1. Everyday Health Economics
for Public Health
Greg Fell
Director of Public Health, Sheffield
@felly500 /
greg.fell@sheffield.gov.uk
2. Contents
1. Basics and introductory stuff
2. Cost effectiveness and ROI
3. Prevention
4. Better value in NHS and social care
3. Acknowledgements for slides
• Prof Chris McCabe
• Dr Chris Gibbons
• Prof Andrew Stevens
• Dr Peter Brambleby
• Prof Muir Gray
• Probably others…..
4. Learning Objectives -
• Basics of health economics. As
applied to health care and health.
• Overview of some theory, some
practical stuff in appraising cost
effectiveness studies, ins and outs of
ROI
• Tips to apply when arguing with folk
about value
5. Caveat
I am not an expert.
I know a little bit (well actually quite a lot,
comparatively) about HE but I am not an
economist and I have had lots of practice
translating all that clever economics.
6. Caveat (2)
The views below are not the views of my
employer or any of the agencies I have
worked with or for
7. Caveat (3)
I am assuming some basic
knowledge of the science of health
economics
This is NOT Health Economics 101
This is NOT a session in the basics
of HE
this is HE that textbooks wont teach
you
8. Caveat (4)
many of the examples here are pertinent
to health care.
Exactly same principles can be applied to
“prevention” or social care
11. • The Cancer programme budget in Bedford is
about £27.5m.
• If more funds are available next year what
would you spend more on (choose one):
• A: not on health care
• B: health care but not on cancer
• C: Cancer Prevention
• D: Diagnosis and treatment
• E: Supportive care for patients and carers?
Why?
12. • If the same budget was available next year,
but you could choose to spend in new ways,
would you spend more on (choose one):
• A: not on health care
• B: health care but not on cancer
• C: Cancer Prevention
• D: Diagnosis and treatment
• E: Supportive care for patients and carers?
Why?
13. • Since the less funds are available next year,
where would you spend less on (choose
one):
• A: not on health care
• B: health care but not on cancer
• C: Cancer Prevention
• D: Diagnosis and treatment
• E: Supportive care for patients and carers?
• F: Cant say or wont say
Why?
14. “A hellish decision is a choice you
have to make that clashes with
personal values, or has predictable
adverse impact on self or others.”
15. “The hottest places in hell are kept
for those who reserve judgement in
times of great crisis”. - JFK
16. Why do we spend so much on poor
value things??
Ash Paul (on twitter) asked me
over weekend something along
lines of "why do we spend so much
on cost ineffective stuff - like
telehealth, proton beam and robotic
surgery"
17. Why do we spend so much on poor
value things??
• my answer was (in 140 characters)
– indifference to overall cost among clinicians
– Moral hazard
– clinicians want to "do the best" for patients
– commercial interests / vested interests in
status quo
– stunning lack of understanding of incremental
cost effectiveness and how it plays out in the
real world....
which about sums up most of my daily life....
19. Why it matters that we spend a
lot on poor value stuff (beyond
the obvious answer of financial
and accounting)
20. The Don (not THAT Don) on the flat
of the curve
http://www.kingsfund.org.uk/audio-video/don-berwick-implementing-new-models-care
http://blogs.bmj.com/bmj/2015/03/23/richard-smith-flat-of-the-curve-healthcare/
Ever spiralling spend (is
normally on lower
value)
Crowds out more
valuable social
investments
Decent housing,
nutrition, education etc
21. Not just modelling
No economic ‘analysis’ – no literature,
costs or outcomes – HE helps to make
qualitative decision
Presentation of basic costs and/or effects
based on systematic review
Critical evaluation of existing economic
evaluation literature (3a: reverse
engineering of published analyses)
Original (de novo) economic modelling
1
2
3
4
Dr Chris Gibbons
23. The basic problem for the NHS
• You are the CCG Accountable Officer
• You have got an uplift in your budget of £1m
• It is recurrent.
• What should you invest in
– Neonatal intensive care – keeps 2 babies alive
– Approx 300,000 patients to be treated with statins who
were not previouly
– 30 patients to get the newest lung cancer drug that will
extend life by approx 4months per patient
– Better asthma inhalers for everyone
– That new indication for anti VEGF - 250 patients
– 10 salaried GPs.
