Top Banner
University of Leeds, 18-20 March 2008 University of Leeds, 18-20 March 2008 Domenica Rasulo ( Domenica Rasulo ( User Fellow User Fellow)
29
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

University of Leeds, 18-20 March 2008University of Leeds, 18-20 March 2008

Domenica Rasulo (Domenica Rasulo (User FellowUser Fellow))

Page 2: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Background concepts – life expectancyBackground concepts – life expectancy

Life expectancy represents the expected number of years Life expectancy represents the expected number of years

to be lived and is computed by using death and population to be lived and is computed by using death and population

counts counts

It does not indicate whether people live in good or bad It does not indicate whether people live in good or bad

health (e.g. the same number of years can be lived with or health (e.g. the same number of years can be lived with or

without diseases)without diseases)

Using measures of disease prevalence, the number of years Using measures of disease prevalence, the number of years

to be lived can be partitioned in years into with and without to be lived can be partitioned in years into with and without

diseasesdiseases

Page 3: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Partitioning life expectancyPartitioning life expectancy

The number of years of life expectancy is generally The number of years of life expectancy is generally

partitioned in years with and without limiting illness:partitioned in years with and without limiting illness:

Life expectancy (LE) = LE illness Life expectancy (LE) = LE illness + LE without illnessLE without illness

Page 4: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Years of life with ‘limiting long term illness’ in Years of life with ‘limiting long term illness’ in EnglandEngland

1991 73.4 59.4 13.9 78.9 61.8 17.0

1992 73.6 59.7 13.9 79.0 61.9 17.1

1993 73.9 59.4 14.5 79.2 61.7 17.5

1994 74.1 59.4 14.6 79.3 61.7 17.6

1995 74.3 59.2 15.1 79.5 61.6 17.9

1996 74.5 … … 79.6 … …

1997 74.8 59.1 15.6 79.8 60.8 19.0

1998 75.0 … … 79.9 … …

1999 75.3 60.6 14.7 80.1 62.6 17.5

2000 75.6 … … 80.3 … …

2001 76.0 60.8 15.1 80.6 62.9 17.7

Source: ONS

FEMALES, birth

Life expectancy

(LE)

LE without limiting

long term illness

LE with limiting long

term illness

MALES, birth

Years Life expectancy

(LE)

LE without limiting

long term illness

LE with limiting long

term illness

Page 5: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Years of life with ‘limiting long term illness’ in Years of life with ‘limiting long term illness’ in England (indexed values)England (indexed values)

1991 100.0 100.0 100.0 100.0 100.0 100.0

1992 100.3 100.4 99.9 100.1 100.1 100.4

1993 100.7 100.0 103.9 100.4 99.8 102.9

1994 100.9 100.0 104.9 100.5 99.8 103.1

1995 101.3 99.6 108.6 100.8 99.5 105.3

1996 101.5 … … 100.9 … …

1997 101.9 99.5 112.2 101.1 98.3 111.3

1998 102.2 … … 101.3 … …

1999 102.6 101.9 105.6 101.5 101.2 102.8

2000 103.1 … … 101.9 … …

2001 103.5 102.4 108.5 102.2 101.6 104.1

Source: Analysis of ONS data

FEMALES, birth - I ndex number (base=1991)

Lif e expectancy (LE)LE without limiting

long term illness

LE with limiting

long term illness

Years

MALES, birth - I ndex number (base=1991)

Lif e expectancy (LE)LE without limiting

long term illness

LE with limiting long

term illness

Page 6: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Policy issues arisingPolicy issues arising

The burden on the health system might be underestimated The burden on the health system might be underestimated

if only ‘long term limiting illness’ is considered if only ‘long term limiting illness’ is considered

People may have a long term illness but not consider it People may have a long term illness but not consider it

‘limiting’ but they still may:‘limiting’ but they still may:

take medicine on a daily basistake medicine on a daily basis require hospitalisation require hospitalisation require hospital visitsrequire hospital visits

a)a) Can we circumvent this by using a more general Can we circumvent this by using a more general

definition of illness? definition of illness?

b)b) Can we break the years in ill health down into different Can we break the years in ill health down into different

conditions e.g. in order to estimate economic cost?conditions e.g. in order to estimate economic cost?

c)c) People might have more than one disease so need to be People might have more than one disease so need to be

able to deal with co-morbidityable to deal with co-morbidity

Page 7: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Different definitions of long term Different definitions of long term illnessillnessThe life expectancy with different diseases can be analysed The life expectancy with different diseases can be analysed

using the Health Survey for England (HSE). It includes using the Health Survey for England (HSE). It includes

questions on questions on

‘ ‘limiting’ long-term illness (from 1997)limiting’ long-term illness (from 1997) long-term illness (from 1991)long-term illness (from 1991) specific diseases (from 1991)specific diseases (from 1991)

Using the HSE it is, therefore, possible to computeUsing the HSE it is, therefore, possible to compute

life expectancy with limiting long-term life expectancy with limiting long-term

illnessillness life expectancy with long-term illness life expectancy with long-term illness life expectancy with any disease life expectancy with any disease

Page 8: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

MethodologyMethodology

The diseases affecting a change in life expectancy with The diseases affecting a change in life expectancy with

disease can be identified using decomposition methods (e.g. disease can be identified using decomposition methods (e.g.

