Science and Policy Making in Public Health

The example of Blood Safety

Albert FarrugiaSenior Director Global Access, Plasma Protein

Therapeutics Association

2003-2008

Senior Principal Research Scientist, Australian Department of Health and AgeingUniversity of Canberra, Faculty of Applied Science

May 2009

In this discussion, I will seek to make the following points

• Policy making in government is complex and often erratic and confused• While most governments promulgate cost-effectiveness in health care spending,

blood safety measures have been relatively insulated from this measure• Cost – effectiveness in blood safety engages blood experts – AND NO ONE ELSE• Public opinion supports a high level of willingness to pay for blood safety – real

and perceived – and this is only modestly influenced by scientific knowledge• Therefore, measures to further enhance the safety of blood should continue to be

introduced, while focus of scrutiny should be the evidence base for the therapies themselves

Pillars of Australian health policy inputs

E

B

M

CE

PRECAUTIONISM

Contributors to policy choices

• Values

• Interests

• Resources

•Mediated through institutions

•Determined by politics

Volatile mixture -

? Agreement on process

Policy choices are Made at a range of levels within governments, which are “An amorphous mass of interlocking organisations attempting to provide for

the many public service demands of a modern industrial society” and have to be

Negotiated across agencies, which Are dispersed and diffuseDisagree and overlap Pursue contrary understandings of policy objectivesObstruct unacceptable (to them) choicesContest decisions at various levels within themselvesRemake decisions during implementations

As a result - policy making is “inherently difficult, frustrating and only partially effective”

Decision making in the Australian blood system

Areas involved in blood:•Acute Care Division•Regulatory Policy Division•Population Health Division•Etc

Areas involved in blood:•Blood & Tissues Unit•Drug Safety Branch•Devices Section•Manufacturing Assessment Branch•Etc

Decision making and implementation• Interests

– Seek to shape procedures– Intervene at strategic moments– Challenge the validity of administrative decisions

• Implementation– Delivery cannot be guaranteed– Need to ensure understanding and respect– Complex processes demand discretion– Discretion means policy process seldom concluded– Few processes are entirely authoritive– Fewer decisions are quite final

Furler 2001“Health Policy

Electoral Populism Research-based evidence

Selectively exploits and reinforces community (mis)-conceptions

•Value for money•equity•health outcomes•etc

May entail political risk

Furler 2001

Public Service Priorities1 Get government re-elected2 Deliver micoreconomic reform (while husbanding votes)3 Deliver better population health outcomes (without affecting 1 &

2)

throughawareness of mood in marginal electoratesunderstanding of swinging voters’ core valuesknowledge of market research mechanisms

HIV Risks in Repeat Donors in Europe andU.S. versus Australia

Residual Risk per 106 Repeat DonationsIncidence Rate per 105 Person-Years

(95% CI)

DatesCountry

0.51.01.6 (1.0 - 2.3)1/96 - 12/00US

0.50.71.6 (1.2 - 2.0)1/00 - 12/01

1.12.33.8 (2.8 - 5.0)1/96 - 12/00 Italy

1.01.93.2 (2.2 - 4.5)1/97 - 12/99Spain

0.20.40.7 (0.5 - 1.1) 1/97 – NA

(EPFA)

Central Europe

0.40.71.2 (0.7 - 2.0)1/98 - 12/00France

Serology / NAT Estimate

Serology Estimate

Glynn, Kleinman, Wright, Busch. Transfusion 42:966-72, 2002; Australian Red Cross Blood Services data courtesy of Dr. Clive Seed, 7/03

Australia 1/00 - 6/03 0.54 (6 SC in 2.6 million donations) 0.12 (1/ 6.1 million)

The other side of NATSimmonds et al 2002

Staginnus 2006CE of platelet bacterial testing

Australian Government review of alternatives to homologous blood donation

http://www.nba.gov.au/PDF/homologous.pdf

Pre-operative autologous donation should not be promoted, for the following reasons:

· While pre-operative autologous donation reduces the need for homologous blood, any benefit from avoiding the adverse effects of homologous blood has to be balanced against the risks associated with the use of autologous blood or any blood product;

· The chance of receiving a transfusion is significantly increased in those with autologous blood available, magnifying these risks and increasing the costs; and

· In the absence of formal modelling or a properly conducted clinical trial, it is difficult to determine whether the benefits of autologous donation will definitely outweigh the harms.

