Examination of Witnesses (Questions 63
- 79)
TUESDAY 13 NOVEMBER 2007
DR TIM
ENSOR, MR
ALEC CUMMING
AND DR
SAM ADJEI
Q63 Chairman: Good morning to you
gentlemen. Thank you very much for coming in to help us with our
second session of public evidence on our inquiry into maternal
health. I wonder, Mr Cumming, first of all, if you could introduce
yourself and your team and then after that it might be helpful
if you gave us a brief introduction to what Immpact has actually
been doing and how it is set up.
Mr Cumming: Thank you very much,
Chairman. I am Alec Cumming, the Chief Executive Officer of the
Immpact project. I have been in that position since 2004. Prior
to that I was chief executive of an NHS Trust and had worked in
the NHS for 30 years. On my immediate left is Dr Sam Adjei who
is the Chief Consultant to the Ghana Health Service but who worked
with the Immpact project as a visiting professor in international
health at the University of Aberdeen until last year. So we have
had a long close association with Dr Adjei. On my far left is
Dr Tim Ensor who is a health economist who works with the Immpact
project. He is based in York and also works with the Oxford Policy
Management Group, but in this context he is here to represent
the experience and interests he has with Immpact.
Q64 Chairman: Thank you for that.
If you could then give us a brief history of Immpact and what
you have been doing.
Mr Cumming: Immpact was established
in 2001/02. It was set up with funding from the Gates Foundation,
from DFID,[1]
from USAID[2]
and from the European Community, with total funding amounting
to £20 million. The aims of Immpact were to identify evidence
of what is effective or what would be effective in reducing maternal
mortality and also to enable that effectiveness to be measured
by ensuring that we can actually count the number of women who
die because the statistics were, and remain, very poor in this
area. So it was established in 2001/02 to look for evidence of
what is effective and during the subsequent five years we have
worked extensively particularly with three countries, Ghana, Burkina
Faso and Indonesia, to assess the effectiveness of strategies
implemented in these countries and to identify universal messages
from these strategies. We are now in a position to share that
evidence which I guess is why we are here.
Q65 Chairman: Thank you very much.
That is extremely helpful. As you know, the more you look at this
issue the more people say that having an effective health infrastructure
is a crucial part of it. DFID, where possible, does provide budget
support to governments towards things like providing their own
health service and education. How effective is budget support
in delivering, and how can you ensure that there is a direct correlation
between supporting the health service and improving maternal health?
Mr Cumming: I think my colleagues
will answer that in more detail. I would reinforce the desirability
of supporting health systems. It is quite clear from the evidence
we have gathered that the key to improving maternal health is
to improve health systems. Maternal health depends on good operating
health systems and therefore, wherever possible, the best route
to improving maternal health does lie in systems support. That
cannot apply in every country's circumstances, but wherever possible,
as DFID indeed do, we believe that systems support is the correct
approach.
Dr Ensor: Obviously there are
challenges with implementing systems of budget support, particularly
with regards to the monitoring. The monitoring tends to be much
more general than the monitoring you would get with project support.
The key point that Alec made is that maternal health is something
that affects the whole system. It is very difficult to set up
vertical programmes that are only targeted at maternal health
outcomes because maternal health tends by its nature to be something
that covers the whole system. It is very difficult to set up individual
projects that are only focusing on maternal health. Budget support
for maternal health makes a lot of sense. The problems occur with
monitoring budget support for maternal health in the same way
that problems occur with any kind of budget support or any other
health or indeed any other kind of programme. That is where the
systems with the governments that we are working with are just
not adequate to deliver the kind of broad, independent monitoring
indicators required to provide the information on how the money
is being used. In addition to the narrow maternal health monitoring
indicators that are needed to monitor the outcome of maternal
health programmes I think it is also important to put in place
systems that allow much more general monitoring of public expenditure.