Sullivan 1971, Nussleder & Looman 2004Sullivan 1971, Nussleder & Looman 2004) )

Life expectancy with disease is computed by using mortality Life expectancy with disease is computed by using mortality

and disease data as followsand disease data as follows

LE with disease = total LE * disease LE with disease = total LE * disease

prevalenceprevalence

Changes over time can be decomposed by separating Changes over time can be decomposed by separating

mortality and disease effectsmortality and disease effects

Page 9: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Decomposition Decomposition (Arriaga 1989, Nussleder & Looman 2004(Arriaga 1989, Nussleder & Looman 2004) ) Let us suppose that life expectancy with disease has Let us suppose that life expectancy with disease has

changed between time changed between time tt and time and time t+nt+n

The change is attributable to a change in the number of The change is attributable to a change in the number of

person-years with disease (Lperson-years with disease (Ldisdis) which can be decomposed ) which can be decomposed

as followsas follows

t

Life

nt

Life

t

Dis

nt

Dist

Dis

nt

Dis

t

life

nt

lifeDis LL*

2*

2

LLL

Where the difference between the Where the difference between the DisDis depends on the depends on the

change of thechange of the prevalence ofprevalence of each cause of disease, and the each cause of disease, and the

difference betweendifference between the L the LLifeLife depends on the change of each depends on the change of each

cause of deathcause of death

Page 10: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Previous work on decompositionPrevious work on decomposition

The method was applied on Dutch data to explain the The method was applied on Dutch data to explain the

difference between the sexes in the period 1990-1994 difference between the sexes in the period 1990-1994

(Nussleder & Looman 2004)(Nussleder & Looman 2004)

The research indicated that the higher life expectancy with The research indicated that the higher life expectancy with

LLT illness among the women was mostly attributable to LLT illness among the women was mostly attributable to

arthritis.arthritis.

This required the identification of exclusive diseases This required the identification of exclusive diseases

conditions (i.e. they could not have more than one disease)conditions (i.e. they could not have more than one disease)

Page 11: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

ObjectivesObjectives

Decompose the differences over time and between the Decompose the differences over time and between the

sexes in the life expectancy with disease by applying the sexes in the life expectancy with disease by applying the

Nussleder & Looman method (2004)Nussleder & Looman method (2004)

Three types of life expectancy with diseaseThree types of life expectancy with disease

life expectancy with limiting long-term life expectancy with limiting long-term

illnessillness life expectancy with long-term illness life expectancy with long-term illness life expectancy with any diseaselife expectancy with any disease

Page 12: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Data usedData used

ONS (death and population counts)ONS (death and population counts)

Health Survey for England (disease prevalence)Health Survey for England (disease prevalence)

Analysis performed on the most complete data series Analysis performed on the most complete data series

available from ONS and the Health Survey for England:available from ONS and the Health Survey for England:

1991 - 20051991 - 2005

The Health Survey for England has included from 1991 The Health Survey for England has included from 1991

people aged 16 and above:people aged 16 and above:

All the analysis performed on people All the analysis performed on people

16+16+

Page 13: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Analysis of mortality dataAnalysis of mortality data

Compute period life tables for each sex consistently with GAD using Compute period life tables for each sex consistently with GAD using

3 years data 3 years data

Code specific causes of death consistently through ICD IX Code specific causes of death consistently through ICD IX

and ICD X and ICD X

Aggregate specific causes of death in categories (e.g. neoplasm) Aggregate specific causes of death in categories (e.g. neoplasm)

Compute mortality rates for each category of cause of death Compute mortality rates for each category of cause of death

by age band and sex by age band and sex

Page 14: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Classification of causes of deathClassification of causes of death

