Transcript Burden of disease:

Burden of disease:
Concepts and applications
Session Aims
1. to introduce the concept “burden of disease”
2. to examine patterns and trends in mortality in
Southern African settings
3. to discuss and evaluate the concept of “health
transition”
4. to introduce the concept of “priority setting”
and its relation to burden of disease studies
5. to examine the implications of South African
mortality patterns for the provision of health
care in the country.
Data to measure burden of disease
Industrialised versus developing settings
National data
eg census, vital registration
Health facilities
Surveys
eg household surveys: DHS
Sentinel site data
eg India, China, HDSS, verbal autopsy
Models
The disability-adjusted life year
(DALY)
A single measure of disease burden
Expresses years of life lost due to
premature death and years lived with a
disability (ie years of healthy life lost
due to poor health)
DALY: Values and methods
How “long” should people live?
Is a year of healthy life now worth more
than in 30 years’ time?
Are we – all people – equal?
How to compare years of life lost due to
premature death, and years lived with
disabilities of differing severities?
Trends in life expectancy
Agincourt 1992-2003
75
72
Female
Life expectancy
70
65
66
Male
60
60
55
52
50
1992-93
1994-95
1996-97
1998-99
Year
2000-01
2002-03
Relative increase in mortality,
Agincourt 2002-2003 compared to baseline 1992-1993
2002-2003 / 1992-1993
7
6
Male
5
Female
4
3
2
1
0
0-4
'5-14
15-24 25-34 35-44 45-54 55-64 65-74 75-84
Age group
Death rate 0-4
Trends in under-five mortality
0.100
0.090
0.080
0.070
0.060
0.050
0.040
0.030
0.020
0.010
0.000
Male
Female
1992-93
1994-95
1996-97
1998-99
Period
2000-01
2002-03
Trends in adult mortality
Age 20-34
Death rate 20-34
0.150
0.100
Male
0.050
Female
0.000
1992-93
1994-95
1996-97
1998-99
Period
2000-01
2002-03
Trends in cause specific mortality:
Infectious & parasitic disease
0.0025
Diarrhoeal diseases
Acute respiratory infection
0.0020
HIV/AIDS
Tuberculosis
ASDR
0.0015
Malaria
Other Infectious and parasitic diseases
0.0010
0.0005
0.0000
1992-1994
1995-1997
1998-2000
Period
2001-2003
Trends in adult mortality
Age 50-64
Death rate 50-64
0.500
0.400
0.300
Male
0.200
Female
0.100
0.000
1992-93
1994-95
1996-97
1998-99
Period
2000-01
2002-03
Trends in cause specific mortality:
Women 50-64, broad categories
0.0090
0.0080
0.0070
Infectious and parasitic
Non-communicable
External
Ill defined or unknown
ASDR
0.0060
0.0050
0.0040
0.0030
0.0020
0.0010
0.0000
1992-1994
1995-1997
1998-2000
Period
2001-2003
Age-standardised death rates, broad cause and
broad health care categories, Agincourt 1992-2005
Top five causes of death, 50-64 years
Agincourt 1992-2005
Top five causes of death, children and older
adults, Agincourt 1992-2005
Prevalence of stroke survivors:
South Africa, Tanzania, New Zealand
Com parison of age-standardised rates in three prevalence studies
Prevalence / 100,000 Age-Standardised to
Segi Population
1000
900
800
700
Auckland, New Zealand
600
Tanzania
500
Agincourt, South Africa
400
300
200
100
0
Male
Female
Total
Males
Female
Total Needing
Needing Help Needing Help
Help
Stroke Survivors
Managing chronic NCDs in Agincourt
Sub-district services based on network of clinics
staffed by primary care nurses with limited support
drug supply irregular
medical supervision sporadic
Poor capacity to manage chronic illness
No functional system secondary prevention
103 stroke survivors – only 1 on aspirin
85 hypertensives
– 8 on treatment; only 1 controlled
General pop ≥ 35
– 43% hypertension; 24% of these treated in past
week; half with BP controlled
Missed diagnoses
Majority of deaths with active TB had previously presented to clinic
2/3 TB patients seen at a clinic self-referred to hospital
Care-seeking pluralistic – allopathic, traditional, faith-based
most first visits to local clinics = pivotal role
Age-standardised death rates by health care
categories, Agincourt sub-district 1992-2005
PHC in Practice: Integrating HAART & chronic NCD care
Age and sex standardized death rates, by social strata,
Agincourt 1992-2000
0.006
0.005
ASDR
0.004
0.003
0.002
0.001
0.000
Highest
Higher
Medium
Strata
Lower
Lowest
Age-specific death rates by nationality of
household head
Age-specific death rates by nationality of household head,
Agincourt, 2000-2001
0.0016
South African
0.0014
Mozambican
Age specific death rate
0.0012
0.001
0.0008
0.0006
0.0004
0.0002
0
0-1
1-2
2-3
Age in ye ar s
3-4
4-5
Reasons given for non-consultation:
no money, ineffective care
100
90
80
70
60
50
40
30
20
10
0
Other non-access barriers
Feeling better
26%
7%
Illness not serious enough
18%
Other access barriers (Too far,
nobody to go with patient, no time)
25%
Reasons for not taking
treatment action
Health system access barriers (Health
care can do nothing, drugs don't
work, no drugs at clinic)
Nomoney
money
No
Household survey data
Implications of mortality patterns for
health system
Shift orientation of service provision: chronic,
long-term care as well as acute, episodic care
Tackle (prevent/control) increasing burden of noncommunicable disease and risk
Strengthen HIV/AIDS (and TB) prevention, treatment
and care
Simultaneously maintain and improve on gains in child
and maternal health
•
•
Strengthen primary care provision + referral
system
Address differential access to care
Epidemiological Transition
Epidemiologic transition theory: 3 stages
Pestilence and famine
Receding pandemics
Man-made or degenerative disease
Critique
Not same direction: reversals in mortality “counter
transition”
Not sequential: stages may overlap, co-existence
different diseases “prolonged/protracted transition”
Too general: insufficient attention to subgroup
differences “epidemiologic polarisation”
Rethinking epidemiologic transition:
mortality patterns in rural South Africa
Counter transition
Mortality increasing in children and young
adults
Protracted or prolonged transition
Simultaneous emergence of HIV/AIDS
together with increasing non-communicable
disease
Epidemiologic polarisation
Poorest experience highest burden of
mortality
Why is burden of disease
information necessary?
“priority setting” and its relation to
burden of disease
Programme planning
Programme evaluation