemma of whose values to measure for cost-effectiveness analysis. Health Economics (in
press).
Ustun TB. Rehm J. Chatterji, S. Trotter R. Room R. Bickenbach J (1999). Are disability
valuations universal? Multiple-informant ranking of the disabling effects of different health
conditions in 14 countries, Lancet (in submission).
Sadana R 1998. A closer look at the WHO/World Bank Global Burden of Disease study's
methodologies: how do poor women's values in a developing country compare with
international public health experts? Paper presented at Eighth Annual Public Health Forum:
Reforming Health Sectors, London School of Hygiene & Tropical Medicine, 21-24 April.
72
79.
Mahapatra P (2000). Ph.D. thesis. Harvard School of Public Health. Cambridge, Massachusetts
(in preparation).
80.
Australian Institute of Health and Welfare (AIHW) (1999). Australia’s Welfare 1999.
Canberra: AIHW.
Eurostat (1999). European Community Household Panel Longitudinal Users’ Database. Waves
1, 2 and 3.Manual. European Commission.
Lopez AD, Salomon J. Ahmad O, Murray CJL (2000). Life tables for 191 countries: data,
methods and results. Geneva, World Health Organization, 2000 (GPE discussion Paper No. 9).
Murray CJL, Lopez AD, Ahmad O, Salomon J (2000). WHO system of model life tables.
Geneva. World Health Organization, 2000 (GPE Discussion Paper No. 8).
81.
82.
83.
84.
Salomon J. Gakidou EE, Murray CJL (2000). Methods for modelling the HIV/AIDS epidemic
in sub-Saharan Africa. Geneva, World Health Organization, 2000 (GPE discussion Paper No.
3).
85.
Salomon J. Murray CJL (2000). Compositional models for mortality by age, sex and cause.
Geneva, World Health Organization, 2000 (GPE Discussion Paper No. 11).
Lozano R, Murray CJL, Lopez AD, Satoh T (2000). Miscoding and misclassification of
ischaemic heart disease mortality. Geneva, World Health Organization, 2000 (GPE Discussion
Paper No. 12).
Boschi-Pinto C. Murray CJL, Lopez AD, Lozano R (2000). Cancer survival by site for 14
regions of the world. Geneva, World Health Organization, 2000 (GPE Discussion Paper No.
13).
Rasmussen N, Gudex C and Christensen S (1999). Survey Data on Disability, Eurostat
Working Papers, Population and Social Conditions 3/1999/E/n 20, European Commission,
Luxembourg
Salomon J. Murray CJL (2000). Methods for life expectancy and disability-adjusted life
expectancy uncertainty analysis. Geneva, World Health Organization, 2000 (GPE Discussion
Paper No. 10).
Woodward A, Mathers C, Tobias M (2000). The health status of New Zealanders and
Australians: different politics, different health? Paper prepared for the Health Inequalities
Research Consortium Conference “Social origins of health and well-being: from the planetary
to the molecular.” Canberra, July 2000.
86.
87.
88.
89.
90.
73
ANNEX TABLE
74
Annex Table A. Disability-adjusted life expectancy (DALE) and life expectancy (LE) at birth and age 60, by sex, WHO Member States, 1999
Uncertainty
Rank rangea Member State
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
1
2 - 10
2-7
2-8
3 - 11
3 - 10
5 - 11
4 - 11
2 - 15
6 - 15
5 - 17
8 - 14
8 - 15
11 - 17
11 - 18
11 - 17
12 - 18
15 - 20
18 - 23
19 - 24
18 - 24
20 - 24
19 - 25
22 - 27
23 - 28
20 - 33
25 - 30
26 - 31
26 - 31
27 - 32
26 - 31
30 - 34
31 - 36
32 - 34
33 - 36
34 - 45
36 - 46
36 - 42
38 - 44
37 - 45
Japan
Australia
France
Sweden
Spain
Italy
Greece
Switzerland
Monaco
Andorra
San Marino
Canada
Netherlands
United Kingdom
Norway
Belgium
Austria
Luxembourg
Iceland
Finland
Malta
Germany
Israel
United States of America
Cyprus
Dominica
Ireland
Denmark
Portugal
Singapore
New Zealand
Chile
Cuba
Slovenia
