Chronic diseases and lost productivity: costs to individuals, the - - PowerPoint PPT Presentation

Faculty of Pharmacy University of Sydney

Faculty of Pharmacy

Faculty of Pharmacy University of Sydney

Chronic diseases and lost productivity: costs to individuals, the government and the society

Dr Rupendra Shrestha Senior Research Fellow Faculty of Pharmacy The University of Sydney

The Treasury Wellington

5 April 2017

Faculty of Pharmacy University of Sydney

• Prof Deborah Schofield, University of Sydney
• Assoc Prof Megan Passey, University of Sydney
• Prof Simon Kelly, University of Canberra
• Richard Percival, University of Canberra
• Prof Robert Tanton, University of Canberra
• Dr Emily Callander, University of Sydney
• Dr Michelle Cunich, University of Sydney
• Prof Theo Vos, University of Queensland
• Dr Lennert Veerman, University of Queensland

2

Collaborators

Faculty of Pharmacy University of Sydney

Population pyramid, Australia

3

Source: ABS (2008) Population Projections, Australia. Cat. No. 3222.0

Faculty of Pharmacy University of Sydney

Demographic change

4

Source: ABS (2008) Population Projections, Australia. Cat. No. 3222.0

Australian Population projections (%)

Faculty of Pharmacy University of Sydney

Economic implications

• Shrinking working age population
• Aged-to-working age ratio increase
• Increase dependency ratio

5

2010: About 5 people of working age ~ one person of pension age (aged over 65) 2050: About 2.7 people of working age ~ one person of pension age (aged over 65)

0% 5% 10% 15% 20% 25% 30% 35% 40% 2010 2020 2030 2040 2050 % of aged to working-age Year

Source: Intergenerational Report 2010

Faculty of Pharmacy University of Sydney

6

Projections of Australian Government spending by category

Source: Intergenerational Report 2015

1 2 3 4 5 6 Health Aged care Age and Service Pensions Other income support Education Defence

Percent of GDP

2014-15 2054-55

Faculty of Pharmacy University of Sydney

• How much money will be available for health spending
• How much pressure will health spending place on the future budget

balance

• Sustainability of future health spending
• Places the future of health funding in the context of the future economy

and budget balance

7

Sustainability of health spending

Faculty of Pharmacy University of Sydney

8

Ripple effects

Health is a social investment

Faculty of Pharmacy University of Sydney

Direct costs of OA

9

Hunter et al. Nat Rev Rheumatol 2014, 10:437-441

Faculty of Pharmacy University of Sydney

Indirect costs of OA

10

Hunter et al. Nat Rev Rheumatol 2014, 10:437-441

Faculty of Pharmacy University of Sydney

Health and labour force priorities

11

Health as a driver for labour force participation

• Poor health excludes people from the labour force
• 58% of men and 26% of women who retire from full-time work early

(before the age of 55 years) do so because of ill health.

• Improving health to increase labour force participation

Faculty of Pharmacy University of Sydney

Impact of chronic disease on labour force participation

12

Faculty of Pharmacy University of Sydney

13

Faculty of Pharmacy University of Sydney

14

Faculty of Pharmacy University of Sydney

• The total number of Australians aged 45 to 64 not working due to a health

condition was estimated to be 663,000 (14%) in 2003

• This lost workforce reduced Australia’s GDP by around $12 billion per

annum.

15

Total impact

Faculty of Pharmacy University of Sydney

Costs of lost productivity

• What are the impacts on personal incomes ?
• How much tax is lost ?
• How much extra social security payments are paid ?
• What are the impacts on savings and lifetime living standards ?
• What are the impacts on poverty ?
• What impacts would interventions to treat chronic conditions have ?

16

Faculty of Pharmacy University of Sydney

17

Health&WealthMOD

Faculty of Pharmacy University of Sydney

Health&WealthMOD

• Health&WealthMOD
• Australian microsimulation model of health, income, savings, tax and government

benefits.

• Development of this economic model was funded by an ARC Linkage Project

Grant with Pfizer Australia as an industry partner

• It estimates:
• lost income and savings, lost tax revenue and payment of benefits to those who

have retired early due to illness.

• the relative costs of diseases that lead to early retirement.

