Tobacco tax modelling and other cost-effectiveness studies for NZ: - - PowerPoint PPT Presentation

Tobacco tax modelling and

• ther cost-effectiveness studies

for NZ: Latest BODE3 Results

Tony Blakely Nick Wilson Treasury 13 August 2014

Overview

• who we are
• what we do
• tobacco endgame – game on!
• other examples:
• HPV vaccination, cancer care coordinators and

costing studies

• BODE3 tools

Who we are

Size of the problem Effectiveness Cost-effectiveness Affordability Safety/Risk Impact on inequalities Feasibility Public acceptability Government priorities Availability of other alternatives Direct cost to users Other criteria

“Is this a sensible thing to do?”

Cost-effectiveness

What do we do more of? What do we do less of? Is this a sensible thing to do? Who should get this?

How we do it

Cost-effectiveness

INPUTS MODEL OUTPUTS

MODEL

Markov models Discrete event simulation Multistate life tables

INPUTS

Same general data inputs, methods and outputs across BODE3 evaluations comparability between interventions

Burden of Disease Epidemiology, Equity and Cost-Effectiveness Programme

Tobacco Endgame – Game On!

15

Tony Blakely, Nick Wilson, Amber Pearson, Linda Cobiac, Nhung Nghiem and Frederieke van der Deen Tobacco Endgame

Structure

• Background
• ‘Business as Usual’ (BAU) – what smoking prevalence do

we forecast out to 2025 and beyond?

• Ongoing 10% (and more) per annum increases in tax:

– What impact will that have on smoking prevalence? – What impact will that have on QALYs and health system costs?

• So what?

– Putting it in context of the endgame – What we should do next – policy and research.

16

Background – tobacco smoking in NZ

• The major single risk factor causing of health loss in NZ

(for 2006 – NZBDS)

• Major contributor to ethnic inequalities in health
• NZ one of 4 countries with endgame goals
• NZ using regular tax increases (annual since 2010), retail

display restrictions, strong cessation activity (Quitline service, pharmacotherapy, counselling by health workers), but fairly minimal in some areas (mass media)

17

BAU smoking prevalence projections to 2025 and beyond in New Zealand

18

Methods

• A dynamic forecasting tobacco model previously built for

Australia1 was adapted for NZ by Ikeda et al. (2013)2

• A Markov model designed in MS Excel

– Input data (by age, sex, and ethnicity)

• Smoking prevalence data from the 2006 and 2013 NZ Census3
• Annual birth projections
• Annual trends in mortality rates
• Relative risks of mortality for current and former smokers from NZCMS4
• 1. Gartner et al. Tob Control 2009;18:183-9.
• 3. Cobiac et al. Tob Control Under review
• 2. Ikeda et al. Tob Control doi:10.1136/tobaccocontrol-2013-051196
• 4. Blakely et al. N Z Med J 2010;123(1320):26-36
• 3. Van der Deen et al. Under review (updates Ikeda from using NZHS to census data – important)

20

Smoking prevalence projections - men

23 18.7% 8.3%

2025

Smoking prevalence projections - women

24 19.3% 6.4% 4.4%

Intermediate conclusion

• The 2025 goal is not achieved by any group under the

projected annual trends in initiation and cessation (assuming no further tax rises after 2014)

• Thus, time to explore scenarios that go beyond business-

as-usual:

– E.g. Ongoing 10% (and more) per annum increases in tax

25

Tobacco taxes and smoking prevalence

26

The average (legal) price of a cigarette

27

GST of 15%

$0.21

Excise tax Wholesale price + retail margin

$0.09 $0.40 2011 New Zealand dollars

Cigarette price projections with 10% tax

29

$0.00 $0.20 $0.40 $0.60 $0.80 $1.00 $1.20 $1.40 $1.60 $1.80 $2.00 2011NZD

$14 pack $40 pack

Tobacco taxes in New Zealand

30

• 1. Will 10% annual tax increases be enough

to reach <5% prevalence by 2025?

• 2. Could growth in the illicit tobacco market

undermine the benefits of tax increases?

How do people respond to increasing price?

31

• More smokers quit (or fewer young people start)
• Smokers cut-down on number of cigarettes smoked
• Response is measured by price elasticity

– International review (IARC 2011): -0.2 to -0.5 – New Zealand study (Tait et al. 2013): -0.47

Smoking prevalence projections

32

2025 0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 2010 2015 2020 2025 2030 2035 2040 2045 2050 2055 2060 Smoking prevalence

No tax increase 5% increase 10% increase 15% increase 20% increase

9.9% 8.7%

Smoking prevalence in 2025

35

8.2% 6.1% 17.9% 18.4%

0% 5% 10% 15% 20% 25%

Non-Māori Men Non-Māori Women Māori Men Māori Women No tax increase 10% increase 20% increase

5%

QALY gains and cost-effectiveness of tobacco taxes

36

Methods – baseline data

• Including the following diseases:

– Ischaemic heart disease (IHD) and stroke; respiratory disease (COPD & LRTI) – Cancers: bladder, cervical, endometrial, head and neck, kidney, liver, lung, melanoma, oesophageal, pancreas, stomach, thyroid.