– A new urgent care centre
Back to contents
24. The basic problem for Local Govt
• You are the LA Chief Exec
• You continue to stare down the barrel of austerity
• What should you disinvest in
– Childrens social care – helps keep children safe and
out of care system
– Home care – helps people stay safe, well and
independent
– Road maintanence budget.
– Selective licencing programme – directly impacting on
housing safety and quality in private rented sector
– Infrastructure in community and voluntary sector.
Provides backbone of social prescribing.
Back to contents
25. Economics in 3 lines
Dr Richard Richards, Assistant Director of Public Health
1. Resources are always scarce
2. Therefore choices have to be made
3. Every choice represents a lost opportunity to
do something else
• We ration organs very carefully indeed.
Organs are scarce, therefore difficult decisions
need to be made about how to allocate. This is
no different to allocating finances to different
competing health care programmes
26. important lessons from
economic theory
1. Diminishing marginal returns
2. Average v incremental
3. Opportunity cost
4. Priority setting
5. Population impact
6. Innovation and shiny toys
29. Average cost of picking up Colon Cancer through
screening - FOBT. US study. American Cancer Society.
no of rounds to try to pick up. total cases detected total costs average costs per case detected
1 65.9 77,511.0 1,175.5
2 71.4 107,690.0 1,507.4
3 71.9 130,199.0 1,810.8
4 71.9 148,116.0 2,058.9
5 71.9 163,141.0 2,267.7
6 71.9 176,331.0 2,451.0
Back to contentsBack to intro and basic concepts
30. But what if we look at the marginal cost of additional
testing rounds for picking up more cases
no of tests
Marginal benefit of the additional
test (number of additional
cases picked up
marginal costs (of
additional test) Marginal cost per case detected
1 65.9 77,511.0 1,175.5
2 5.5 30,179.0 5,487.1
3 0.5 22,509.0 48,932.6
4 0.0 17,917.0 465,376.6
5 0.0 15,025.0 4,695,312.5
6 0.0 13,190.0 43,966,666.7
use the term marginal to refer to the fact that we rarely start from scratch / a blank sheet.
More often we are concerned with the question 'what is the value of adding or removing one or more
components' in terms of the added outcome.
thus in the example above, one round of screening test will pick up 65 cases, at a marginal cost per
case of 1175
6 rounds will pick up 72 cases at marginal cost per case of 43m…..
31. no of
roun
ds to
try to
pick
up.
total cases
detec
ted total costs
average
costs
per
case
detec
ted
Marginal benefit
of the
additional
test
(number of
additional
cases
picked up
marginal costs
(of
additional
test)
Marginal cost per
case detected
1 65.9 77,511.00 1,175.50 65.9 77,511.00 1,175.50
2 71.4 107,690.00 1,507.40 5.5 30,179.00 5,487.10
3 71.9 130,199.00 1,810.80 0.5 22,509.00 48,932.60
4 71.9 148,116.00 2,058.90 0 17,917.00 465,376.60
5 71.9 163,141.00 2,267.70 0 15,025.00 4,695,312.50
6 71.9 176,331.00 2,451.00 0 13,190.00 43,966,666.70
New COPD drug.
Add into coctail of drug for treating unstable COPD.
What is the incremental cost and incremental effectiveness (?reduction in
acute admits) from adding this drug into a treatment regimen.
32. IVF no of cycles in QALYs
Duncan Cooper
NICE IVF CG 2013
33. And in metrics that are more
readily understood
Duncan Cooper
NICE IVF CG 2013
34. Opportunity cost
The opportunity cost of any
choice is the value of the best
alternative choice that you
have to sacrifice in order to
pay for it
35. Locally we fund ONE cycle of IVF
• NICE recommend 3
• We fund 1.
• Opportunity cost of moving to 3 cycle IVF = £0.5m
• 100 PPCI
• 8 midwives
• 30% increase in funding to stop smoking services
• 16 classroom assistants
• Every investment should be considered in the
context of the commensurate service that will be
displaced.
36. An aside – but an important one
as it burns your budgets like
theres no tomorrow
Innovation – old innovations, new
innovations.