18

Faculty of Pharmacy University of Sydney

Model

19

2003 ABS Survey of Disability, Ageing and Carers

Income group, disability, long term conditions, retirement, demographic variables

Health&WealthMOD Base Population

Reweighting

STINMOD

Tax Centrelink payments Personal income and wealth

Simulate economic impacts

Personal: Lost earnings Government: Lost tax Centrelink payments

Simulate impact of interventions

e.g. prevention of diabetes

Synthetic matching

Faculty of Pharmacy University of Sydney

Health&WealthMOD

20

Matching Variables:

• Income quintile
• Income unit type
• Type of government pension/support
• Labour force status
• Sex
• Age group
• Hours worked per week
• Highest educational qualification
• Home ownership

Synthetic matching of STINMOD output with the base population of Health&WealthMOD

Faculty of Pharmacy University of Sydney

Cost of retiring early due to illness in Australia

21

Labour force status No in Survey Income# Transfer income$ Tax liability! Employed full-time 4,266 (46%) 1,167.0 0.2 166.3 Employed part-time 1,738 (19%) 482.9 0.9 8.9 Not in labour force due to ill health 661 (7%) 217.8 74.2 0.0

Geometric means of weekly income, transfer payments and tax liability by labour force status for the Australian population aged 45-64 years, 2009

# including transfer income $ including family payments ! including Medicare levy

Schofield et al. BMC Public Health 2011, 11:418

Faculty of Pharmacy University of Sydney

22

Income# Transfer payments$ Tax liability! Not in labour force due to ill health 17,989,175,000 1,468,007,000 2,052,384,000

National annual impact of persons not in the labour force due to ill health for the Australian population aged 45-64 years, 2009

# including transfer payments $ including family payments ! including Medicare levy Note: Based on the differences between persons not in the labour force due to ill health and the weighted average of persons employed full time and part time.

Impact of ill health:

Exiting the workforce as a result of ill health is costly to both individuals and

• governments. The amount of lost income was significant, adding to the hardship

experienced due to illness itself.

Cost of retiring early due to illness in Australia

Faculty of Pharmacy University of Sydney

23

Income# Transfer payments$ Tax liability! Not in labour force due to back pain (compared to those with no

chronic condition and employed)

4,816,000,000 622,200,000 496,900,000 Not in labour force due to back pain (compared to those with no

back pain and employed)

4,286,900,000 537,300,000 440,400,000

National annual impact of persons not in the labour force due to back pain for the Australian population aged 45-64 years, 2009

# including transfer payments $ including family payments ! including Medicare levy

Cost of retiring early due to back pain in Australia

Faculty of Pharmacy University of Sydney

Diabetes and lost wealth

24

Odds ratios of having any wealth by different classes of wealth, Australians 45-64 years old, 2009, adjusted for age, gender and education

Labour force status Total wealth Income producing assets$ Non-income producing assets# Employed full-time, no health condition Ref Ref Ref Employed part-time, no health condition 1.16 (0.27 – 4.94) 0.45 (0.23 – 0.94) 1.59 (0.99 – 2.56) Not in labour force due to diabetes 0.03 (0.00 – 0.30) 0.09 (0.02 – 0.50) 0.20 (0.06 – 0.71)

$ includes cash, superannuation, shares and property investments # includes the value of owner occupied home

Schofield et al. Br J Diabetes Vasc Dis 2010, 10: 300-304

Faculty of Pharmacy University of Sydney

Chronic conditions and lost wealth

25

Labour force status Total wealth References Employed full time, no chronic condition Ref Not in labour force due to other mental illness

• 93%

(-98 – -71)

Schofield et al. Brit J of Psy 2011, 198: 123-128

Not in labour force due to back pain

• 87%

(-90 – -84)

Schofield et al. Eur Spine J 2011, 20: 731-736

Not in labour force due to CVD

• 84%

(-89 – -77)

Schofield et al. Int J Cardiol 2011, 146(1): 125-126

Not in labour force due to arthritis

• 85%

(-88 – -81)

Schofield et al. Rheumatology Int 2015, 35: 1175-1181

Not in labour force due to diabetes

• 90%

(-94 – -81)

Schofield et al. Br J Diabetes Vasc Dis 2010, 10: 300-304

% difference (95% CI) in value of wealth, Australians 45-64 years old, 2009, adjusted for age, gender and education

Faculty of Pharmacy University of Sydney

Lifetime costs of exiting workforce early due to chronic conditions

26

Total savings (cash, super, shares, other property) at age 65 Annuity at age 65 Mean ($) Median ($) Mean ($) Median ($) Male, 45 – 54 Actual 56,740 1,810 3,280 100 Counterfactual 377,330 249,160 21,800 14,400 Male, 55 – 64 Actual 145,710 34,400 8,420 1,990 Counterfactual 322,220 208,220 18,620 12,030 Female, 45 – 54 Actual 76,100 830 3,850 40 Counterfactual 246,480 170,040 12,470 8,600 Female, 55 – 64 Actual 103,440 9,410 5,230 480 Counterfactual 189,200 120,650 9,570 6,100

Comparisons of actual total savings and annuity at age 65 for those not in the labour force due to ill health with the counterfactual total savings and annuity at age 65 if they were employed and had no chronic condition

Kelly et al. The Economic Record 2012, 88(283):576-584

Faculty of Pharmacy University of Sydney

Summary

• Chronic conditions are associated with people being out of the labour force
• Economic costs of labour market withdrawal because of ill health (costs of

productivity lost) are substantial.