If you do not have a system of public expenditure reviews that
report on how money is spent and then used throughout the system
it is very difficult to ensure that you are going to get good
outcomes at the end of it. I think the point here is that in order
to make budget support work you need both the independent indicators
to monitor maternal health and you also need the much more general
public expenditure framework and monitoring indicators to allow
the system to function properly.
Q66 Chairman: Dr Adjei, I think DFID
were involved in supporting the health service in Ghana. You have
had direct experience of that. How effective were they?
Dr Adjei: DFID supported us very
effectively. I must say at the outset that we at the country level
tend to look at the health sector as one. We do not distinguish
between the interventions and what is called systems because when
you do that you create two health political parties, one the interventionist
group and the other the systems group and it is not particularly
helpful to us at a country level. We are interested in the strategies
for service delivery and in integrating those strategies as much
as the systems that are required to support the delivery of the
interventions. We like to look at them as a whole. I think that
is how we have been working with DFID. We would discuss the health
system in the context of what are the priority interventions and
then what are the important systems issues relating to technical
resources or planning or even addressing poverty and looking at
the resources that are required for women to access services.
It has been quite effective in the sense that we can look at all
the interventions that are critical and look at what the support
system is that is critical to carry the interventions forward.
Over the years we have had these health sector reforms going on
and we have been dialoguing together. We have set up the systems
to be able to track performance from the district right through
to national level in terms of how the interventions are performing,
like immunisation or access to maternal and child services, as
well as tracking the human resource needs that are required for
these, tracking logistics, tracking the financing and how funds
are dispersed together as a team and we do independent reviews
on an annual basis. We have monthly discussions with our donor
partners, including DFID, we have business meetings on a quarterly
basis and twice a year we have major summits where we bring all
the donors and the governments and other sectors that have an
interest together to look at the evidence and review the evidence
that has been generated in the country. We even field independent
review teams to look at what our performances have been in the
previous years and bring this to the table.
Q67 Chairman: You are talking about
the Ghana Health Service. In the evidence you say that the release
of funds by DFID has been unpredictable and erratic, that DFID
closed its health offices in Ghana without discussion and this
created a vacuum in the policy dialogue and deprived the health
sector of good quality technical support.
Dr Adjei: This has been the problem
within the past few years. Where it has been very, very difficult—
and this came up during our review and our dialogue and our discussions
with a view to reducing transactional costs— is that DFID
closed its office in the country without discussion and the Ministry
was really concerned about that because they felt that having
DFID at the table and discussing together was very important for
us, passing on that responsibility to another government and it
seemed not to have worked very well. So we were concerned about
that. Now is the budget period and so this week we are having
a meeting with all our donors, but it has always been difficult
to confirm the monies that are coming in. People will pledge and
say okay, we will give you so much, but when the time comes we
have not really had the releases that we have been promised or
been expecting over the past two or three years. It has been quite
erratic over the last couple of years. It has been improving over
the past few months, but that has been the problem.
Chairman: I think that is something we
may want to ask DFID about themselves directly when we have them
here to give evidence.
Q68 Hugh Bayley: I am interested
in looking at what the priorities should be for DFID in terms
of strengthening the health system as a whole or providing specific
programmes and interventions on maternal and child health. If
you look at their published statistics, overwhelmingly their spend
is on health systems, excluding budget support. They spend a sum
of almost £400 million on health overall, but only £16
million is spent on maternal and newborn health. In your evidence
you say "the temptation to target resource through vertical
initiatives, experienced by our Ghanaian colleagues, needs to
be resisted by DFID and as far as DFID can influence them, by
other donors."[3]
It seems to me that DFID has the right policy, but you are saying
it is not happening on the ground.
Mr Cumming: The reason for expressing
ourselves so strongly is that maternal health, as I said earlier,
does depend on a functioning health system. You cannot provide
emergency obstetric care without having blood banks, without having
decent transport for patients and without having sufficient hospital
beds for emergency cases. These are all independent indicators
of a functioning health system and it is not impossible, but it
is difficult to see, if the governance of the country is adequate,
that there is a case for vertical intervention in these circumstances.