Czech Republic
Jamaica
Uruguay
Croatia
Argentina
Costa Rica
Total
population
at birth
74.5
73.2
73.1
73.0
72.8
72.7
72.5
72.5
72.4
72.3
72.3
72.0
72.0
71.7
71.7
71.6
71.6
71.1
70.8
70.5
70.5
70.4
70.4
70.0
69.8
69.8
69.6
69.4
69.3
69.3
69.2
68.6
68.4
68.4
68.0
67.3
67.0
67.0
66.7
66.7
At
birth
71.9
70.8
69.3
71.2
69.8
70.0
70.5
69.5
68.5
69.3
69.5
70.0
69.6
69.7
68.8
68.7
68.8
68.0
69.2
67.2
68.4
67.4
69.2
67.5
68.7
67.2
67.5
67.2
65.9
67.4
67.1
66.0
67.4
64.9
65.2
66.8
64.1
63.3
63.8
65.2
Disability-adjusted life expectancy (DALE) in years
Males
Females
Uncertainty At age Uncertainty
Uncertainty At age
a
a
a
range
60
range
At birth
range
60
71.6 - 72.3
17.5
17.3 - 18.1
77.2
76.9 - 78.0
21.6
70.5 - 71.3
16.8
16.6 - 17.3
75.5
75.2 - 76.2
20.2
69.0 - 69.7
16.8
16.5 - 17.4
76.9
76.5 - 77.8
21.7
70.9 - 71.8
16.8
16.5 - 17.3
74.9
74.4 - 75.7
19.6
69.1 - 70.6
16.8
16.4 - 17.6
75.7
75.3 - 76.6
20.1
69.7 - 70.5
16.2
16.0 - 16.8
75.4
75.0 - 76.2
19.9
70.2 - 70.9
16.9
16.6 - 17.3
74.6
74.2 - 75.2
18.8
69.0 - 70.2
16.0
15.7 - 16.7
75.5
75.0 - 76.5
20.6
67.5 - 69.6
16.4
15.9 - 17.2
76.3
75.6 - 77.3
21.5
68.6 - 70.2
16.3
15.9 - 17.0
75.2
74.6 - 76.2
20.0
68.6 - 70.5
15.7
15.3 - 16.5
75.0
74.4 - 76.0
19.6
69.7 - 70.5
16.0
15.8 - 16.6
74.0
73.6 - 74.9
18.9
69.3 - 70.1
15.4
15.3 - 16.0
74.4
74.0 - 75.3
19.7
69.4 - 70.1
15.7
15.5 - 16.2
73.7
73.5 - 74.4
18.6
68.5 - 69.3
15.1
15.0 - 15.7
74.6
74.2 - 75.3
19.7
68.4 - 69.2
15.8
15.6 - 16.4
74.6
74.2 - 75.3
19.6
68.4 - 69.4
15.2
15.0 - 15.8
74.4
74.1 - 75.1
18.7
67.6 - 68.7
15.8
15.2 - 16.8
74.2
73.7 - 75.2
19.7
68.6 - 70.1
14.9
14.2 - 15.9
72.3
71.7 - 73.4
17.0
66.9 - 67.7
14.5
14.2 - 15.0
73.7
73.4 - 74.4
18.5
67.9 - 69.2
14.8
14.5 - 15.6
72.5
72.0 - 73.4
17.3
67.1 - 67.9
14.3
14.1 - 14.9
73.5
73.2 - 74.1
18.5
68.9 - 69.7
15.6
15.3 - 16.3
71.6
71.2 - 72.4
16.9
67.0 - 68.1
15.0
14.7 - 15.7
72.6
72.2 - 73.3
18.4
68.2 - 69.4
15.9
15.6 - 16.6
70.9
70.4 - 71.7
17.3
66.2 - 68.2
15.0
14.3 - 15.6
72.3
71.0 - 73.4
17.9
67.0 - 68.2
13.9
13.6 - 14.6
71.7
71.2 - 72.5
16.6
66.8 - 67.9
14.2
13.9 - 14.8
71.5
71.2 - 72.2
17.2
65.6 - 66.6
14.0
13.7 - 14.6
72.7
72.4 - 73.4
17.7
66.9 - 68.2
14.4
14.1 - 15.2
71.2
70.7 - 72.2
16.8
66.8 - 67.6
14.4
14.1 - 15.0
71.2
70.8 - 72.0
17.0
65.2 - 67.0
14.3
13.6 - 15.3
71.3
70.9 - 72.2
17.8
66.8 - 68.1
15.4
14.9 - 16.1
69.4
68.9 - 70.3
16.1
64.6 - 65.4
12.7
12.6 - 13.4
71.9
71.5 - 72.6
16.8
64.9 - 65.7
12.7
12.6 - 13.2
70.8
70.5 - 71.5
16.4
65.5 - 68.0
18.9
18.1 - 19.7
67.9
66.5 - 69.3
18.2
63.1 - 65.0
15.3
14.8 - 15.8
69.9
68.8 - 71.0
18.3
63.1 - 63.8
11.4
11.3 - 11.9
70.6
70.3 - 71.3
16.0
63.5 - 64.3
14.7
14.4 - 15.3
69.6
69.2 - 70.3
18.1
64.6 - 66.0
14.2
13.9 - 15.0
68.1
67.5 - 69.1
16.6
75
Life expectancy
at birth (years)
Uncertainty
a
range
21.3 - 22.4
19.9 - 20.9
21.4 - 22.7
19.4 - 20.5
19.8 - 21.0
19.6 - 20.7
18.6 - 19.5
20.3 - 21.6
21.1 - 22.5
19.6 - 20.9
19.2 - 20.5
18.6 - 19.8
19.4 - 20.6
18.3 - 19.2
19.4 - 20.6
19.3 - 20.4
18.4 - 19.4
19.0 - 21.0
16.4 - 18.3
18.3 - 19.3
17.0 - 18.2
18.2 - 19.1
16.7 - 17.8
18.1 - 19.2
17.0 - 18.1
17.2 - 18.7
16.3 - 17.4
16.9 - 18.0
17.3 - 18.5
16.5 - 17.8
16.8 - 17.9
17.3 - 18.8
15.8 - 16.9
16.5 - 17.6
16.2 - 17.1
17.3 - 19.1
17.6 - 19.0
15.8 - 16.7
17.8 - 19.0
16.2 - 17.6
Males Females
77.6
84.3
76.8
82.2
74.9
83.6
77.1
81.9
75.3
82.1
75.4
82.1
75.5
80.5
75.6
83.0
74.7
83.6
75.4
82.2
75.3
82.0
76.2
81.9
75.0
81.1
74.7
79.7
75.1
82.1
74.5
81.3
74.4
80.4
74.5
81.4
76.1
80.4
73.4
80.7
75.7
80.8
73.7
80.1
76.2
79.9
73.8
79.6
74.8
78.8
74.1
80.3
73.3
78.3
72.9
78.1
72.0
79.5
75.1
80.8
74.0
79.4
73.4
79.9
73.5
77.4
71.6
79.5
71.5
78.3
75.2
77.4
69.4
77.2
70.5
77.8