• In addition to the health burden caused by these conditions, they also

contribute to economic disadvantage due to impacts on labour force participation.

• People who retire early due to chronic condition are left with reduced

savings to finance their retirement years resulting in lower living standards in retirement.

• Exiting the workforce as a result of ill health is costly not only to individuals

and but also to governments (i.e. lost tax revenue and increased transfer payments)

27

Faculty of Pharmacy University of Sydney

Health&WealthMOD

• Health&WealthMOD: a cross-sectional model

Does not capture the long-term impacts of

• demographic change
• the changing labour force participation patterns of older working aged population
• change in wealth accumulation patterns
• trends in major chronic conditions affecting the older workforce

28

What will be the economic impacts of early retirement due to ill health in future?

Faculty of Pharmacy University of Sydney

29

Health&WealthMOD2030

Faculty of Pharmacy University of Sydney

Health&WealthMOD2030

• Funded by ARC linkage project grant with Pfizer Australia as an industry

partner

• A microsimulation model of long term economic impacts of ill health on
• lder working aged Australians and the Australian government
• Takes into account the significant socio-demographic and economic

changes and trends in major chronic conditions predicted to occur between 2010 and 2030

• Estimates the economic impacts of early retirement due to ill health to

individuals and the government in every five years from 2010 to 2030

30

Purpose

Faculty of Pharmacy University of Sydney

Health&WealthMOD2030

• Australian Bureau of Statistics Surveys of Disability, Ageing and Carers

2003 and 2009

• The SDACs include the following variables:
• demographics; e.g. age, sex, family type, region and state of residence
• socioeconomics; level and field of education, income, benefits received
• labour force; labour force participation, employment restrictions,

retirement, average weekly hours of paid employment, reason for leaving the labour force (e.g. due to own ill health / to become a carer)

• care responsibilities; caring status (primary, other or non-carer), average

weekly hours caring

• 45 to 64 years age group
• 25,104 unit records (8864 from SDAC 2003 and 16,240 from SDAC 2009)

31

Data

Faculty of Pharmacy University of Sydney

Health&WealthMOD2030

32

ABS Surveys of Disability, Ageing and Carers 2003, 2009 Australian Population and Policy Simulation Model (APPSIM) 2003 Australian Burden of Disease and Injury Study

Health&WealthMOD2030

• utput datasets

Long-term health trends Socio-demographic and economic trends

The Australian Treasury

Population, labour force projections

STINMOD

Economic data including personal incomes, welfare payment, tax

Faculty of Pharmacy University of Sydney

Reweight

• A reweighting algorithm GREGWT (SAS macro) from the ABS

Benchmarks:

• For socio-demographic and economic changes : Projected Australian

population

• by age group, sex, labour force participation
• by age group, sex and income quintiles
• by age group, sex and highest level of education
• by age group, sex and whether or not receiving disability support pension
• by income unit type
• by age group, sex and whether or not owns a home
• For disease trends: Age, sex specific prevalence projections based on the

Australian Burden of Disease and Injury Study

• heart diseases; stroke; diabetes; COPD; and cancer

33

Reweight the dataset to the projected Australian population

Faculty of Pharmacy University of Sydney

34

Synthetic matching of STINMOD output with Health&WealthMOD2030 base population

Synthetic dataset

Matching variables Number of categories Labour force status 4 Income unit type 4 Income quintile 5 Receiving age pension 2 Receiving disability support pension 2 Sex 2 Age group 4 Hours worked per week 5 Highest educational qualification 2 Home ownership 2

Faculty of Pharmacy University of Sydney

35

Synthetic dataset

Employed full time; Married couple with dependents; 2nd income quintile; No age pension; No disability support pension Home

• wnership

Employed full time; Married couple with dependents; 4th income quintile; No age pension; No disability support pension Sex Male Female Home

• wnership

Highest level

• f education

Yes No Highest level

• f education

Sex Cell 2 Cell 4

Faculty of Pharmacy University of Sydney

• Stopping rules:
• less than 10 records in a cell
• used all matching variables
• Standardized Euclidean Distance

36

Synthetic dataset

Faculty of Pharmacy University of Sydney

Productive Life Years (PLYs)

• Productive life years of an individual
• number of years an individual is in employment before reaching the retirement

age (panel data)

• Prevalence based productive life years
• total number of individuals in employment in any given year (cross-sectional

data)

• Individuals who responded as “not in the labour force” due to “their own

ill-health or disability” were identified as having productive life years (PLYs) lost due to chronic conditions

37

How did we measure it ?