The temptation clearly is that you can guarantee returns if you
put money in for a specific purpose and monitor the achievement
of that purpose, but the experience of our colleagues in Ghana
and elsewhere is that that undermines the efforts of local health
services to develop the whole service and distorts priorities.
Although clearly more is needed to be spent on maternal health,
in our view and in our experience that should be in the context
of supporting health systems, where the governance of the country
permits the spending to be tracked and DFID to have confidence
that the money is being properly spent.
Q69 Hugh Bayley: You seem to be saying
80 % of the money should be channelled through the government
and taken from the basic health system, but you may have some
particular initiatives in a particular clinical discipline or
other that you would want to support. The last time the Committee
was in Ghana was probably about four or five years ago and we
visited the Kintampo Health Research Centre just before the tragic
death of the technical director, Paul?
Dr Adjei: Paul Arthur.
Q70 Hugh Bayley: He was an inspirational
doctor. They were doing a lot of work on very, very cheap interventions,
high doses of Vitamin C, which had a big impact on neonatal morbidity
if I remember correctly. Are those lessons being taken on board?
Have those lessons been fed in to the protocols for maternal and
child health in Ghana? Do you need targeted funding for relatively
cheap but effective interventions of that kind or would you still
say "put the money in the health system and let the Ghanaian
health system decide the priorities"?
Dr Adjei: The maternal assessment
on Vitamin A will be completed next year. On the child health
side, that is part of the overall package of interventions that
we deliver to children. It is integrated as part of the primary
healthcare package of interventions. We like to look at it as
a whole so that the districts get the funding to organise how
they deliver that without somebody coming in with the money and
saying they should go out there and just give Vitamin A to the
kids. The districts can organise it together as part of their
whole healthcare delivery arrangements. That is how we would like
to have resources organised.
Q71 Hugh Bayley: Has the government
or the Department of Health in Ghana published any reports on
the impact of spreading out this therapy more widely?
Dr Adjei: The annual review process
will describe in detail all the interventions and how they are
carried out and it will take measurements on implementing child
nutrition. We do have evidence to show that the nutritional status
of children has improved for a couple of years now.
Hugh Bayley: If you could share it with
us that would be helpful.
Q72 Ann McKechin: I want to continue
on this issue of integrating vertical funding with overall health
systems. The Global Fund to Fight AIDS, Tuberculosis and Malaria
obviously takes up a huge amount of donor funding. To what extent
do you believe that government and donors alike are sufficiently
alert to the need to integrate this properly with maternal health
systems, and what has been your experience in countries such as
Ghana as to whether or not there have been any problems about
that issue of integration?
Mr Cumming: I would reiterate
that insufficient resources have been made available to tackle
maternal health. It is the single biggest measurable difference
in health status between the developed and the developing countries.
We clearly do not want to get into a debate of HIV/AIDS versus
maternal health, but maternal health above all requires functioning
health systems and is the least susceptible of all the conditions
to vertical funding. In that sense its interests may not coincide
with interests in tackling the specifics of HIV and AIDS and that
is a problem because these priorities are set externally by governments
and by donor agencies when they give money for these specific
purposes and to that extent it is a distortion, one would have
to say.
Q73 Ann McKechin: I wonder if Dr
Ensor has any comments about how integration of these systems
is working in practice.
Dr Ensor: International agencies
are generally quite good at working together at one level. They
are pretty good at working together in investing in primary care.
Where I think that international agencies are much weaker at working
together is at the hospital level. One of the reasons for that
has been that primary healthcare has become such a political necessity
to invest money in and rightly so because that is where most diseases
can be tackled, but as a result the secondary level tends to be
a bit neglected. This maybe does not directly answer your question
because you are asking about general donor co-ordination perhaps.
My observation would be that certainly in a couple of countries
in Asia where I have been working the emphasis from the donor
side is all about putting money into the primary level. The government
is left on its own to try and do the best it can with the secondary
care level.