70.6
77.8
72.3
77.1
Expected years Per cent of total
lost to disability life expectancy
b
at birth (DLE ) lost to disability
Males Females
5.7
7.1
6.0
6.7
5.6
6.7
5.9
7.0
5.5
6.4
5.4
6.7
5.0
5.9
6.1
7.5
6.2
7.3
6.1
7.0
5.8
7.0
6.2
7.8
5.4
6.7
5.0
6.0
6.3
7.6
5.8
6.7
5.6
6.0
6.5
7.2
6.8
8.1
6.2
7.0
7.3
8.3
6.3
6.6
7.1
8.3
6.3
7.0
6.1
7.9
6.8
8.0
5.8
6.6
5.7
6.6
6.1
6.8
7.7
9.6
6.8
8.1
7.4
8.6
6.2
8.0
6.7
7.6
6.3
7.5
8.4
9.5
6.4
7.9
6.0
6.6
6.8
8.2
9.0
10.8
Males Females
7.3
8.4
7.8
8.1
7.5
8.0
7.7
8.5
7.3
7.7
7.1
8.2
6.7
7.4
8.1
9.1
8.3
8.7
8.0
8.5
7.7
8.6
8.1
9.6
7.2
8.2
6.7
7.5
8.4
9.2
7.8
8.2
7.5
7.4
8.7
8.8
9.0
10.0
8.4
8.6
9.6
10.3
8.6
8.3
9.3
10.4
8.6
8.8
8.2
10.0
9.2
10.0
8.0
8.4
7.9
8.4
8.4
8.6
10.2
11.8
9.2
10.2
10.1
10.8
8.4
10.3
9.4
9.6
8.8
9.5
11.2
12.3
9.1
10.2
8.7
8.6
9.6
10.6
12.1
13.7
Annex Table A (continued). Disability-adjusted life expectancy (DALE) and life expectancy (LE) at birth and age 60, by sex, WHO Member States, 1999
Uncertainty
Rank range Member State
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
36 - 46 Armenia
39 - 45 Slovakia
36 - 54 Saint Vincent and the
Grenadines
39 - 50 Georgia
41 - 48 Poland
39 - 51 Yugoslavia
40 - 53 Panama
39 - 61 Antigua and Barbuda
40 - 63 Grenada
46 - 58 United Arab Emirates
46 - 58 Republic of Korea
49 - 57 Venezuela, Bolivarian Rep. of
47 - 59 Barbados
45 - 67 Saint Lucia
48 - 58 Mexico
49 - 64 Bosnia and Herzegovina
52 - 62 Trinidad and Tobago
52 - 66 Saudi Arabia
51 - 69 Brunei Darussalam
55 - 63 Bulgaria
50 - 64 Bahrain
57 - 66 Hungary
56 - 67 Lithuania
a
60 - 69 TFYR Macedonia
60 - 72 Azerbaijan
55 - 80 Qatar
61 - 78 Cook Islands
61 - 82 Kuwait
65 - 77 Estonia
68 - 76 Ukraine
64 - 82 Paraguay
62 - 86 Oman
67 - 82 Turkey
64 - 82 Colombia
65 - 83 Tonga
68 - 83 Sri Lanka
67 - 85 Suriname
66 - 88 Mauritius
68 - 86 Dominican Republic
74 - 84 Romania
Total
population At
at birth
birth
66.7
65.0
66.6
63.5
66.4
66.3
66.2
66.1
66.0
65.8
65.5
65.4
65.0
65.0
65.0
65.0
65.0
64.9
64.6
64.5
64.4
64.4
64.4
64.1
64.1
63.7
63.7
63.5
63.4
63.2
63.1
63.0
63.0
63.0
62.9
62.9
62.9
62.8
62.7
62.7
62.5
62.3
65.0
63.1
62.3
64.2
64.9
63.4
62.4
65.0
62.3
62.9
62.4
62.4
62.4
63.4
62.8
65.1
63.4
61.2
63.9
60.4
60.6
61.8
60.6
64.2
62.2
63.0
58.1
58.5
60.7
61.8
64.0
60.3
61.4
59.3
60.2
59.0
62.1
58.8
Life expectancy
at birth (years)
Expected years Per cent of total
lost to disability life expectancy
at birth (DLE)
lost to disability
Disability-adjusted life expectancy (DALE) in years
Males
Females
Uncertainty At age Uncertainty
Uncertainty At age
range
60
range
At birth
range
60
64.4 - 65.9
14.5
14.2 - 15.5
68.3
67.6 - 69.3
15.5
63.2 - 64.0
12.7
12.6 - 13.1
69.7
69.4 - 70.3
16.0
Uncertainty
range
15.1 - 16.5
15.9 - 16.7
Males Females
74.2
78.9
68.9
76.7
Males Females
7.3
8.8
5.4
7.0
Males Females
10.1
11.4
7.8
9.1
63.8 - 66.2
62.2 - 64.0
61.6 - 63.0
63.1 - 65.3
63.6 - 66.1
62.0 - 64.6
61.1 - 63.6
64.0 - 65.9
61.6 - 63.1
62.4 - 63.6
61.2 - 63.8
61.1 - 63.6
61.6 - 63.3
62.3 - 64.5
62.2 - 63.5
64.3 - 65.9
63.0 - 64.5
60.8 - 61.6
63.0 - 64.8
59.6 - 61.2
59.7 - 61.6
61.2 - 62.6
59.9 - 61.4
63.4 - 65.1
61.0 - 63.4
61.6 - 64.3
57.3 - 59.0
58.1 - 59.0
59.2 - 62.0
60.8 - 62.8
62.9 - 65.0
59.2 - 61.5
60.2 - 62.6
58.3 - 60.3
58.9 - 61.4
57.9 - 60.2
60.7 - 63.6
58.4 - 59.2
15.9 - 17.5
16.2 - 17.4
16.4 - 17.3
16.8 - 18.1
16.5 - 18.3
15.9 - 17.6
16.2 - 17.7
11.8 - 13.5
14.5 - 15.8
15.2 - 16.7
15.8 - 17.9
15.2 - 16.5
16.4 - 17.9
14.5 - 15.9
13.7 - 14.6
12.2 - 13.5
11.9 - 14.4
14.8 - 15.7
11.8 - 13.3
15.3 - 16.3
15.9 - 17.0
13.2 - 14.3
15.4 - 16.7
8.9 - 11.3
13.0 - 14.5
10.6 - 12.9
15.5 - 16.6
15.3 - 16.1
15.1 - 17.0