Faculty of Pharmacy University of Sydney

Productive life years

Labour force status 2010 2015 2020 2025 2030 Growth (%) Employed full-time with a chronic condition 1,452,000 1,565,000 1,722,000 1,830,000 1,962,000 35.1 Employed full-time without a chronic condition 1,572,000 1,653,000 1,786,000 1,864,000 2,017,000 28.3 Employed part-time with a chronic condition 621,000 682,000 769,000 825,000 888,000 43.0 Employed part-time without a chronic condition 501,000 536,000 594,000 632,000 688,000 37.3 Productive Life Years (PLYs) lost due to chronic conditions 347,000 380,000 413,000 434,000 459,000 32.3 Total Population 5,580,000 5,945,000 6,374,000 6,677,000 7,130,000 27.8

38

Labour force status of Australians aged 45-64 years projected to 2030

Faculty of Pharmacy University of Sydney

39

Drivers of increase in PLYs

Growth in PLYs = f (population growth, labour force participation trends, disease trends)

Stepwise reweight to the projected Australian population

• A reweighting algorithm GREGWT (SAS macro) developed at the Australian

Bureau of Statistics

Faculty of Pharmacy University of Sydney

Productive life years

Driver Employed full-time Employed part-time PLYs lost due to chronic conditions With a chronic Condition Without a chronic condition With a chronic condition Without a chronic condition

Population growth (total) Due to population ageing (a component of total population growth)

82.0%

• 0.9%

94.0%

• 3.3%

63.2% 1.2% 70.4%

• 3.3%

89.2% 4.0%

Chronic disease trends

1.8%

• 4.3%

0.5%

• 2.7%

8.3%

Labour force trends

10.4% 8.4% 26.4% 26.8% NA

Interaction effects

5.8% 1.9% 9.9% 5.5% 2.5%

Population growth, chronic condition trends, labour force trends and interactions

100% 100% 100% 100% 100%

40

Contribution of main drivers to growth in PLYs lost and labour force participation between 2010 and 2030, Australians aged 45-64 years

Faculty of Pharmacy University of Sydney

Projections of cost of retiring early due to illness

41

Schofield et al. BMJ Open 2016

Faculty of Pharmacy University of Sydney

GDP lost

2010 2015 2020 2025 2030 37.8 44.5 50.5 56.4 63.7

42

Estimated loss of GDP due to productive life years lost caused by chronic ill health in Australians aged 45-64 years, ($ billions)

Schofield et al. MJA 2015, 203(6): 260.e1:e6

Faculty of Pharmacy University of Sydney

43

Impact of diabetes intervention

• n labour force participation

(Passey et al. BMC Public Health 2012, 12:16)

Faculty of Pharmacy University of Sydney

Impact of diabetes prevention on labour force participation

• Investigated the impact on labour force participation and personal incomes
• f diabetes intervention using a screening program and metformin or

lifestyle intervention for those identified as pre-diabetics

• Estimated the extra number of years in the labour force and increase in

personal incomes of Australians aged 45 – 64 years in 2003, who would not have developed diabetes if a screening and intervention program to prevent the onset of diabetes were in place for the 20 years from 1983

44

Faculty of Pharmacy University of Sydney

Model

45

Epidemiological Microsimulation Model ABS Survey of Disability, Ageing and Carers Estimate extra number of person years in labour force due to intervention Increase in total income Health&WealthMOD Income

Faculty of Pharmacy University of Sydney

Results

46

Differences in labour force participation rates of people without diabetes and with diabetes, 2003

Age group 40 – 44 45 – 49 50 – 54 55 – 59 60 – 64 Male 0.10 0.16 0.10 0.12 0.13 Female 0.30 0.10 0.27 0.33 0.09

Faculty of Pharmacy University of Sydney

Impacts of intervention to prevent diabetes

47

Increased number of person years in the labour force & the associated increase in total incomes over the ten years

Age group in 2003 Using metformin Lifestyle intervention Total person years Total income Total person years Total income Male 45-49 28 1,263,000 43 1,896,000 50-54 97 4,319,000 125 5,595,000 55-59 282 12,578,000 358 15,967,000 60-64 683 30,486,000 753 33,599,000 Female 45-49 11 347,000 11 347,000 50-54 42 1,329,000 42 1,329,000 55-59 679 21,629,000 816 25,983,000 60-64 790 25,144,000 890 28,334,000 Total 2,612 97,095,000 3,038 113,049,000

$44,600 per person per year $31,800 per person per year

Faculty of Pharmacy University of Sydney

Summary

• Health is a social investment.
• Investment in health can result in savings in other portfolios in addition to

health.

• Preventing / managing chronic conditions is one way to increase the

workforce to grow economy.

• This may contribute towards a sustainability of the future health funding.

48

Faculty of Pharmacy University of Sydney

49

Thank you !!!