Causes of death Groups Label of the group 1 I nfectious & parasitic diseases 1 I nfectious disease 2 Accidents 2 Accidents & I njuries 3 Suicides 2 Accidents & I njuries 4 Neoplasms 3 Cancers 5 Endocrine & immunity disorders 6 Other chronic diseases 6 Blood diseases 6 Other chronic diseases 7 Mental disorders 5 Respiratory & Other 8 Nervous system 6 Other chronic diseases 9 Circulatory system 4 Circulatory system 10 Respiratory system 5 Respiratory & Other 11 Digestive system 5 Respiratory & Other 12 Genitourinary system 6 Other chronic diseases 13 Pregnancy complications 6 Other chronic diseases 14 Skin diseases 6 Other chronic diseases 15 Musculoskeletal diseases 5 Respiratory & Other 16 Congenital anomalies 6 Other chronic diseases 17 Conditions f rom perinatal period 6 Other chronic diseases 18 I ll-defi ned conditions 6 Other chronic diseases 19 Diseases of the eye 6 Other chronic diseases 20 Diseases of the ear 6 Other chronic diseases

Page 15: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Mortality rates by cause (Males)Mortality rates by cause (Males)

0

25

50

75

100

125

150

175

200

225

250

275

300

325

350

375

400

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Years

Accidents Cardiovascular

I nfections Neoplasm

Respiratory Chronic

Mortality rate per 1,000, Age 16+

Page 16: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Mortality rates by cause (Females)Mortality rates by cause (Females)

0

25

50

75

100

125

150

175

200

225

250

275

300

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Years

Accidents Cardiovascular

I nfections Neoplasm

Respiratory Chronic

Mortality rate per 1,000, Age 16+

Page 17: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Health Survey for EnglandHealth Survey for England

Measure the prevalence of different disease seriousness:Measure the prevalence of different disease seriousness:

prevalence of diseases (from 1991) prevalence of diseases (from 1991) prevalence of long-term illness (from 1991) prevalence of long-term illness (from 1991) prevalence of limiting long-term illness (from prevalence of limiting long-term illness (from

1997)1997)

Identify the diseases based on survey questionsIdentify the diseases based on survey questions

Compute exclusive disease categories for each year of Compute exclusive disease categories for each year of

survey for each person in the surveysurvey for each person in the survey

Page 18: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Information collected in different years (HSE 1991-2005)Information collected in different years (HSE 1991-2005)

Questions 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 LT illness List of illnesses if LT illness is present

LLT illness Taking medicine for 1) Cardiovascular diseases 2) Digestive diseases 3) Respiratory diseases 4) Nervous system diseases 5) I nfections 6) Endocrine diseases 7) Genitourinary diseases 8) Neoplasm 9) Blood diseases 10) Musculoskeletal diseases 11) Eye/Ear diseases 12) Skin diseases Cardiovascular diseases Blood pressure Diabetes Asthma Bladder

Page 19: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Disease categories analysed (A)Disease categories analysed (A)

Number of diseases Disease categories

One Two Three Four Five

Cardiovascular diseases I nfections Neoplasm Respiratory Diseases & Other Chronic Diseases

Cardiovascular & Respiratory

Cardiovascular & Chronic

Cardiovascular & Neoplasm

Cardiovascular & I nfections

Respiratory & Chronic

Respiratory & Neoplasm

Respiratory & I nfections

Chronic & Neoplasm

Chronic & I nfections

Neoplasm & I nfections

Page 20: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Disease categories analysed (B)Disease categories analysed (B)

Number of diseases Disease categories

One Two Three Four Five

Cardiovascular & Respiratory & Chronic

Cardiovascular & Respiratory & Neoplasm

Cardiovascular & Respiratory & I nfections

Respiratory & Chronic & Neoplasm

Respiratory & Chronic & I nfections

Respiratory & Neoplasm & I nfections

Cardiovascular & Chronic & Neoplasm

Cardiovascular & Chronic & I nfections

Chronic & Neoplasm & I nfections

Cardiovascular & Respiratory & Chronic & Neoplasm

Cardiovascular & Respiratory & Chronic & I nfections

Cardiovascular & Respiratory & Neoplasm & I nfections

Cardiovascular & Chronic & Neoplasm & I nfections

Respiratory & Chronic & Neoplasm & I nfections

Cardiovascular & Resp. & Chronic & I nfections & Neoplasm

Page 21: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Prevalence of diseases by %Prevalence of diseases by %

Disease categories, Males Life expectancy

with disease(s)

Life expectancy

with LT illness

Life expectancy

with LLT illness

Cardiovascular diseases 16.0 15.1 9.8

I nfections 1.4 0.3 0.2

Neoplasm 1.0 1.1 1.0

Respiratory diseases 34.6 37.4 36.7

One disease

Chronic diseases 17.8 16.2 11.7

Cardiovascular & respiratory

diseases 7.7 8.3 11.3

Cardiovascular & chronic

diseases 5.8 5.7 5.9

Two diseases

Respiratory & chronic diseases 8.3 8.7 12.2

Three diseases Cardiovascular, respiratory &

chronic diseases 4.5 4.5 7.5

Two diseases

or more Other categories of diseases

3.0 2.7 3.6

Total 100.0 100.0 100.0

any disease LT illness LTLImales

Page 22: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

101112131415161718192021222324252627282930313233

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Years

LE with disease LE with LT illness LE with LLT illness

Years of life with diseases Males, age 16

Results: Life expectancy with disease(s), Results: Life expectancy with disease(s), LT illness & LLT illness (Males)LT illness & LLT illness (Males)