11.2 - 12.8
14.6 - 15.9
14.6 - 16.2
12.6 - 14.1
15.4 - 16.7
14.6 - 16.3
13.2 - 14.4
15.2 - 16.8
14.3 - 15.3
71.8
69.4
67.9
71.8
72.6
71.4
69.1
72.2
68.8
70.9
72.7
68.9
71.0
71.2
68.7
71.0
74.3
67.4
70.7
66.3
67.0
69.8
67.8
71.5
69.3
71.9
64.4
64.4
69.6
70.4
69.7
68.1
68.3
65.8
68.2
66.7
71.4
65.1
6.8
6.3
5.6
7.6
7.8
8.0
6.7
7.3
6.4
8.1
10.3
6.5
8.6
7.9
5.9
5.8
10.9
6.3
6.8
5.9
6.4
8.0
7.2
7.4
7.0
8.9
6.3
5.8
8.9
8.6
5.7
7.8
6.8
6.5
7.9
7.7
9.2
6.4
9.5
9.1
8.2
10.6
10.7
11.2
9.7
10.0
9.3
11.4
14.2
9.4
12.2
11.0
8.5
8.2
14.6
9.3
9.7
9.0
9.5
11.4
10.6
10.3
10.1
12.4
9.8
9.1
12.9
12.2
8.2
11.5
10.0
9.9
11.6
11.6
12.9
9.8
15.9
13.8
12.5
15.1
17.3
14.4
14.1
11.7
12.1
13.4
14.5
14.1
14.7
13.3
12.0
12.7
12.4
12.2
11.6
11.7
13.4
11.7
12.7
10.8
12.2
11.1
11.2
11.5
14.2
10.6
16.2
13.5
11.5
12.7
14.4
10.2
17.1
12.0
15.2 - 16.7
13.5 - 14.6
12.1 - 13.1
14.4 - 15.7
16.4 - 18.1
13.7 - 15.2
13.5 - 14.8
10.9 - 12.5
11.6 - 12.7
13.1 - 14.2
13.8 - 15.8
13.5 - 14.8
14.4 - 15.6
12.7 - 14.1
11.8 - 12.6
12.1 - 13.4
12.0 - 13.6
12.0 - 12.7
11.0 - 12.3
11.4 - 12.3
13.1 - 14.2
11.4 - 12.5
12.4 - 13.5
10.1 - 11.4
11.5 - 12.8
10.2 - 12.0
10.9 - 11.9
11.4 - 12.0
13.3 - 15.1
9.9 - 11.3
15.5 - 16.8
12.7 - 14.2
10.9 - 12.2
12.1 - 13.4
13.8 - 15.2
9.8 - 11.2
16.1 - 18.2
11.8 - 12.6
67.8
69.4
70.1
68.1
67.2
68.3
68.5
65.8
67.7
67.1
67.6
67.6
67.6
66.4
66.4
64.0
65.4
67.7
64.9
67.9
67.5
65.6
66.7
62.8
64.5
63.4
68.1
67.5
65.3
64.1
61.8
65.5
64.3
66.3
65.2
66.3
62.9
65.8
76
66.4 - 69.0
68.7 - 70.3
69.7 - 70.7
66.9 - 69.2
65.9 - 68.5
66.9 - 69.6
67.2 - 69.7
64.6 - 67.0
66.7 - 68.7
66.5 - 68.1
66.9 - 68.7
66.4 - 68.7
67.1 - 68.5
65.2 - 67.5
65.9 - 67.1
62.6 - 65.2
64.7 - 66.9
67.4 - 68.3
63.8 - 66.0
67.5 - 68.5
66.9 - 68.3
65.1 - 66.5
66.1 - 67.7
61.2 - 64.3
63.1 - 65.9
61.6 - 65.1
67.4 - 69.0
67.1 - 68.0
63.9 - 66.7
62.8 - 65.3
60.6 - 63.0
64.2 - 66.7
62.9 - 65.6
65.2 - 67.4
64.0 - 66.5
65.7 - 67.3
61.8 - 63.7
65.4 - 66.5
16.7
16.6
16.6
17.5
17.4
16.8
16.9
12.6
15.2
15.7
16.6
15.8
16.8
15.3
13.9
12.8
12.6
15.1
12.6
15.5
16.2
13.5
15.7
10.2
13.7
11.8
15.8
15.5
16.0
12.1
15.2
15.4
13.3
16.0
15.5
13.5
16.1
14.6
75.2
76.7
76.6
76.4
75.8
76.8
75.9
75.6
76.0
76.2
77.8
74.9
77.2
75.0
73.4
72.6
79.5
74.8
73.6
75.1
77.9
74.1
75.3
74.6
73.3
75.2
75.3
74.2
74.1
73.8
69.9
74.1
72.9
73.4
73.6
74.1
72.8
73.5
7.4
7.3
6.5
8.2
8.6
8.5
7.4
9.8
8.3
9.0
10.2
7.3
9.6
8.6
7.0
8.7
14.3
7.1
8.7
7.2
10.4
8.5
8.6
11.8
8.8
11.9
7.2
6.9
8.8
9.7
8.1
8.6
8.6
7.1
8.3
7.7
9.9
7.6
9.8
9.5
8.5
10.8
11.4
11.1
9.7
13.0
10.9
11.8
13.1
9.8
12.4
11.5
9.5
12.0
18.0
9.4
11.8
9.6
13.3
11.5
11.4
15.8
12.0
15.8
9.5
9.3
11.8
13.1
11.6
11.6
11.8
9.7
11.3
10.4
13.6
10.4
Annex Table A (continued). Disability-adjusted life expectancy (DALE) and life expectancy (LE) at birth and age 60, by sex, WHO Member States, 1999
Total
population
at birth
China
62.3
Latvia
62.2
Belarus
61.7
Algeria
61.6
Niue
61.6
Saint Kitts and Nevis
61.6
El Salvador
61.5
Republic of Moldova
61.5
Malaysia
61.4
Tunisia
61.4
Russian Federation
61.3
Honduras
61.1
Ecuador
61.0
Belize
60.9
Lebanon
60.6
Iran, Islamic Republic of
60.5
Samoa
60.5
Guyana
60.2
Thailand
60.2
Uzbekistan
60.2
Jordan
60.0
Albania
60.0
Indonesia
59.7
59.6
Micronesia, Federated States of
Peru
59.4
Fiji
59.4
Libyan Arab Jamahiriya
59.3
Seychelles
59.3
Bahamas
59.1
Morocco
59.1
Brazil
59.1
Palau
59.0
Philippines
58.9
Syrian Arab Republic
58.8
Egypt
58.5
Viet Nam
58.2
Nicaragua
58.1
Cape Verde
57.6
Tuvalu
57.4
Tajikistan
57.3
Uncertainty
Rank range Member State
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
72 - 83
72 - 88
79 - 90
79 - 95
70 - 101
74 - 100
79 - 97
82 - 91
81 - 96
82 - 96
84 - 95