Cost of Australian Blood Sector

0

2

4

6

8

10

12

14

16

18

$A

us p

er

cap

ita

1990 1995 2000

0.71% of total health care expenditure

29%

Optimal versus maximal safety• Conference - 9/11/96 - by the Medical Advisory Commission to the Blood Transfusion

Council of the Netherlands Red Cross, addressing the issues of 'maximal' versus 'optimal' safety measures for the blood supply.

• Invited were blood transfusion specialists, clinicians, representatives of patient interest groups, the Ministry and Inspectorate of Health and members of parliament.

• Transfusion experts and clinicians were found to advocate an optimal course, following strategies of evidence-based medicine, cost-benefit analyses and medical technology assessment.

• Patient groups depending on blood products, such as haemophilia patients would rather opt for maximal safety.

• Insurance companies would choose likewise, to exclude any risk if possible. • Health care juridical advisers would advise to choose for optimal safety, but to reserve funds

covering the differences with 'maximal safety' in case of litigation. • Politicians and the general public would sooner choose for maximal rather than optimal

security. • The overall impression persists that however small the statistical risk may be, in the eyes of

many it is unacceptable. “This view is very stubborn.”

van der Poel at al Ned Tijdschr Geneeskd. 1998 Feb 7;142(6):285-8

•Cross section of Australians questioned about the importance of costs in setting priorities in health care.

•Generally, respondents felt unfair to discriminate against patients with high cost illness and that costs should not be a major factor in prioritising.

•Majority maintained this view when confronted with its implications in terms total number of people who could be treated and their own chance of receiving treatment if they fall ill.

•Results suggest that

Concern with allocative efficiency, as usually envisaged by the economists, is not shared by the general public

Cost-effectiveness approach may be an excessively simple value system upon resource allocation decision making.

Willingness to pay for autologous donationMoxey at al Transfusion Medicine, 2005, 15, 19–32

0

10

20

30

40

50

60

70

80

90

100

10 10000

Price $A

% w

illi

ng

to

pa

y

Perceptions and preferences regarding PADLee et al Transfusion 38:757-763, 1989

Study group

Study questions Informed Uninformed p

Concern re allogeneic transfusions – 0 to 10 (10 = extreme worry) 6 7 0.75

% selecting 10 28 27 0.73

Median willingness to pay $1100 $1900 0.095

Prefer EPO over PAD %

Yes 44 41 0.545

No 19 23

Not sure 37 36

Prefer PAD despite elimination of all infectious agents from allogeneic blood 56 66 0.202

Yes 56 66 0.202

No 20 17

Not sure 24 18

Prefer PAD even if not recommended by a physician

Yes, definitely 47 44 0.323

Yes, probably 29 30

Not sure 11 10

Probably not 13 12

Definitely not 1 4

24

Evidence-based decision making versus (precautionary) risk management approach

Definitive clinical or epidemio-logical evidence of benefit before application of the measure

Attempt to balance risk prevented by implementation of the measure, versus risk remaining without its introduction, based on incomplete information

The “threat” of pharmaco-economics“What we are seeing, in both Factor VIII and Factor IX is in fact the market has grown rather than

necessarily plasma-derived being used less. So when we were the sole fractionator before recombinant there was a lot of rationing of haemophilic patients, probably as low as - we only probably had about 1.6 IUs per head of population. We now have a target in Australia of 3 international units per hit of population, of which I think plasma will represent 1.8 to 2 and recombinant 1……. But, in essence, I know Canada has gone to a different solution in that scenario. I think it is an expensive solution. I'm not sure clinically it is superior, but it is what it is. I think that in Australia, because we believe in pharmaco-economics, the recombinant products for patients who are well stabilized on plasma-derived there is no evidence that it is a useful thing to do…. because in very few other areas of medicine have we seen such an expensive - such a lot of money spent for such little clinical gain.”

