[afro-nets] RFI: Malaria Management (10)

[Peter Burgess <Profitinafrica@aol.com>]

RFI: Malaria Management (10)
----------------------------

Clarification of the idea of cost?

Dear colleagues,

Dr Bruno Moonen has asked a very simple question... "Can I clar-
ify my position on cost for malaria treatment?"

But the answer is far from simple.

I think a lot about "value chain". I hope this will help.

If I buy a diamond ring for my wife from Tiffany's in Manhattan,
the cost is likely to be $10,000 up. That is the amount of money
I have to pay to get the diamond ring. That is my cost.

For Tiffany's the cost is what they pay the wholesaler to get
the ring, plus their costs (allocated) for running a 5th Avenue
store in Manhattan. There cost is perhaps $3,000 for the ring
itself and say $2,000 for the cost of running the store with all
the costs associated with that. Their "price" for the ring is
$10,000. Their price becomes my cost! And the difference between
the price and the cost is their profit.

The same process is going on all the way along the value chain.
The wholesaler has a price, and it has costs for the item and
costs for running the warehouse and service.

Eventually one gets back to the diamond in the river in Sierra
Leone, and a villager selling the stone to a local trader/en-
trepreneur for $3.00. The value chain has costs and profits all
the way to Tiffany's.

I also think a lot about behaviour of cost. And this can get
quite complex. It is very simple accounting on top of whatever
complexity is reality.

Once upon a time computer systems had custom developed soft-
ware... which was very expensive. But in the 1980s it started to
be possible to buy software "out-of-the-box". Lotus 1-2-3 cost
about $4 million to develop... but because of mass marketing I
could buy it for around $400... so that was real progress to-
wards productivity. For Lotus the cost of making the first one
unit was $4 million for development plus say $50 for reproduc-
tion, packaging, etc. When they had sold a million units the
cost was $4 million for development plus $50 million for repro-
duction and packaging. Their profit after selling 1 million
units was $350 million... their loss after selling one unit was
$4 million... minus a little bit.

Quantity has a huge impact on cost. This is sometimes associated
with the idea of economy of scale... but it is a far simpler
concept. Unit cost is total cost divided by quantity... and it
changes a lot of analytical arithmetic when applied correctly.

While we are thinking about economies of scale, note that there
are also major dis-economies of scale... which are quite often
bigger than the economies of scale, and often ignored.

Unit cost is easy to calculate... but having a unit cost have
any meaning or be useful is another matter.

If we use the example of aerial spraying of insecticides there
are a series of unit costs that all need to be understood: Cost
per hour... very high, especially the cost of chemicals Cost per
acre... very low, because in an hour aerial spraying covers a
very large amount of ground Cost per capita... very variable...
depends on the population density per acre and in the case of
highly populated locations is very low. Cost per capita per
year... even more variable, because it also depends on how many
times the same space needs to be sprayed. Cost per mosquito
killed... varies again depending on the efficacy of the chemi-
cals, the skill of the operators, the entomological information
used for planning operations.

The analysis gets even more interesting, because frankly, we are
interested not in killing mosquitoes, but getting rid of the
scourge of malaria.

The challenge is to relate all of the costs just discussed with
the impact on the malaria in the community.

This is not a simple relationship... but the fact that it is not
simple does not mean it should not be thought about and cost and
management information compiled to measure the relationship.

A good management information system can use measures of mos-
quito population and relate the different practical interven-
tions so that a useful cost relationship can be developed be-
tween the mosquito population and the cost of interventions.

At the same time data about the malaria situation in the commu-
nity can be collected and also recorded in a time series.

It is my understanding that the medical community has compiled a
lot of information about what different medications do when ap-
plied to people with malaria... so with medication one would ex-
pect that malaria in the population would go down. But it seems
to come back again... WHY? And then the medications seem to lose
their potency... WHY? What is the problem?

The problem is almost certainly to do with re-infection... new
mosquito bites... and the malaria is back again.

The cost of a medical (medication) intervention where there is a
100% certainty of re-infection is indeterminate... but think of
it as rising to infinity.

We do it because we think that doing nothing in the face of ma-
laria is wrong... but when you do the costing rigorously with an
attempt to understand the process... we are actually doing worse
than nothing... we are achieving nothing at huge cost.

It does not seem to me to matter very much how effective the
drugs are when re-infection is the dominant aspect of the dis-
ease... and from what I understand of the malaria situation this
is the challenge.

The question I am struggling to answer is what is the least cost
way to end malaria re-infection... and my work to this point
suggests that a wide use of aerial spraying together with spe-
cific local spraying interventions (adulticides and larvacides)
together with deep data collection about mosquito populations
and data collection about malaria cases... and medicating the
cases as much as possible... and limiting mosquito access to ma-
laria infected people will give the best results.

This is an accountant talking... not an entomologist, or a medi-
cal scientist, or a pilot. Accountants are hardly every right...
but they do help get the right questions answered.

And accountants do accounting, not statistics. An accountant's
numbers tend to be right... they may be useless... but they
should be right. Statistics often give more useful numbers, but
they have low probability of either being right or accurate. A
healthy mix of accounting numbers and statistical numbers might
get us closer to knowing what we need to know.

This all started with a simple question about cost.

There is a lot of catch up to do to get a good cost and informa-
tion regime around the global relief and development assistance
sector.

Sincerely,

Peter Burgess
Tr-Ac-Net in New York
Tel.: +1-212-772-6918
mailto:peterbnyc@gmail.com
The Transparency and Accountability Network
http://tr-ac-net.blogspot.com
http://www.tr-ac-net.org