83 - 100
85 - 100
82 - 104
90 - 103
89 - 104
87 - 103
90 - 109
92 - 107
94 - 104
95 - 108
94 - 107
96 - 112
95 - 114
101 - 114
95 - 116
100 - 114
102 - 112
99 - 117
101 - 114
102 - 115
103 - 116
104 - 116
105 - 117
107 - 118
110 - 119
110 - 119
110 - 121
115 - 122
116 - 123
At
birth
61.2
57.1
56.2
62.5
61.0
58.7
58.6
58.5
61.3
62.0
56.1
60.0
59.9
58.5
61.2
61.3
58.7
57.1
58.4
58.0
60.7
56.5
58.8
58.7
58.0
57.7
59.7
56.4
56.7
58.7
55.2
57.4
57.1
58.8
58.6
56.7
56.4
54.6
57.1
55.1
Disability-adjusted life expectancy (DALE) in years
Males
Females
Uncertainty At age Uncertainty
Uncertainty At age
range
60
range
At birth
range
60
60.7 - 62.0
11.6
11.3 - 12.3
63.3
62.8 - 64.2
13.5
55.9 - 58.2
11.4
11.0 - 12.2
67.2
66.4 - 68.2
15.9
55.4 - 57.1
10.1
9.8 - 10.8
67.2
66.7 - 68.0
15.1
61.4 - 63.5
12.9
12.3 - 13.6
60.7
59.4 - 62.0
12.0
59.2 - 62.6
12.2
11.7 - 13.3
62.2
60.4 - 63.8
13.2
57.4 - 59.9
12.8
12.2 - 13.3
64.4
63.2 - 65.6
14.3
57.4 - 59.7
13.9
13.1 - 14.6
64.5
63.2 - 65.7
15.8
58.0 - 59.0
10.7
10.6 - 11.2
64.5
64.0 - 65.2
13.0
60.2 - 62.1
9.7
9.2 - 10.2
61.6
60.5 - 62.7
9.7
61.2 - 62.9
11.2
10.8 - 11.7
60.7
59.7 - 61.8
10.3
55.4 - 56.9
10.5
10.3 - 11.2
66.4
65.8 - 67.2
14.9
58.8 - 61.2
15.0
14.2 - 15.7
62.3
61.1 - 63.5
14.4
58.9 - 60.9
12.6
11.9 - 13.2
62.1
61.1 - 63.3
12.9
56.9 - 60.1
13.6
12.7 - 14.4
63.3
61.5 - 65.0
15.2
60.2 - 62.2
10.1
9.6 - 10.6
60.1
58.8 - 61.2
9.2
60.2 - 62.3
11.9
11.3 - 12.5
59.8
58.6 - 61.1
10.9
57.5 - 59.8
9.5
9.0 - 10.1
62.3
60.9 - 63.6
12.3
55.8 - 58.6
15.4
14.5 - 16.4
63.3
61.9 - 64.7
16.8
57.1 - 59.6
13.7
12.9 - 14.5
62.1
60.9 - 63.3
13.9
57.4 - 58.8
11.5
11.3 - 12.2
62.3
61.6 - 63.1
13.4
59.8 - 61.5
9.5
9.1 - 10.0
59.3
58.2 - 60.4
8.9
55.8 - 57.4
10.1
9.9 - 10.9
63.4
62.7 - 64.4
13.9
57.5 - 60.1
16.3
15.3 - 17.2
60.6
59.3 - 61.8
15.8
57.2 - 60.0
11.1
10.4 - 11.8
60.6
59.0 - 62.0
11.5
56.9 - 59.0
12.3
11.7 - 13.0
60.8
59.6 - 62.0
13.1
56.1 - 59.1
8.3
8.0 - 9.1
61.1
59.8 - 62.3
9.8
58.7 - 60.7
9.7
9.2 - 10.2
58.9
57.6 - 60.2
9.3
55.7 - 57.3
8.6
8.3 - 9.4
62.1
61.5 - 63.0
11.7
55.1 - 58.1
11.3
10.5 - 12.1
61.6
59.9 - 63.4
13.0
57.9 - 59.6
11.5
11.0 - 12.0
59.4
58.4 - 60.4
11.4
54.4 - 56.1
11.8
11.5 - 12.7
62.9
62.2 - 63.9
14.8
56.1 - 58.5
8.0
7.5 - 8.5
60.7
59.2 - 61.9
9.7
56.0 - 58.1
10.3
9.7 - 10.9
60.7
59.4 - 61.9
12.4
57.7 - 59.9
9.7
9.2 - 10.2
58.9
57.6 - 60.2
10.0
57.7 - 59.5
11.8
11.2 - 12.2
58.3
57.1 - 59.6
11.7
55.6 - 57.9
9.7
9.1 - 10.4
59.6
58.4 - 60.9
10.8
55.3 - 57.4
11.1
10.4 - 11.8
59.9
58.7 - 61.1
12.5
53.0 - 56.2
11.4
10.6 - 12.3
60.6
58.8 - 62.4
15.3
55.7 - 58.3
10.3
9.7 - 10.9
57.6
56.2 - 58.8
9.4
53.5 - 56.5
12.3
11.4 - 13.2
59.4
57.9 - 60.9
15.6
77
Life expectancy
at birth (years)
Uncertainty
range
13.2 - 14.4
15.5 - 16.7
14.8 - 15.9
11.3 - 12.6
12.7 - 15.2
13.7 - 15.0
14.9 - 16.6
12.8 - 13.7
9.1 - 10.3
9.8 - 10.8
14.6 - 15.7
13.5 - 15.1
12.3 - 13.7
14.3 - 16.2
8.7 - 9.7
10.2 - 11.6
11.7 - 13.0
15.8 - 17.8
13.1 - 14.7
13.1 - 14.3
8.5 - 9.4
13.6 - 14.7
15.0 - 16.6
10.7 - 12.4
12.3 - 13.9
9.5 - 10.8
8.7 - 10.0
11.4 - 12.4
12.0 - 14.0
10.8 - 12.0
14.4 - 15.8
9.1 - 10.4
11.6 - 13.1
9.4 - 10.6
11.1 - 12.4
10.1 - 11.5
11.7 - 13.2
14.2 - 16.4
8.8 - 10.0
14.7 - 16.4
Males Females
68.1
71.3
63.6
74.6
62.4
74.6
68.2
68.8
68.3
70.9
65.0
71.2
66.9
73.0
64.8
71.9
67.6
69.9
67.0
67.9
62.7
74.0
68.2
70.8
67.4
70.3
69.6
75.0
66.2
67.3
66.8
67.9
65.4
70.7
65.6
72.2
66.0
70.4
65.8
71.2
66.3
67.5
65.1
72.7
66.6
69.0
66.4
70.1
65.6
69.1
64.0
69.2
65.0
67.0
64.9
70.6
67.0
73.6
65.2
66.8
63.7
71.7
64.5
69.7
64.1
69.3
64.6
67.1
64.2
65.8
64.7
68.8
64.8
68.8
64.2
71.8
63.9
65.5
65.1
70.1
Expected years Per cent of total
