‘thresholds’ for Canadian - NOAHE · ‘thresholds’ for Canadian Provinces Claxton K., Ochalek J. and Lomas J. 13/9/2018 . £20,000 per QALY Price = P* £40,000 Cost-effectiveness

Estimating cost-effectiveness ‘thresholds’ for Canadian

Provinces

Claxton K., Ochalek J. and Lomas J.

13/9/2018

£20,000 per QALY

£40,000 Price = P*

Cost-effectiveness Threshold £20,000 per QALY

QALYs gained

Cost

£60,000 £30,000 per QALY

Price > P*

3

£20,000

2

£10,000 per QALY

Price < P*

1

Net Health Benefit 1 QALY

Net Health Benefit -1 QALY

What price should we pay?

Price

Quantity

P*

Q*

Value of the innovation = P*.Q* All value is appropriated by manufacturer during patent


Price

Quantity

P*

Q*

Net harm done at list price Q(L-P*) up to Q* Q.L beyond Q*

L

Value of the innovation = P*.Q* All value is appropriated by manufacturer during patent


• Patent expiry (technology and comparators)

– Price later brands relative to generic version of old

– If not no benefit to the HCS even in the long run

£0

£20,000,000

£40,000,000

£60,000,000

£80,000,000

£100,000,000

£120,000,000

£140,000,000

£160,000,000

£180,000,000

£200,000,000

0 5 10 15 20 25 30

Total value

Manufacturers share

HCS share

52% of surplus (HCS perspective) 68% of surplus (consumption, v=2k) Patent expires and generic entry at t=15 Generic prices are 25% of the brand All prescribing switches to generic Discounted at 3.5%

How is total value shared?

Cost-effectiveness ‘thresholds’

• Norms describing how recommendations are made

– NICE (UK), £20,000 to £30,000 per QALY

– WHO 1-3 GDP per capita (now withdrawn)

• Consumption value of health (demand side)

– Value of a statistical life • Revealed and expressed preferences

– Willingness to pay to gain QALY/avert a DALY • Expressed preferences (contingent valuation and DCE)

• Health opportunity costs (supply side)

– Health effects of changes in health expenditure

PBC 23 GMS

How can we estimate health opportunity

costs?

Ch

ange

in o

vera

ll ex

pe

nd

itu

re

How are changes in

expenditure allocated to

PBCs?

11 PBCs

PBC without mortality

signal

11 PBCs

How does a change in PBC expenditure effect PBC mortality?

11 PBCs

PBC 23 GMS


costs?

Ch

ange

in o

vera

ll ex

pe

nd

itu

re

How are changes in


PBCs?

11 PBCs


signal

11 PBCs


11 PBCs 11 PBCs

% effect of a change in PBC expenditure

on burden of disease (LY)

PBC 23 GMS


costs?

Ch

ange

in o

vera

ll ex

pe

nd

itu

re

How are changes in


PBCs?

11 PBCs


signal

11 PBCs


11 PBCs 11 PBCs



PBC 23 GMS

11 PBCs

Same % effect on burden of

disease (totals from

the other 11PBCs)

PBC 23 GMS


costs?

Ch

ange

in o

vera

ll ex

pe

nd

itu

re

How are changes in


PBCs?

11 PBCs


signal

11 PBCs


11 PBCs 11 PBCs



PBC 23 GMS

11 PBCs

Same % effect on burden of

disease (totals from

the other 11PBCs)

PBC 23 GMS

Measures of QALY burden

of disease

Life years (ONS)

Quality of life

(HoDAR MEPS)

Age, gender and duration

of disease (GBD)

Cost per QALY (life year and

quality effects) Surrogacy

Extrapolation

What are the expected health consequences of £10m? Change in spend Additional deaths LY lost Total QALY lost Due to premature death Quality of life effects