Q74 Ann McKechin: Would this be one
of the reasons why mother to child transmissions of HIV, for example,
are still very high, because there has been such an emphasis on
the primary care level rather than at delivery?
Dr Ensor: It could be. The other
related observation to this is that I have observed that a number
of development partners, possibly DFID in some countries but certainly
some of the other European development partners, have then begun
to realise that coverage for catastrophic care is important, that
coverage for things like Caesarean sections and other emergency
obstetric needs are important and they have begun to think how
they can develop almost parallel systems of financing and sometimes
delivery. So the government is struggling along trying to develop
its district hospital structure and you end up with a parallel
structure. I think that is an important issue for the development
community. Bangladesh is a good example of this where there has
been far too little focus on that level by developing partners.
Q75 Mr Crabb: Let us move on to the
issue of user fees for maternal health services. What complementary
policies should donors like DFID be promoting and supporting in
those countries which are looking to abolish user fees in order
to ensure that lost revenues are replenished?
Dr Ensor: It is very difficult.
When we are looking at user fees we should not only focus on the
fees that are faced by the user once they have reached the health
facility. It varies from context to context, but in some contexts
by far the greatest burden on households is the cost of getting
to the facility, having to give up time, having to travel. Nepal
is a very good example of this where studies have shown that 60
% of the cost is in getting to the facility. Any policy for free
delivery has to focus on the main financial barrier. In Nepal
it is transport. In other countries it will be other things. A
policy needs to begin with the biggest barrier. I think the second
part of that is that in a resource-constrained environment it
is very tempting to target and probably some sort of targeting
is almost inevitable. You cannot implement a policy that allows
free delivery in all circumstances for everyone immediately. However,
the type of targeting is very important. The experience with individually
based targeting where you try and work out who is rich and who
is poor on an individual basis does not have a very good pedigree.
It has worked more or less in some countries, but in many contexts
it has not worked well. I think the experience from Immpact and
from elsewhere suggests that in designing a careful policy of
targeting there is almost a hierarchy of targeting that begins
by looking at the poorest areas and targeting those areas with
universal obstetric care. That will mean the rich in those areas
benefit, but the cost of targeting areas tends to be much less
than if you try and work out at an individual level who is rich
and who is poor. The second part of it is that for emergency obstetric
care individual targeting does not work well at all because there
is plenty of evidence that shows that even the non-poor can be
pushed into poverty as a result of large charges. Our evidence
from Indonesia, for example, shows that without free emergency
obstetric care at a couple of district hospitals 13 % of users
would have been pushed into poverty if they were not exempt. They
were the near poor that would have been pushed into poverty. The
beginnings of any user fee policy must be, firstly, to look at
the biggest financial barrier and, secondly, to look at the type
of targeting, and then thirdly, which is perhaps where you started
from, to ensure the resource flow. Maybe Sam would like to say
something about this on Ghana. The Ghanaian policy for providing
free delivery care failed almost entirely because the funds that
were promised at the national level were not properly channelled
to the facility level.
Q76 Chairman: Have you looked into
the role of providing, for example, maternal grants, in other
words to give a woman who is pregnant a sum that would help her
to cover transport or other costs? Has that been looked into?
Dr Ensor: That is interesting.
One of the problems with that is that the grant is not used for
the right purposes. However, the Nepal intervention is actually
very interesting in this area. The cost of transport is around
60 % of the overall cost to users. The Nepal Government a couple
of years ago developed a policy that pays a cash amount to every
woman that reaches a health facility. If she gets to the facility
she gets money in her hands and it offsets the cost of transport.
A full evaluation of that is yet to be done, in fact it is ongoing.
Early results from the management information systems suggest
that skilled attendance is increasing. We will wait to see whether
it is directly caused by the system of grants to women.