Dr Brian McNamee

Open Forum

Plasma Self-sufficiency in Canada - is it a matter of safety?

National Blood Safety CouncilMarch 29-30, 2001

Blood infections in US hemophiliac birth cohortsCDC survey

HBV (▪), HCV ( ), and HIV-1 ( )▴ ◯ The proportion was zero for HIV after 1984, for HCV after 1992, and for HBV after 1993.

Efficacy of recombinant vs plasma concentrates

Blood 109, 546-551. 2007

28

Variant CJD•BSE emerged as major epidemic in United Kingdom in 1980s and 1990s

– related to feed material used – >180,000 cattle infected– hundreds of cases of BSE in Europe

•exported British cattle •native born bovines fed materials derived from UK

•Magnitude of human epidemic unknown

– early prediction of many thousands of deaths– latest scientific data suggest no more than few hundred

deaths– incubation period very long

•BSE emerged as major epidemic in United Kingdom in 1980s and 1990s

– related to feed material used – >180,000 cattle infected– hundreds of cases of BSE in Europe

•exported British cattle •native born bovines fed materials derived from UK

•Magnitude of human epidemic unknown

– early prediction of many thousands of deaths– latest scientific data suggest no more than few hundred

deaths– incubation period very long

•Transfusion and vCJD

• state of knowledge - 2003

–evidence of transfusion transmissibility elusive

• theoretical considerations suggest possible risk

• scientific data show some cross species transmissibility by blood in

animal experiments

• applicability of animal data remains controversial

• epidemiology has not supported the risk

• state of knowledge - 2009

–evidence of at least 3 transfusion derived vCJD from fresh

components

–most recently putative transmission to hemophiliac

•Transfusion and vCJD

• state of knowledge - 2003

–evidence of transfusion transmissibility elusive

• theoretical considerations suggest possible risk

• scientific data show some cross species transmissibility by blood in

animal experiments

• applicability of animal data remains controversial

• epidemiology has not supported the risk

• state of knowledge - 2009

–evidence of at least 3 transfusion derived vCJD from fresh

components

–most recently putative transmission to hemophiliac

0

5

10

15

20

25

30

Donor deferral policies:

• entirely precautionary (and highly controversial) when implemented

• are they now justified in light of new knowledge?

Donor deferral policies:

• entirely precautionary (and highly controversial) when implemented

• are they now justified in light of new knowledge?

http://www.nba.gov.au/pdf/factor_VIII_IX.pdf

Special Expert Committee on TransmissibleSpongiform Encephalopathies

June 2002

" Although the theoretical risks from plasma-derived AHF are very small, they cannot be saidto totally negligible. It is prudent to recommendthat, as soon as feasible, AHF be made availablein recombinant form, or as a product of apurification process that is proven to reduce prioncontent by at least 7 logs."

“……..The Government has agreed to fund access to recombinant clottingfactors for haemophilia patients. A small number will not be able to userecombinant products and will continue to be provided with plasma-derivedclotting factors. The Government will provide ongoing funding of $80.7million over four years, to be supplemented by the States andTerritories……”

P 57

Reality check• Power is diffuse - no single “decision-making process”• Implementation generates options - and new policy• Process must bend to meet circumstances• Problems vary so policy formulation adjusts over time and

place• Political processes are by nature ad hoc, pragmatic and

confused• Long term financial strategies may need to cede to

immediate political priorities• ELECTORAL CONCERNS CAN OVVERRIDE THE MOST

CAREFUL POLICY PROCESS