Totals 10 (£m) 51 233 773 150 623

Cancer 0.45 3.74 37.5 26.3 24.4 1.9

Circulatory 0.76 22.78 116.0 107.8 73.7 34.1

Respiratory 0.46 13.37 16.1 229.4 10.1 219.3

Gastro-intestinal 0.32 2.62 24.7 43.9 16.2 27.7

Infectious diseases 0.33 0.72 5.3 15.7 3.6 12.1

Endocrine 0.19 0.67 5.0 60.6 3.2 57.3

Neurological 0.60 1.21 6.5 109.1 4.3 104.8

Genito-urinary 0.46 2.25 3.3 10.6 2.1 8.5

Trauma & injuries* 0.77 0.00 0.0 0.0 0.0 0.0

Maternity & neonates* 0.68 0.01 0.4 0.2 0.2 0.1

Disorders of Blood 0.21 0.36 1.7 21.8 1.1 20.7

Mental Health 1.79 2.83 12.8 95.3 8.3 87.0

Learning Disability 0.10 0.04 0.2 0.7 0.1 0.6

Problems of Vision 0.19 0.05 0.2 4.2 0.2 4.1

Problems of Hearing 0.09 0.03 0.1 14.0 0.1 13.9

Dental problems 0.29 0.00 0.0 6.8 0.0 6.8

Skin 0.20 0.24 1.1 1.9 0.7 1.2

Musculo skeletal 0.36 0.39 1.8 23.2 1.2 22.1

Poisoning and AE 0.09 0.04 0.2 0.8 0.1 0.7

Healthy Individuals 0.35 0.03 0.2 0.7 0.1 0.6

Social Care Needs 0.30 0.00 0.0 0.0 0.0 0.0

Other (GMS) 1.01 0.00 0.0 0.0 0.0 0.0

• Scale of health opportunity costs

• Type of health effects (mortality, survival and morbidity)

• Where these are likely to occur (disease, age, gender)

• Severity of disease (burden, absolute and proportional)

• Net production effects (marketed and non marketed)

• Impact on health inequality

• Affordability and the scale of budget impact

• Elicitation from clinical and policy experts (surrogacy and extrapolation)

• Re-estimated for all waves of data

Recent UK estimates

• Claxton, K., Martin, S., Soares, M., et al.. Methods for the estimation of the NICE cost effectiveness threshold. Health Technology Assessment, 2015; 19(14) (see web page for more materials about this research https://www.york.ac.uk/che/research/teehta/thresholds/)

• Martin S, Rice N, Smith PC. Comparing costs and outcomes across programmes of health care. Health Economics. 2012 Mar;21(3):316-337. • Martin S, Rice N, and Smith PC. Does health care spending improve health outcomes? Evidence from English programme budgeting data. Journal of Health

Economics 2008; 27:826–42 • Claxton, K., Sculpher, M., Palmer, S., et al. (2015). Causes for concern: is NICE failing to uphold its responsibilities to all NHS patients? Health Economics, 2015;

24: 1–7. • Love-Koh J, Cookson R, Claxton K, Griffin S. Who gains most from public healthcare spending? Estimated health impacts of changes in English NHS expenditure

by age, sex and socioeconomic status. Submitted to Social Science and Medicine August 2016. • Lomas J, Claxton K, Martin S and Soares M. Resolving the ‘cost-effective but unaffordable’ ‘paradox’: estimating the health opportunity costs of non-marginal

changes in available expenditure. Submitted to Value in Health, November 2016

Re-estimated for all waves of data

https://www.york.ac.uk/che/research/teehta/health-opportunity-costs/re-estimating-health-opportunity-costs/#tab-2

https://www.york.ac.uk/che/research/teehta/health-opportunity-costs/re-estimating-health-opportunity-costs/
















Alternative approach to identification

• Exogenous elements in funding allocation rules (Andrews et al 2017)

• Estimate all cause mortality elasticities

• Disease specific elasticities for changes in total spend

• Implied all cause elasticities

Claxton K, Lomas J, Martin S. The impact of NHS expenditure on health outcomes in England: Alternative approaches to identification in all-cause and disease specific models of mortality. Health Economics. 2018 Apr 2. Available from, DOI: 10.1002/hec.3650

Estimated elasticity of mortality with respect to total PB expenditure

2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13

Implied all-cause

mortality -0.883 -1.23 -1.364 -1.413 -1.307 -1.531 -1.329 -1.094 -0.914 -0.682

All-cause mortality -0.704 -0.826 -1.004 -0.866 -0.749 -0.985 -0.668 -0.607 -0.642 -0.821 Implied all-cause

mortality

socio-economic

instruments

-1.246 -1.615 -1.372 -1.496 -1.269 -0.795 -0.941 -1.328 -1.386 -1.028

Alternative approach to identification

Martin S. Lomas J and Claxton K. How effective is marginal health care expenditure? Evidence from England for 2003/04 to 2012/13. Submitting to Journal of Health Economics. September 2018

• Australia (Edney et al)

– Cost per QALY of $28,033 AUD ($20,758 to $37,667)

• Spain (Vallejo-Torres et al)

– Cost per QALY of 22,000€ to 25,000€

• Netherlands

– Cost per QALY of 41,000€ (CVD hospital care) (van Baal)

– In submission

• Indonesia

• South Africa

Other estimates using within country data

• UK estimate represents 50% of UK GDP pc.