• All-cause mortality from SNZ lifetables with 1.75% (non-Māori)

and 2.25% (Māori) p.a. ↓ mortality rates to 2026, then 1% ↓ p.a.

• Disease-specific incidence, case fatality, prevalence from range of

sources, brought together with DISMOD to ensure consistency:

– cancer registry, mortality data, HealthTracker, NZ Burden of Disease Study (NZBDS), NZCMS, CancerTrends

• Morbidity incorporated using years of life lost (YLDs) from NZBDS
• Costs in each state from HealthTracker, 2011 $

37

Methods – multistate lifetable

• A multistate lifetable is literally that – a lifetable in which subjects

(proportions of a cohort) can be in multiple states simultaneously

38

Methods – intervention parameterization

39

• ↑Tax → ↑price → ↓prevalence and cigs/day:

– Using price elasticities applied in year of increase only (and in subsequent year in scenario analyses = ‘persistence’ scenario).

• Relative risks for smoking (NZCMS, other) applied to changing

prevalence and cigs/day to calculate population impact fractions (PIFs; aka PAR%), that are then ‘fed into’ the lifetables to de(in)crease disease incidence.

• Difference in QALYs and cost for 2011 population between

comparator and intervention tallied up for rest of their life (max 110 years). 3% discount rate.

Costs and QALYs – all 2011 population

40 PRELIMINARY RESULTS – will change a little with pending improvements. Not for citation

Non-Māori Māori Total Health gain Cost Offsets Health gain Cost Offsets Health gain Net costs QALYs Millions QALYs Millions QALYs Millions 10% tax increase 32,030 (28,650 to 35,330) $-406.2 (-462.7 to - 353.9) 17,550 (15,560 to 19,460) $-154.4 ($-177.0 to $-132.8) 49,580 (45,790 to 53,390) $-560.6 ($-619.5 to $- 504.3) 20% tax increase 61,610

• $777.6

33,430

• $292.5

95,030

• $1,070.1

QALYs gained per capita

41 PRELIMINARY RESULTS – may change a little with pending improvements. Not for citation

Māori Non-Māori RR 10% tax 0.0260 0.0086 3.04

• Strong health inequality reduction:

– in relative terms – but not so much in absolute terms as only about 3 (non-Māori) to 9 (Māori) quality day of life gained per person in the population

• [gain much more for the actual person who quits]

Timing of health gains

43 PRELIMINARY RESULTS – will change a little with pending improvements. Not for citation

• 6
• 5
• 4
• 3
• 2
• 1

1 2 3 4 5 6 2011 2031 2051 2071 2091 2111 QALYs gained Thousands Year

Non-Māori, QALYs gained, undiscounted Māori, QALYs gained, undiscounted Non-Māori, QALYs gained, discounted Māori, QALYs gained, discounted

Timing of health gains and costs averted

44 PRELIMINARY RESULTS – will change a little with pending improvements. Not for citation

• 80
• 60
• 40
• 20

20 40 60 80

• 6
• 5
• 4
• 3
• 2
• 1

1 2 3 4 5 6 2011 2031 2051 2071 2091 2111 Net health cost savings (NZD) Millions QALYs gained Thousands Year

Non-Māori, QALYs gained, undiscounted Māori, QALYs gained, undiscounted Non-Māori, QALYs gained, discounted Māori, QALYs gained, discounted Māori, Health system costs averted, discounted Māori, Health system costs averted, undiscounted Non-Māori, Health system costs averted, discounted Non-Māori, Health system costs averted, undiscounted

Timing of health gains & cost by age

45

• $80
• $60
• $40
• $20

$0 $20 $40 $60 $80

• 6
• 4
• 2

2 4 6 2011 2031 2051 2071 2091 2111

$ QALYs

0-14 yr olds

• $80
• $60
• $40
• $20

$0 $20 $40 $60 $80

• 6
• 4
• 2

2 4 6 2011 2031 2051 2071 2091 2111

$ QALYs

15-24 yr olds

• $80
• $60
• $40
• $20

$0 $20 $40 $60 $80

• 6
• 4
• 2

2 4 6 2011 2031 2051 2071 2091 2111

$ QALYs

25-44 yr olds

• $80
• $60
• $40
• $20

$0 $20 $40 $60 $80

• 6
• 4
• 2

2 4 6 2011 2031 2051 2071 2091 2111

$ QALYs

45-64 yr olds

• $80
• $60
• $40
• $20

$0 $20 $40 $60 $80

• 6
• 4
• 2

2 4 6 2011 2031 2051 2071 2091 2111

$ QALYs

65+ yr olds Legend

QALYs gained (thousands), discounted QALYs gained (thousands), undiscounted Health system costs averted (millions), discounted Health system costs averted (millions), undiscounted

PRELIMINARY RESULTS – will change a little with pending improvements. Not for citation