Page 23: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Results: Life expectancy with disease(s), Results: Life expectancy with disease(s), LT illness & LLT illness (Females)LT illness & LLT illness (Females)

131415161718192021222324252627282930313233343536373839

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Years

LE with disease LE with LT illness LE with LLT illness

Years of life with diseases Females, age 16

Page 24: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Life Expectancy with disease(s)Life Expectancy with disease(s)

Males Females Life expectancy at age 16

1991- 93 2003- 05 1991- 93 2003- 05

Lif e expectancy 58.4 61.5 63.7 65.7

Lif e expectancy with disease(s) 26.7 31.9 34.5 37.2

Lif e expectancy without disease(s) 31.7 29.6 29.2 28.5

Disease/ LE (%) 45.7 51.9 54.2 56.6

No disease/ LE (%) 54.3 48.1 45.8 43.4

Page 25: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Life Expectancy with long-term illnessLife Expectancy with long-term illness

Males Females Life expectancy at age 16

1991- 93 2003- 05 1991- 93 2003- 05

Lif e expectancy 58.4 61.5 63.7 65.7

Lif e expectancy with LT illness 22.3 28.4 26.2 31.6

Lif e expectancy without LT illness 36.1 33.1 37.5 34.1

LT illness/ LE (%) 38.2 46.2 41.1 48.1

No LT illness/ LE (%) 61.8 53.8 58.9 51.9

Page 26: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Life Expectancy with limiting LT illnessLife Expectancy with limiting LT illness

Males Females Life expectancy at age 16

1997- 99 2003- 05 1997- 99 2003- 05

Lif e expectancy 59.7 61.5 64.5 65.7

Life expectancy with LLT illness 12.0 13.1 14.5 16.4

Life expectancy without LLT illness 47.7 48.4 50.0 49.3

LLT illness\ LE (%) 20.1 21.3 22.5 25.0

No LLT illness\ LE (%) 79.9 78.7 77.5 75.0

Page 27: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Emerging conclusions (Males, age 16)Emerging conclusions (Males, age 16)

Between 1992 and 2004 the total life expectancy has Between 1992 and 2004 the total life expectancy has

increased by 3.1 yearsincreased by 3.1 years

In the same period the life expectancy with any disease In the same period the life expectancy with any disease

has increased by 5.2 years and the life expectancy with has increased by 5.2 years and the life expectancy with

long-term illness by 6.1 yearslong-term illness by 6.1 years

The life expectancy with limiting long-term illness has The life expectancy with limiting long-term illness has

increased by 1.1 years between 1998 and 2004increased by 1.1 years between 1998 and 2004

Page 28: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Emerging conclusions (Females, age 16)Emerging conclusions (Females, age 16)

Between 1992 and 2004 the total life expectancy has Between 1992 and 2004 the total life expectancy has

increased by 2 yearsincreased by 2 years

In the same period the life expectancy with any disease In the same period the life expectancy with any disease

has increased by 2.7 years and the life expectancy with has increased by 2.7 years and the life expectancy with

long-term illness by 5.4 yearslong-term illness by 5.4 years

The life expectancy with limiting long-term illness has The life expectancy with limiting long-term illness has

increased by 1.9 years between 1998 and 2004increased by 1.9 years between 1998 and 2004

Page 29: University of Leeds, 18-20 March 2008 Domenica Rasulo (User Fellow)

Emerging conclusions (Co-morbidity)Emerging conclusions (Co-morbidity)

Between 1992 and 2004, out of 100 men with any Between 1992 and 2004, out of 100 men with any

disease or a long-term illness 29 have a co-morbidity disease or a long-term illness 29 have a co-morbidity

condition condition

Out of 100 men with a limiting long-term illness, 40 Out of 100 men with a limiting long-term illness, 40

have a co-morbidity condition have a co-morbidity condition

Out of 100 women with any disease or a long-term Out of 100 women with any disease or a long-term

illness, 32 have a co-morbidity condition illness, 32 have a co-morbidity condition

Out of 100 women with a limiting long-term illness, 44 Out of 100 women with a limiting long-term illness, 44

have a co-morbidity condition have a co-morbidity condition