• Value placed on health varies with income – Published estimates of income elasticity of demand

– Greater proportion of income devoted to health care as income rises

• Assume – UK estimate is correct

– Estimates of income elasticity based on mortality risk

• Assume same for survival and morbidity effects

– Proportionate ‘underfunding’ is similar across countries

• Heath care financing more difficult but donated funds

– Limited data and uncertain assumptions

– Indicative possible implications

Possible implications for other countries

Woods BS, Revill P, Sculpher MJ, Claxton K. Country-level cost-effectiveness thresholds: initial estimates and the need for further research. Value in Health. 2016 Feb 20.

L/M IC = 2% - 56%

M/H IC = 20% - 77%

Canada $28,089 ($26,596 - $33,560 2013 CAN)

• Bokhari et al 2007

– Estimated elasticities for 127 countries

– Effect of expenditure on under 5 and maternal mortality

– Account for endogeneity in health expenditure and GDPpc

– Interaction with measures of infrastructure and donor funding

• Ochalek et al 2015 and 2018

– Re-estimate effect on adult mortality (male and female)

– Population (age and gender), mortality rates (age and gender),

conditional life expectancies (age and gender), total health care

expenditure

– Country specific cost per life year and costs per DALY

– Directly re-estimated for direct effects on YLL, YLD and DALY

Estimates of the effect of expenditure on mortality

Ochalek JM, Lomas J, Claxton K. Cost per DALY averted thresholds for low- and middle-income countries: evidence from cross country data. York, UK: Centre for Health Economics, University of York. 2015 Dec, p. 1-50. (CHE Research Paper; 122).

Direct and indirect cost per DALY estimates

Survival (YLL) Morbidity (YLD) DALY

Based on mortality estimates

Directly estimated

Mortality as surrogate for

morbidity effects

Directly estimated ,or

adjusted

Directly estimated

DALY

DALY 1 X X

DALY 2 X X

DALY 3 X X

DALY 4 X

Estimating health opportunity costs in Canada

Estimated elasticities Canada

Average for high

income countries

Mortality (deaths per 1,000)

Children under-5 -0.3412 -0.3549

Adults females -0.1924 -0.1944

Adult males -0.1928 -0.2000

DALYs -0.2137 -0.1929

YLLs -0.3032 -0.2765

YLDs -0.0294 -0.0246

• Elasticities estimated from country level data

• Province specific mortality, conditional LE by age and gender

• Province specific public health expenditure

• Morbidity burden for Canada (GBD) applied to provinces (mortality burden)



Cost per DALY averted for DALY 1 (2013 C$)

Claxton et al

(-1.028)

Andrews et al

(-0.705)

Bokhari et al

(-0.193)

Canada $19,914 $29,032 $97,321

Alberta $26,060 $37,991 $125,997

British Columbia $19,227 $28,029 $96,042

Manitoba $21,722 $31,667 $104,498

New Brunswick $18,265 $26,628 $90,166

Newfoundland and

Labrador $21,392 $31,186 $104,902

Northwest Territories $52,191 $76,087 $249,536

Nova Scotia $18,002 $26,244 $89,814

Nunavut $41,776 $60,903 $177,375

Ontario $19,606 $28,582 $95,706

Prince Edward Island $16,425 $23,945 $82,939

Quebec $17,936 $26,147 $87,446

Saskatchewan $20,804 $30,329 $99,467

Yukon $30,633 $44,659 $155,899

Woods et al 2016 , $28,089 ($26,596 - $33,560)

• Range of potential values for Canada and for most provinces

is in the region of $20,000 to $100,000 per DALY averted

• A cost per DALY threshold is likely to be less than $50,000 for

Canada as a whole and is likely to be similar across most

provinces.