Paying for innovation
technical (biotech/pharma/IT / AI / genomics)
service-level
knowledge-driven (ie changing how we think)
37. Here are two definitions.
• a) From “shiny new thing”. usually, but not
always involves new technology.
• b) To disruption in the established way of
doing things that improves population
relevant and / or individual outcomes in a
way that is cost saving, cost neutral or
objectively cost effective from a
commissioner viewpoint.
38. We spend a vast sum on “innovation”
in the hope of solving problems
here is the story of the curious case
of the Bradford Robot…..
39. Example The Da Vinci Robot.
• New?
• Radical?
• Game changing?
• c£2m fixed cost, £1500 per
procedure expendables.
• Of course you should
expect good evidence of
substantial improvement in
prostate cancer survival
and reduced medium term
complications
40. There is no such evidence.
There is US evidence that the existence of
“the robot” is influencing patient decisions
more towards radical prostatectomy when
those same patients might have
historically elected for watchful waiting
(with no poorer outcomes as a result)
41. All that glitters is not gold
Conclusions and Relevance During its initial national
diffusion, MIRP was associated with diminished perioperative
patient safety. To promote safety and protect patients, the
processes by which surgical innovations disseminate into
clinical practice require refinement.
43. It kind of, sort of, works out at
annual volume n=150 or so
see p71
more costly, slightly more
QALYs (tho no survival benefit,
assumed survival based on
modelled data about margin of
removal), HIGHLY sensitive to
cost, which is highly sensitive
to volume done (sweating
asset)
44. And it is now NICE recommended –
if n=150
Commissioners should ensure that robotic systems for the surgical treatment of localised
prostate cancer are cost effective by basing them in centres that are expected to perform at
least 150 robot-assisted laparoscopic radical prostatectomies per year. [new 2014]
45. What is “n” in your local centre?
Has your local centre thought about
fixed and variable costs.
Have you thought about payments
and tariff.
46. A Da Vinci robot is £2m capital +
tariff uplift on consumables
47. Hip Arthroscopy
• Impingement or labral tear
• Alternatives are pain management physio or
replacement
• NICE IPG – poor quality evidence but enough
to recommend. Taken as green light.
• HA = £3.5k (JRS = £5k)
• In 2017 RCT – improvement in pain scores of
6 points (0-100 scale) against physio.
• Worth it?
48. Some further examples of innovations
• Primary PCI
• Thrombolysis in stroke
• Telemedicine and Telehealth
• Assisted living technology
• Expensive new lipid agents PSK9s
• Renal denervation for resistant
hypertension
49. Shiny things take £ and distract
attention
• ……..from the things that make MOST difference
– Speed of PPCI / system to sort this
– stroke thrombolytics & HASU vs primary prev
– Aggressive implementation of lifestyle intervention – stop
smoking!
– Telestuff – evidence and system in which you drop the telestuff is
chaotic
– Optimisation of medicine in Heart Failure patients!
– 10% risk vs under statinisation of high risk
50. The areas where we have
TRANSFORMED outcomes in last
20 years
• Cancer
• CVD
• Renal?
• System development
• National standards. Relentlessly
implemented. Standardised approach
• Products have helped. Systems have
MATTERED
51. Paying for innovation
• From lower value interventions in the
same programme area. PBMA
• Period.
• Unfair that glaucoma patients pay for
innovation (by rote of poorer care /
services / lack of investment) as a result of
req to invest in AMD
• Unfair that asthma patients pay as a result
of requirement to invest in Cancer care
53. Oh really? • is enalapril + 10mg TD dose etc
the best comparator in this group,
• composite outcomes - all the
normal statistical jiggery pokery
that goes with them
• powered to detect difference in
composite outcome
• 5 odd % abs risk difference
(ARD) NNT = 20, but this is the
composite
• c3% ARD in deaths (power?), but
of the deaths a relatively low %
from CVD causes....
• risk of hospitalization reduced by
21%
• this is a relative risk- that will be
misplayed in the marketing spin.