Future mortality inequalities reduced from 10% p.a. tax 2011-31 – moderately

46

• 5%
• 4%
• 3%
• 2%
• 1%

0% 2011 2021 2031 2041 2051 Percentage difference in Māori:non-Māori SRD for tobacco tax versus BAU Year Men 45-64 Men 65-84 Men 85+ Women 45-64 Women 65-84 Women 85+ PRELIMINARY RESULTS – will change a little with pending improvements. Not for citation

Is this a big impact? BODE3 evaluations

47

266 54,000 99,000 77,000 137,000 2,038,000

100 1000 10000 100000 1000000

Current HPV vaccination 12 year old girls in 2011 10% tobacco tax p.a. 2011-31, total 2011 population followed to death 20% tobacco tax 2011-31, total 2011 population followed till death 25% mandatory salt reduction three processed food groups, total 2011 population 25% mandatory salt reduction in all processed foods, total 2011 population Eliminate all CVD for 2011 population followed till death

PRELIMINARY RESULTS – will change a little with pending improvements. Not for citation

Summary: Tobacco Endgame – Game On!

• There is inherent uncertainty in modelling like this:

– but it is the best we can do to inform the endgame.

• Regular increases in tax:

– are an effective endgame strategy – but almost certainly not enough

• n their own

– are unlikely to be undermined by illicit trade – but are only one component of any endgame strategy to reach the smokefree goal

• More needed, particularly for Māori
• More needed for older smokers (gains from tax for younger)

– Generate big increases in health gain… and save $.

• Need more modelling research on other strategies: eg, outlet

phase-out, denicotinisation & combinations

48

What we’ve studied

Cancer care coordinators

Cancer services a national priority

• ‘continuity of care’
• ‘a single point of contact’
• 2012 Budget:

$16 million for care coordination nurses

patient navigators key workers liaison nurses case managers care coordinators

Cancer care coordinators in stage III colon cancer

Clinical nurse specialist Hospital-based versus ‘usual practice’ Reduce time to surgery Reduce time to chemotherapy Increase chemo coverage Improve patient QOL

Cancer care coordinators in stage III colon cancer

CE threshold $45,000 per QALY

But not the same for everyone

ICER $9,400 per QALY ICER $23,600 per QALY

Cancer care coordinators in early breast cancer

Improve persistence with tamoxifen

Burden of Disease Epidemiology, and Cost-Effectiveness Programme

HPV vaccination

Human papillomavirus (HPV)

• Cancer of cervix, anus, oropharynx
• Genital warts

NZ HPV vaccination programme

School age girls 3 doses In school or through primary care provider

HPV vaccination in girls

Vaccination girls cost effective at:

• $18,800 per QALY

gained for current school-programme (c.f. no HPV vaccination)

• $34,700 for intensified

school programme c.f. above

• Not cost-effective to be

mandatory ($122,500 per QALY gained)

QALYs gained (both sexes)

HPV vaccination in boys

Vaccination boys not cost effective at $247,000 ($119,000 - $474,000) per QALY gained for BOYS added to intensive GIRLS

Knowledge dissemination & translation: BODE3 website = ‘hub’

www.otago.ac.nz/bode3

Publications Presentations Videos Plain language summaries Public Health Expert Blog Rapid assessment tools Data

tools

Main page

Evaluation summaries

Tools

Average female citizen health system costs

$- $2,000 $4,000 $6,000 $8,000 $10,000 $12,000 $14,000 $16,000 1-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90-94 95-99 Publically funded health system costs Age A* = Citizen cost p.a. not last 6 months of life B* = Citizen cost in last 6 months

• f life

Colorectal cancer (CRC; female) excess health system costs

$- $5,000 $10,000 $15,000 $20,000 $25,000 $30,000 $35,000 'Excess' health system costs compared to expected citizen costs (if no tobacco-related disease) Age C - A = 'Exess' costs

• f CRC in first year of

diagnosis E -A = 'Excess' cost

• f being in last 6

months of life if dying from CRC D -A = 'Excess' p.a. cost of having prevalent CRC

Knowledge dissemination & translation: BODE3 website = ‘hub’

www.otago.ac.nz/bode3

Publications Presentations Videos Plain language summaries Public Health Expert Blog Rapid assessment tools Data

What next?

• Evaluations:

– CT screening smokers lung cancer, CRC screening – More tobacco, diet and CVD – We will soon reconvene Advisory Groups

• National Science Challenge 3 and MBIE
• Next HRC Programme Grant

Thanks for the opportunity to present – comments and discussion?

Wider tobacco market: Marlborough (NZ) man selling tobacco seeds on Trade Me (2010) – but probably a very niche pursuit?

76

Smoking prevalence in 2025

77

Illegal price

5% 6% 7% 8% 9% 10%

Base case (10% annual increase in excise) Absorb price change (20% of excise) Added price increase (20% of excise) Stable (1% of market) Increases (+5% per year) Rapidly increases (+20% per year) Very cheap (25% of legal price) Cheap (65% of legal price) Best case combination Worst case combination

8.7% Illicit market share Tobacco industry