lost to disability life expectancy
at birth (DLE)
lost to disability
Males Females
6.9
8.0
6.5
7.4
6.2
7.3
5.7
8.1
7.3
8.7
6.3
6.8
8.3
8.5
6.3
7.4
6.3
8.3
5.0
7.2
6.6
7.6
8.2
8.5
7.5
8.2
11.1
11.6
5.1
7.2
5.5
8.1
6.7
8.4
8.4
8.9
7.6
8.3
7.7
8.9
5.6
8.2
8.6
9.3
7.8
8.4
7.8
9.5
7.6
8.2
6.3
8.1
5.3
8.1
8.4
8.4
10.3
12.0
6.4
7.4
8.5
8.8
7.1
9.0
7.1
8.7
5.8
8.2
5.6
7.5
8.0
9.2
8.4
8.9
9.6
11.2
6.8
7.9
10.1
10.6
Males Females
10.2
11.2
10.2
9.9
9.9
9.9
8.4
11.7
10.7
12.2
9.7
9.5
12.4
11.6
9.7
10.3
9.4
11.9
7.4
10.6
10.5
10.3
12.0
12.0
11.1
11.6
15.9
15.5
7.7
10.7
8.2
11.9
10.2
11.9
12.9
12.3
11.6
11.8
11.7
12.6
8.4
12.1
13.3
12.8
11.7
12.2
11.7
13.5
11.6
11.9
9.8
11.7
8.2
12.1
13.0
11.9
15.4
16.3
9.8
11.0
13.3
12.3
11.0
13.0
11.0
12.5
8.9
12.3
8.8
11.4
12.3
13.3
13.0
13.0
15.0
15.5
10.6
12.1
15.5
15.2
Annex Table A (continued). Disability-adjusted life expectancy (DALE) and life expectancy (LE) at birth and age 60, by sex, WHO Member States, 1999
Uncertainty
Rank range Member State
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
117 - 125
120 - 124
119 - 125
122 - 127
123 - 129
123 - 129
123 - 133
125 - 133
126 - 132
127 - 134
127 - 135
129 - 135
129 - 136
130 - 136
130 - 138
133 - 138
134 - 139
138
139
140
141
142
143
144
145
146
147
135 - 139
136 - 139
140 - 142
140 - 142
140 - 142
142 - 144
143 - 145
144 - 147
145 - 147
145 - 150
148
149
150
151
152
153
154
155
146 - 150
147 - 151
148 - 152
149 - 153
150 - 153
149 - 156
153 - 157
153 - 156
Marshall Islands
Kazakhstan
Kyrgyzstan
Pakistan
Kiribati
Iraq
Solomon Islands
Turkmenistan
Guatemala
Maldives
Mongolia
Sao Tome and Principe
Bolivia
India
Vanuatu
Nauru
Democratic People's
Republic of Korea
Bhutan
Myanmar
Bangladesh
Yemen
Nepal
Gambia
Gabon
Papua New Guinea
Comoros
Lao People's Democratic
Republic
Cambodia
Ghana
Congo
Senegal
Equatorial Guinea
Haiti
Sudan
Côte d'Ivoire
Life expectancy
at birth (years)
Expected years Per cent of total
lost to disability life expectancy
at birth (DLE)
lost to disability
Total
population
at birth
56.8
56.4
56.3
55.9
55.3
55.3
54.9
54.3
54.3
53.9
53.8
53.5
53.3
53.2
52.8
52.5
At
birth
56.0
51.5
53.4
55.0
53.9
55.4
54.5
51.9
52.1
54.4
51.3
52.1
52.5
52.8
51.3
49.8
Disability-adjusted life expectancy (DALE) in years
Males
Females
Uncertainty At age Uncertainty
Uncertainty At age
range
60
range
At birth
range
60
54.4 - 57.4
10.7
10.0 - 11.4
57.6
55.9 - 59.0
11.1
50.9 - 52.2
8.8
8.7 - 9.5
61.2
60.8 - 62.0
13.1
52.6 - 54.2
9.6
9.4 - 10.4
59.1
58.3 - 60.1
12.4
53.8 - 56.3
11.3
10.5 - 12.1
56.8
54.6 - 57.9
12.6
52.4 - 55.3
9.4
8.7 - 10.1
56.6
55.0 - 58.0
11.0
54.4 - 56.4
9.2
8.7 - 9.8
55.1
53.9 - 56.2
8.2
53.0 - 55.8
8.8
8.2 - 9.5
55.3
53.7 - 56.7
9.2
50.6 - 53.3
9.0
8.7 - 10.0
56.7
55.3 - 58.0
10.9
51.1 - 53.1
9.1
8.6 - 9.8
56.4
55.4 - 57.5
10.1
53.0 - 55.9
12.1
11.3 - 13.0
53.3
51.8 - 54.7
11.5
49.7 - 52.7
11.8
11.0 - 14.4
56.3
54.7 - 57.7
14.3
51.1 - 53.3
11.4
11.1 - 12.5
54.8
53.8 - 55.8
11.7
51.3 - 53.7
11.6
10.9 - 12.3
54.1
52.8 - 55.2
11.2
52.1 - 53.5
10.6
10.4 - 11.3
53.5
52.8 - 54.3
12.1
49.8 - 52.7
8.0
7.4 - 8.6
54.4
52.8 - 55.8
9.2
48.7 - 50.9
3.6
3.1 - 4.0
55.1
53.8 - 56.2
5.9
52.3
51.8
51.6
49.9
49.7
49.5
48.3
47.8
47.0
46.8
51.4
51.4
51.4
50.1
49.7
49.4
47.2
46.6
45.5
46.1
49.8 - 53.1
50.0 - 52.7
50.0 - 52.6
48.7 - 51.3
48.1 - 51.1
48.1 - 50.7
46.3 - 48.2
45.4 - 47.6
44.3 - 46.8
45.1 - 47.1
9.6
11.4
12.5
9.9
8.5
10.3
9.9
10.3
8.2
8.9
8.7 - 10.6
10.7 - 12.0
11.7 - 13.3
9.2 - 10.5
7.9 - 9.2
9.6 - 10.9
9.3 - 10.6
9.8 - 10.8
7.5 - 8.9
8.3 - 9.6
53.1
52.2
51.9
49.8
49.7
49.5
49.4
49.0
48.5
47.5
51.3 - 55.0
50.7 - 53.6
50.5 - 53.2
48.3 - 51.2
48.2 - 51.1
48.2 - 50.9