ARD please??
http://www.nejm.org/doi/pdf/10.1056/NEJMoa1409077
56. Epidemiology
• Incidence / prevalence
• is it a first line thing.... in which case
incidence important
• or a "in the pathway" thing... in which case I
and P important
• which sub groups, or all
57. • What is the PICO
• Incremental effectivness and cost. Including
implementation costs
• compared to what
• NNT / ARD - risks and benefits
• time frame
• Perspective – cost and return to who
• common sense - does the clinical and cost data
feel like the real world
58. Five handy questions to ask the
Professor
• Do you believe that NHS is cash limited
• Do you think there should be objective
evidence that benefit outweigh harm prior to
introduction
• Do you believe that should be objective
evidence re cost effective prior to intro
• Are you okay that others will be denied care if
the evidence suggests that the introduction of
the treatment is a net cost for the nhs
• Can you identify which group you'd like to
deny the care of
59. The “This is cost saving”
business case
If I had a pound for every time….
5 things to say
60. “this is cost saving”
A note on ‘invest to save’ proposals
• The world is full of proposals that start with ‘if only we made this
investment, then we could save’. The savings rarely come to
fruition. Reality of assumptions, differential timing of investment
and savings complicates further. Doesn’t address the
fundamental problem. Unmet need.
• ‘Invest to save’ proposals only if:
1. clear and good q evidence re clinical and cost
effectiveness - it could work
2. Clear business case demonstrating is HAS worked
somewhere else
3. Clear proposal - applying Population, intervention,
Comparator, outcome
4. Management capacity and the right incentives and levers
to actually make it work
5. A clear plan for realizing any savings in hard cash
61. “But its an effective treatment, you
MUST pay for it”
Five questions to ask
1. Do you believe that NHS is cash limited
2. Do you think there should be objective evidence that
benefit outweigh harm prior to introduction
3. Do you believe that should be objective evidence re cost
effective prior to intro
4. Are you okay that others will be denied care if the
evidence suggests that the introduction of the treatment is
a net cost for the NHS
5. Can you identify which group you'd like to deny the care of
62. Innovation and economics. The
opportunity cost of innovation
JAMA November 26, 2014 Volume 312, Number 20
Do good
First do no harm
Yes of course
What about the harm done
to those that bear the
opportunity cost of
marginal benefit high cost
innovations
http://newsatjama.jama.com/2014/09/17/jama-forum-high-tech-care-can-save-lives-but-it-also-may-
create-incentives-that-result-in-lives-lost/
64. Ticagrelor
• Effective
• Highly cost
effective – c£2,500
/ QALY
• NICE TA approved
in Unstable Angina
and MI
• NNT – CV Death
c55
• The new clopidogrel
• The new (new)
aspirin
65. Int J Clin Pract, November 2013, 67, 11, 1210–1212
Gain in life expectancy?
CV events?
CV mortality ?
66. • The evidence that it is better in UA is equivocal.
– Forrest plot - page 654 of the online supplementary
appendix
• It is better than clopidogrel in MI.
• Implementing this medicine in the UA / MI population
will cost Bradford B&A c£1m
• 5 less deaths
• It will shift crude CV mortality from 2.37 / 1000
registered pop to 2.36 / 1000.
• And add 0.02y to life expectancy.
• So it is innovative, effective and highly cost
effective…..but does it make a difference to
population relevant outcomes.
• And is it worth it.
67. Population impact of Herceptin
• 5.5 days of added life in women
• At £182m
Richard Richards
68. “But its not fair to measure whole
pop outcomes – only in the treated
pop”
• It should be only impact in those treated
with that thing
• Yes sure
• But it's the pop more broadly that bear the
opportunity cost
• is that fair?
• Which is least unfair?
71. Fundamental principles of priority setting
1. Prioritization will be adopted as the methodology for
decision making (often referred to within commissioning
organizations as the ‘primary of prioritization’)
2. Stay within budget
3. Only invest in treatments which are cost-effective
4. Optimize health gain
5. Do not allow third parties to dictate priorities
6. Do not fund one individual if others with the same clinical
need cannot be funded
7. Do not fund treatments of unproven clinical effectiveness
unless it is in the context of well-designed clinical study
8. Social factors should not be taken into account at the
level of the individual
Daphne Austin IDEAL
72. The theory of priorities Vs the practice
– disinvestment this time.
Decisions about whether to decommission
services may be based largely on cost and
government intervention rather than
effectiveness, safety and cost-effectiveness.
principles by themselves won't get you out of a fix
doi:10.1186/s13012-014-0123-y