Dr Adjei: User fees account for
almost 20 % of their expenditure on health and if it is to be
abolished then you have to find a way of replacing it because
that is a minus of revenue for the hospitals using the procurement
of grants and essential supplies. We started off with an exemption
policy and a lot of the units supported that and we were paying
from the debt relief money, but that became unpredictable. We
were running out of funds from the debt relief money to pay for
the exemptions. Now we have started an ambitious health insurance
system so that we can get women enrolled into the health insurance
system as a way of covering their costs for deliveries when they
come in. Quite a number of them get exempted because of the nature
of the laws. You can get exempted, but you have to go and get
enrolled and get a card and then you are able to access insurance
funds to cover your costs. One of the interesting observations
is that the women come to the paediatric clinics four times and
you have a coverage of about 80 %, but less than 50 % come to
deliver. The difference for a long time between the two was the
deliveries that are charged to it. Now that we are taking out
the charge for the delivery we are hoping the utilisation rate
will go up because it is within the same facilities that the prenatal
and the deliveries take place. If they can come four times then
perhaps they would come for delivery. We are already beginning
to show that the exemption policy has increased uptake. There
are still issues about the quality of the services and some cultural
factors that we have to deal with as part of the overall policy
of removing fees. We will have to go with information and education
and ensure the poor have the confidence to use those services
because if they go there and they are discriminated against because
they are poor then they will not go even if the fees are removed.
I guess the policy that I am talking about is support to ensure
that other mechanisms come into play to provide for the cost of
removing the exemption from the woman of having to pay.
Q77 Mr Crabb: What kinds of costs
are built in to the user fees? Do they tend to be made up of drugs
costs, other supplies, salary costs?
Dr Ensor: It varies considerably.
If we are talking just about the facility costs, then the salary
costs might be anything between 30 and 50 %, the drugs costs might
be another 30 % and then the rest is the overheads and that is
a supply side cost. In addition to that you have whatever demand
side costs there are, such as transport, which, as I have indicated,
can sometimes be the dominant cost in certain contexts.
Q78 Mr Crabb: Are you aware of any
work that has been done to try and identify average costs for
a birth in the developing world?
Dr Ensor: Yes. It varies enormously.
The figures that I have seen show overall aggregates fall by continent
and also for individual countries. The supply side cost is between
perhaps $30 and $100 per birth. Obviously it is much higher where
there are complications.
Q79 Sir Robert Smith: I want to pursue
a bit more the issue of cash transfers or vouchers. In societies
where women's status may not be that strong is there a concern
that whilst the cash is made available to them they do not then
get the ability to control it in that direction? Do vouchers have
an added advantage in that they are for women?
Dr Ensor: There are pros and cons.
Clearly with just giving cash to a family you run the risk that
it will not be used for the right things. If you can somehow condition
that cash, for example in Nepal you do not get the cash until
you reach the facility, then that guarantees that at least the
woman has got there and it is basically mitigating the cost of
transport and provides some extra incentive. Vouchers lose that
problem. In certain contexts I think vouchers can be appropriate.
The problem with voucher schemes is that often they require the
development of a parallel infrastructure, ie administration, that
adds a lot to the cost. There are a number of examples. Bangladesh
has a system of vouchers now for reaching a facility and they
have three levels of committee which, firstly, filter the women
because it is only targeted at poor women, so they have to decide
who is poor, and then once the woman has received the care they
allow the transfer of resources from a fund to the facility and
also to practitioners in certain cases. With all of that you are
adding an extra parallel structure. It is obviously not sustainable
beyond the end of project funding unless somehow that structure
can be absorbed within government. Sometimes— and I think
this is true in Bangladesh— the parallel structure has not
been created in order to serve the needs of the poor or to develop
competition, but it has basically been set up to circumvent what
is seen as a very bureaucratic and unyielding government system
of financial distribution, so you create your own system, but
that has problems.
1 Department for International Development (DFID) Back
2
United States Agency for International Development (USAID) Back
3
Ev 10 Back
| 