48.4 - 50.4
47.8 - 50.1
47.1 - 49.8
46.5 - 48.5
11.6
12.6
12.3
10.5
8.2
10.0
11.7
12.3
8.7
9.8
10.7 - 12.6
12.0 - 13.3
11.6 - 12.9
9.8 - 11.1
7.5 - 8.8
9.4 - 10.7
11.2 - 12.4
11.8 - 12.8
8.0 - 9.4
9.4 - 10.4
58.0
59.6
58.4
57.5
57.3
57.3
56.0
54.6
53.4
56.0
60.7
60.8
59.2
58.1
58.0
57.8
58.9
57.5
56.6
58.1
6.6
8.2
7.1
7.4
7.6
7.9
8.8
8.0
7.8
9.9
7.6
8.7
7.4
8.3
8.3
8.3
9.5
8.5
8.1
10.6
11.3
13.8
12.1
12.9
13.2
13.7
15.7
14.6
14.7
17.7
12.5
14.2
12.4
14.3
14.3
14.3
16.1
14.8
14.3
18.3
46.1
45.7
45.5
45.1
44.6
44.1
43.8
43.0
42.8
45.0
43.9
45.0
44.3
43.5
42.8
42.4
42.6
42.2
43.5 - 46.5
42.6 - 45.1
43.8 - 46.2
43.1 - 45.5
42.5 - 44.5
41.7 - 43.9
41.0 - 43.6
41.2 - 43.7
41.2 - 43.3
8.9
7.4
9.9
10.7
8.8
9.4
7.4
5.6
11.9
8.0 - 9.7
6.6 - 8.2
9.3 - 10.6
10.0 - 11.3
8.1 - 9.5
8.9 - 10
6.8 - 8.0
5.1 - 6.0
11.5 - 12.5
47.1
47.5
46.0
45.9
45.6
45.4
45.2
43.5
43.3
45.5 - 48.6
46.1 - 48.9
44.8 - 47.2
44.6 - 47.1
44.6 - 46.7
44.4 - 46.6
43.7 - 46.7
42.1 - 44.6
42.3 - 44.4
8.8
9.3
10.2
12.8
11.3
11.0
8.0
6.0
12.7
7.9 - 9.6
8.7 - 10.0
9.6 - 10.8
12.2 - 13.4
10.6 - 11.9
10.5 - 11.7
7.3 - 8.7
5.6 - 6.5
12.2 - 13.2
54.0
52.2
54.2
53.6
53.5
51.4
50.5
53.1
47.3
56.6
55.4
55.6
55.2
56.2
55.4
55.0
54.7
48.3
9.0
8.3
9.2
9.3
10.0
8.6
8.2
10.5
5.1
9.5
7.9
9.6
9.3
10.6
9.9
9.8
11.2
5.0
16.6
15.8
16.9
17.4
18.7
16.7
16.2
19.8
10.8
16.7
14.2
17.2
16.9
18.8
17.9
17.8
20.5
10.3
78
Uncertainty
range
10.3 - 12.0
12.8 - 13.9
12.1 - 13.3
11.9 - 13.2
10.3 - 11.7
7.6 - 8.8
8.6 - 9.9
10.6 - 11.8
9.5 - 10.7
10.8 - 12.2
13.4 - 15.1
11.4 - 12.6
10.6 - 11.9
11.8 - 12.8
8.5 - 9.8
5.4 - 6.4
Males Females
64.0
67.1
58.8
69.9
61.6
69.0
62.6
64.9
61.4
65.5
62.0
63.4
62.0
64.0
61.0
65.3
60.2
64.7
63.3
62.6
58.9
64.8
62.1
64.9
60.7
62.2
59.6
61.2
58.7
63.0
56.4
63.3
Males Females
7.9
9.5
7.2
8.7
8.2
9.9
7.6
8.2
7.4
8.9
6.2
7.7
7.5
8.7
9.1
8.6
8.1
8.3
8.9
9.3
7.7
8.5
10.0
10.1
8.3
8.1
6.8
7.7
7.4
8.6
6.6
8.1
Males Females
12.4
14.2
12.3
12.4
13.3
14.3
12.1
12.6
12.1
13.6
10.0
12.2
12.2
13.7
14.9
13.2
13.4
12.8
14.0
14.9
13.0
13.1
16.1
15.5
13.6
13.1
11.3
12.5
12.7
13.7
11.6
12.9
Annex Table A (continued). Disability-adjusted life expectancy (DALE) and life expectancy (LE) at birth and age 60, by sex, WHO Member States, 1999
Disability-adjusted life expectancy (DALE) in years
Rank
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
Uncertainty
Member State
range
154 - 158
154 - 158
157 - 159
158 - 161
158 - 161
160 - 163
160 - 164
162 - 169
162 - 170
163 - 170
163 - 170
164 - 170
163 - 173
164 - 171
166 - 174
167 - 175
169 - 176
169 - 178
171 - 177
172 - 178
172 - 178
172 - 180
173 - 179
176 - 182
178 - 183
179 - 183
180 - 185
181 - 186
181 - 187
181 - 187
183 - 187
185 - 187
188 - 189
188 - 190
189 - 190
191
Cameroon
Benin
Mauritania
Togo
South Africa
Chad
Kenya
Nigeria
Swaziland
Angola
Djibouti
Guinea
Afghanistan
Eritrea
Guinea-Bissau
Lesotho
Madagascar
Somalia
Democratic Republic of the
Congo
Central African Republic
United Republic of Tanzania
Namibia
Burkina Faso
Burundi
Mozambique
Liberia
Ethiopia
Mali
Zimbabwe
Rwanda
Uganda
Botswana
Zambia
Malawi
Niger
Sierra Leone
Total
population
at birth
42.2
42.2
41.4
40.7
39.8
39.4
39.3
38.3
38.1
38.0
37.9
37.8
37.7
37.7
37.2
36.9
36.6
36.4
At
birth
41.5
41.9
40.2
40.0
38.6
38.6
39.0
38.1
37.8
37.0
37.7
37.0
36.7
38.5
36.8
36.6
36.5
35.9
Males
Uncertainty At age
range
60
40.4 - 42.5
9.6
40.9 - 42.9
9.6
39.2 - 41.2
9.2
38.8 - 41.3
9.5
37.7 - 39.5
6.8
37.2 - 39.8
9.2
37.9 - 40.2
9.2
36.9 - 39.2
8.7
36.5 - 39.0
8.1
35.7 - 38.1
8.9
36.2 - 38.8
6.9
36.1 - 38.0
8.5
34.9 - 38.5
7.9
37.6 - 39.5
8.2
35.6 - 37.9
9.1
35.3 - 38.0
9.9
35.5 - 37.4
6.7
34.4 - 37.2
