Eli Lilly 2006 Annual Report Download - page 25

FINANCIALS
23
nifi cant products sold to major wholesalers in the U.S.
and in major markets outside the U.S., primarily by
reviewing periodic inventory reports supplied by our
major wholesalers and available prescription volume
information for our products, or alternative approaches.
We attempt to maintain wholesaler inventory levels at
an average of approximately one month or less on a
consistent basis across our product portfolio. Causes
of unusual wholesaler buying patterns include actual
or anticipated product supply issues, weather patterns,
anticipated changes in the transportation network,
redundant holiday stocking, and changes in wholesaler
business operations. An unusual buying pattern com-
pared with underlying demand of our products outside
the U.S. could also be the result of speculative buying
by wholesalers in anticipation of price increases. When
we believe wholesaler purchasing patterns have caused
an unusual increase or decrease in the sales of a major
product compared with underlying demand, we dis-
close this in our product sales discussion if the amount
is believed to be material to the product sales trend;
however, we are not always able to accurately quantify
the amount of stocking or destocking.
As a result of restructuring our arrangements with
our U.S. wholesalers in early 2005, reductions occurred
in wholesaler inventory levels for certain products
(primarily Strattera, Prozac, and Gemzar) that reduced
our 2005 sales by approximately $170 million. The
modifi ed structure eliminates the incentive for specula-
tive wholesaler buying and provides us improved data
on inventory levels at our U.S. wholesalers. Wholesaler
stocking and destocking activity historically has not
caused any material changes in the rate of actual prod-
uct returns, which have been approximately 1 percent
of our net sales over the past three years and have not
fl uctuated signifi cantly as a percent of sales.
We establish sales rebate and discount accruals in
the same period as the related sales. The rebate/discount
amounts are recorded as a deduction to arrive at our
net sales. Sales rebates/discounts that require the use
of judgment in the establishment of the accrual include
Medicaid, managed care, Medicare, chargebacks, long-
term-care, hospital, discount card programs, and various
other government programs. We base these accruals
primarily upon our historical rebate/discount payments
made to our customer segment groups and the provisions
of current rebate/discount contracts. We calculate these
rebates/discounts based upon a percentage of our sales
for each of our products as defi ned by the statutory rates
and the contracts with our various customer groups.
The largest of our sales rebate/discount amounts
are rebates associated with sales covered by Medicaid.
Although we accrue a liability for Medicaid rebates at the
time we record the sale (when the product is shipped),
the Medicaid rebate related to that sale is typically billed
up to six months later. Due to the time lag, in any particu-
lar period our rebate adjustments may incorporate revi-
sions of accruals for several periods. In determining the
appropriate accrual amount, we consider our historical
Medicaid rebate payments by product as a percentage
of our historical sales as well as any signifi cant changes
in sales trends, an evaluation of the current Medicaid
rebate laws and interpretations, the percentage of our
products that are sold to Medicaid recipients, and our
product pricing and current rebate/discount contracts.
Most of our rebates outside the U.S. are contractual
or legislatively mandated and are estimated and recog-
nized in the same period as the related sales. In some
large European countries, government rebates are
based on the anticipated pharmaceutical budget defi cit
in the country. A best estimate of these rebates, updated
as governmental authorities revise budgeted defi cits, is
recognized in the same period as the related sale. If our
estimates are not refl ective of the actual pharmaceuti-
cal budget defi cit, we adjust our rebate reserves.
We believe that our accruals for sales rebates and
discounts are reasonable and appropriate based on
current facts and circumstances. Federally mandated
Medicaid rebate and state pharmaceutical assistance
programs (Medicaid) and Medicare rebates reduced
sales by $571.7 million, $637.1 million, and $641.0 mil-
lion in 2006, 2005, and 2004, respectively. A 5 percent
change in the Medicaid and Medicare rebate amounts
we recognized in 2006 would lead to an approximate
$29 million effect on our income before income taxes.
As of December 31, 2006, our Medicaid and Medicare
rebate liability was $259.0 million.
Approximately 85 percent and 90 percent of our
global rebate and discount liability resulted from sales
of our products in the U.S. as of December 31, 2006 and
2005, respectively. The following represents a roll-for-
ward of our most signifi cant U.S. rebate and discount
liability balances, including Medicaid (in millions):
2006 2005
Rebate and discount liability,
beginning of year. . . . . . . . . . . $ 379.4 $ 367.9
Reduction of net sales
due to discounts and
rebates 1 . . . . . . . . . . . . . . . 1,246.1 1,300.1
Cash payments of
discounts and rebates . . . . (1,242.2) (1,288.6)
Rebate and discount
liability, end of year. . . . . . . . . $ 383.3 $ 379.4
1 Adjustments of the estimates for these rebates and discounts to actual
results were less than 0.3 percent of net sales for each of the years
presented.
Product Litigation Liabilities and Other Contingencies
Product litigation liabilities and other contingencies are,
by their nature, uncertain and are based upon complex
judgments and probabilities. The factors we consider
in developing our product litigation liability reserves