• A cost per QALY threshold is likely to be similar or lower than

a cost per DALY averted threshold

• A cost per QALY threshold of $30,000 per QALY would be a

reasonable assessment of the health effects of changes in

health expenditure for Canada as a whole.

Summary and recommendations

• Public and private expenditure

– Elasticity of private likely to be similar or higher

– Using public expenditure estimate for pricing avoids undermining

publically funded health care

– May offer additional surplus to patients in private sector

• Estimating mortality elasticities for Canada using within country

data

– Disease specific

– Province specific

• Province specific estimates of QALY burden of disease

• Estimating the effect of changes in health expenditure on QALY

outcomes for each province

– Evidence that tend to underestimate effects using mortality estimates

Considerations and further research

Implications for pharmaceutical pricing

15 years of patent Generic 25% of brand

Discount 3.5%

Pro

po

rtio

n o

f to

tal s

urp

lus

reta

ined

by

HC

S

Cost per QALY ‘threshold’ used for pricing and reimbursement



Pro

po

rtio

n o

f to

tal s

urp

lus

reta

ined

by

HC

S


Discount 3.5%


Discount 1.5%

Implications for pharmaceutical pricing P

rop

ort

ion

of

tota

l su

rplu

s re

tain

ed b

y H

CS



Discount 3.5%


Discount 1.5%


Discount 1.5%

Implications for pharmaceutical pricing P

rop

ort

ion

of

tota

l su

rplu

s re

tain

ed b

y H

CS



Discount 3.5%


Discount 1.5%


Discount 1.5%

Incidence grows at 2% pa 10 years of patent

Generic 25% of brand Discount 1.5%

How should value be shared? P

rop

ort

ion

of

tota

l su

rplu

s re

tain

ed b

y H

CS



Discount 3.5%


Discount 1.5%


Discount 1.5%

Incidence grows at 2% pa 10 years of patent

Generic 25% of brand Discount 1.5%

• Ignore patents protection? – Limited impact on global demand (UK is 3%)

– Don’t believe get any dynamic benefits

• Future innovations are not valuable to HCS

• Surplus wasted in rent seeking

– Do believe are benefits but able and willing to free ride

– Requires compulsory licencing

• Respect current patents (being a ‘good citizen’) – Pay up to temporary monopoly price p* only during patent (differs by HCS)

– P* requires comparison of new brand to generic versions of old ones

– Renegotiate price/rebate on all patent expiry

• Requires considerable rebates to list price

• May require the threat of compulsory licencing

– Different shares retained for different products


• But what should the share for HCS be – Should not be zero!

• Consumer surplus is the point of innovation

• Private capital is not the only contribution to innovation

– Should not be 100% unless

• Future innovations are not valuable to HCS – US market demand

– Diminishing returns (best opportunities have already been exploited)

• Surplus wasted in competitive rent seeking – Does not create new opportunities (shift the production function)

• What additional future consumer surplus is generated by offering more producer surplus now? – Marginal effect of producer surplus on net health effects of future innovations

– Existing but unexploited developments?

– Create new currently unknown areas to exploit

– How does this compare to other policy options • Same resources invested in basic and biomedical science, translational and

evaluative research

• Public franchise of elements of development for licencing


• Limits to list price changes and per product discounts – Reference pricing and parallel trade

– Rebate mechanisms

• Accept harm or restrict access – Political/social consequences and risks of private insurance

• Cost do not determine price, price expectations determine costs

• Respecting patents? – Pay up to p* (but no more) and only during patent (differs by HCS)

– Compare new brands to generic versions of the old

– Renegotiate price/rebate on all patent expiry (compulsory licence)

– Evidence of what the share should be (differs by HCS)

• Competitive generics market is critical – Barriers to entry (e.g., regulation)

– Anti competitive behaviour (collusion and predatory pricing)

– Regulate (generic reference pricing) when competition fails

• ‘Safety authorisation’ requests to the regulator – Avoid strategic use of licencing as a barriers and prevent competition

Critical issues to over come

Documents

‘thresholds’ for Canadian - NOAHE · ‘thresholds’ for Canadian Provinces Claxton K., Ochalek J. and Lomas J. 13/9/2018 . £20,000 per QALY Price = P* £40,000 Cost-effectiveness