6.1
36.3
36.0
36.0
35.6
35.5
34.6
34.4
34.0
33.5
33.1
32.9
32.8
32.7
32.3
30.3
29.4
29.1
25.9
36.4
35.6
35.9
35.8
35.3
34.6
33.7
33.8
33.5
32.6
33.4
32.9
32.9
32.3
30.0
29.3
28.1
25.8
35.5 - 37.3
34.6 - 36.7
35.1 - 36.8
34.3 - 37.4
34.1 - 36.6
33.0 - 36.2
32.3 - 35.3
32.7 - 34.9
32.5 - 34.5
31.6 - 33.7
32.3 - 34.5
31.6 - 34.3
32.1 - 33.9
31.7 - 32.9
28.9 - 30.9
28.3 - 30.2
27.1 - 29.0
24.5 - 26.8
7.3
8.8
7.8
9.8
7.9
7.6
8.3
7.3
7.5
7.7
8.8
6.9
6.2
6.1
7.6
6.8
6.6
6.0
Uncertainty
range
9.0 - 10.2
9.0 - 10.3
8.6 - 9.9
8.8 - 10.1
6.4 - 7.3
8.6 - 9.9
8.6 - 13.4
8.1 - 9.4
7.7 - 8.6
8.3 - 9.6
6.4 - 7.4
7.9 - 9.2
7.0 - 8.8
7.6 - 8.7
8.5 - 13.4
9.3 - 10.4
6.1 - 7.2
5.6 - 6.5
At birth
43.0
42.6
42.5
41.4
41.0
40.2
39.6
38.4
38.4
38.9
38.1
38.5
38.7
36.9
37.5
37.2
36.8
36.9
6.8 - 7.9
8.3 - 9.3
7.2 - 8.4
9.3 - 10.4
7.3 - 8.5
6.9 - 8.3
7.7 - 8.9
6.7 - 8.0
7.0 - 8.1
7.1 - 8.3
8.3 - 9.4
6.2 - 7.6
5.6 - 6.9
5.6 - 6.6
6.9 - 8.3
6.2 - 7.5
5.8 - 7.4
5.4 - 6.7
36.2
36.5
36.1
35.4
35.7
34.6
35.1
34.2
33.5
33.5
32.4
32.7
32.5
32.2
30.7
29.4
30.1
26.0
Life expectancy
at birth (years)
Females
Uncertainty At age
range
60
41.8 - 44.2
11.9
41.5 - 43.6
10.6
41.5 - 43.5
11.0
40.1 - 42.6
11.0
39.9 - 42.1
9.3
38.8 - 41.5
10.6
38.4 - 41.0
12.0
37.1 - 39.6
10.1
36.9 - 39.9
9.5
37.7 - 40.0
10.8
36.6 - 39.3
7.9
37.5 - 39.5
9.6
36.9 - 40.5
7.9
35.9 - 37.9
7.9
36.4 - 38.6
10.0
35.7 - 38.7
11.3
35.7 - 37.7
6.6
35.3 - 38.1
7.5
Uncertainty
range
11.3 - 12.5
9.9 - 11.2
10.3 - 11.7
10.5 - 11.6
8.9 - 9.8
10.0 - 11.2
11.5 - 12.5
9.5 - 10.7
9.0 - 10.0
10.1 - 11.4
7.6 - 8.2
9.0 - 10.3
7.2 - 8.7
7.4 - 8.4
9.4 - 10.6
10.8 - 11.9
6.0 - 7.2
7.2 - 7.9
Males Females
49.9
52.0
51.3
53.3
49.5
53.0
48.9
50.8
47.3
49.7
47.3
50.1
47.3
48.1
46.8
48.2
45.8
46.8
46.3
49.1
45.0
45.0
46.2
48.9
45.3
47.2
46.6
46.5
45.0
47.0
44.1
45.1
45.0
47.7
44.0
44.7
Males Females
8.4
9.0
9.4
10.7
9.3
10.5
8.9
9.4
8.7
8.8
8.7
9.9
8.4
8.5
8.7
9.7
8.0
8.4
9.3
10.2
7.3
7.0
9.2
10.4
8.5
8.4
8.2
9.6
8.2
9.5
7.5
8.0
8.5
10.9
8.2
7.8
Males Females
16.9
17.3
18.4
20.1
18.8
19.7
18.2
18.6
18.4
17.6
18.4
19.8
17.7
17.6
18.5
20.2
17.5
17.9
20.0
20.7
16.2
15.5
19.9
21.2
18.9
17.9
17.5
20.6
18.1
20.2
17.0
17.7
19.0
22.9
18.6
17.4
35.4 - 37.3
35.3 - 37.7
35.2 - 37.1
33.8 - 37.4
34.4 - 37.0
32.8 - 36.3
33.5 - 36.9
33.1 - 35.3
32.3 - 34.7
32.5 - 34.5
31.3 - 33.7
31.3 - 34.3
31.6 - 33.5
31.6 - 33.0
29.5 - 31.7
28.4 - 30.4
29.0 - 31.1
24.8 - 27.1
7.3 - 8.4
10.1 - 11.1
8.7 - 9.8
11.5 - 12.6
8.5 - 9.8
8.8 - 10.0
10.1 - 11.4
7.7 - 8.9
8.0 - 9.2
8.4 - 9.7
9.6 - 10.6
6.9 - 8.1
6.8 - 8.0
9.3 - 10.0
10.1 - 11.4
7.7 - 8.9
8.8 - 10.6
5.3 - 6.7
45.1
43.3
44.4
43.3
44.1
43.2
41.8
42.5
41.4
41.3
40.9
41.2
41.9
39.5
38.0
37.3
37.2
33.2
8.7
7.7
8.5
7.5
8.8
8.6
8.1
8.7
7.9
8.7
7.5
8.4
9.0
7.2
8.0
8.0
9.0
7.4
19.3
17.7
19.1
17.4
19.9
19.9
19.3
20.4
19.1
21.0
18.4
20.3
21.4
18.2
21.1
21.3
24.3
22.4
7.8
10.6
9.2
12.1
9.1
9.4
10.7
8.3
8.6
9.0
10.1
7.4
7.4
9.7
10.7
8.3
9.6
6.0
(a) Uncertainty ranges show 10th and 90th percentiles of the uncertainty distribution of the indicator (rank or DALE).
(b) DLE (Expected years lost to disability) is calculated as total life expectancy minus DALE. Per cent of life expectancy lost to disability is DLE/LE as a per cent.
(c) The Former Yugoslav Republic of Macedonia
79
Expected years Per cent of total
lost to disability life expectancy
at birth (DLE)
lost to disability
46.5
44.9
45.6
43.0
45.7
43.8
44.0
44.9
43.1
44.0
40.0
42.3
42.4
39.3
39.0
38.4
40.6
35.4
10.3
8.4
9.5
7.6
10.0
9.2
8.9
10.7
9.5
10.5
7.6
9.6
9.9
7.1
8.3
9.0
10.5
9.5
22.1
18.7
20.8
17.7
21.9
21.1
20.3
23.8
22.1
23.8
18.9
22.6
23.4
18.0
21.3
23.3
25.8
26.7