October 11, 2006
Mr. Jim B. Rosenberg
Senior Assistant Chief Accountant
Division of Corporate Finance
United States Securities and Exchange Commission
100 F Street, N.E.
Washington, DC 20549-7010
Re: | Hospira, Inc. |
| Form 10-K for Fiscal Year Ended December 31, 2005 |
| Filed March 15, 2006 |
| File No. 1-31946 |
Dear Mr. Rosenberg:
We have received the Staff’s comment letter dated September 14, 2006. Our responses to the comment letter are set forth below. The numbered paragraphs below correspond to the numbered comments in the comment letter.
Comment 1. With regards to your revenue recognition policy documented here as a critical accounting policy and within your notes to the consolidated financial statements, we believe that your disclosure related to estimates of items that reduce gross revenue such as chargebacks, customer rebates and other discounts and allowances could be improved. Please provide us in disclosure-type format the following:
a) The effect that could result from using other reasonably likely assumptions than what you used to arrive at each accrual such as a range of reasonably likely amounts or other type of sensitivity analysis. In this regards, we believe that a reasonably likely analysis should be provided rather than the hypothetical one percentage point discount analysis that you provided.
b) If applicable, discuss to what extent information used to arrive at each accrual is from external sources (e.g., end-customer prescription demand, third-party market research data comparing wholesaler inventory levels to end-customer demand).
c) A roll forward of the accrual for each estimate for each period presented showing the following:
Hospira Inc.
275 North Field Drive
Dept. 9640, Bldg. H1
Lake Forest, IL 60045
www.hospira.com
1
· Beginning balance,
· Current provision related to sales made in current period,
· Current provision related to sales made in prior periods,
· Actual returns or credits in current period related to sales made in current period,
· Actual returns or credits in current period related to sales made in prior periods, and
· Ending balance.
d) A results of operations discussion of the amount of and reason for the period to period fluctuations for each type of reduction of gross revenue, such as product returns, chargebacks, customer rebates and other discounts and allowances including the effect that changes in your estimates of these items had on your revenues and operations.
HOSPIRA RESPONSE:
Revised Disclosure. In response to your comment, we propose the following revised disclosure regarding estimates for items that reduce gross sales:
Upon recognizing revenue from a sale, Hospira records an estimate for certain items that reduce gross sales in arriving at its reported net sales for each period. These items include chargebacks, rebates and other items (such as cash discounts and returns). Provisions for chargebacks and rebates represent the most significant and complex of these estimates.
Chargebacks— Hospira sells a significant portion of its generic injectable pharmaceutical products through wholesalers, which maintain inventories of Hospira products and later sell those products to end customers. In connection with its sales and marketing efforts, Hospira negotiates prices with end customers for certain products under pricing agreements (including, for example, group purchasing organization contracts). Consistent with industry practice, the negotiated end customer prices are typically lower than the prices charged to the wholesalers.
When an end customer purchases a Hospira product that is covered by a pricing agreement from a wholesaler, the end customer pays the wholesaler the price determined under the pricing agreement. The wholesaler is then entitled to charge Hospira back for the difference between the price the wholesaler paid Hospira and the contract price paid by the end customer (a “chargeback”). This process is necessary to enable Hospira to track actual sales to the end customer, which is essential information to run the business effectively. Settlement of chargebacks generally occurs between 30 and 40 days after the sale to wholesalers.
To account for the chargeback, Hospira records the initial sale to a wholesaler at the price invoiced to the wholesaler and at the same time, records a provision equal to the estimated amount the
2
wholesaler will later charge back to Hospira, reducing gross sales and trade receivables. This provision must be estimated because the actual end customer and applicable pricing terms may vary at the time of the sale to the wholesaler. Accordingly, the most significant estimates inherent in the initial chargeback provision relate to the volume of sales to the wholesalers that will be subject to chargeback and the ultimate end customer contract price. These estimates are based primarily on an analysis of Hospira’s product sales and most recent historical average chargeback credits by product, estimated wholesaler inventory levels, current contract pricing, anticipated future contract pricing changes and claims processing lag time. Hospira estimates the levels of inventory at the wholesalers through analysis of wholesaler purchases and inventory data obtained directly from certain of the wholesalers.
Hospira regularly monitors the provision for chargebacks and makes adjustments when it believes the actual chargebacks may differ from estimates. At December 31, 2005 and 2004, chargebacks of $64.2 million and $76.1 million, respectively, were recorded as a reduction in trade receivables. The methodology used to estimate and provide for chargebacks was consistent across all periods presented.
Rebates— Hospira primarily offers rebates to direct customers, customers who purchase from certain wholesalers at end customer contract prices and government agencies, which administer various programs such as Medicaid. Direct rebates are generally rebates paid to direct purchasing customers based on a contracted discount applied to the direct customer’s purchases. Indirect rebates are rebates paid to “indirect customers” which have purchased Hospira products from a wholesaler under a pricing agreement with Hospira. Governmental agency rebates are amounts owed based on legal requirements with public sector benefit providers (such as Medicaid), after the final dispensing of the product by a pharmacy to a benefit plan participant. Rebate amounts are usually based upon the volume of purchases. Hospira estimates the amount of the rebate due at the time of sale, and records the liability and a reduction of gross sales at the same time the product sale is recorded. Settlement of the rebate generally occurs from three to 12 months after sale.
In determining provisions for rebates to direct customers, Hospira considers the volume of eligible purchases by these customers and the rebate terms. In determining rebates on sales through wholesalers, Hospira considers the volume of eligible contract purchases, the rebate terms and the estimated level of inventory at the wholesalers that would be subject to a rebate, which is estimated as described above under “Chargebacks.” Upon receipt of a chargeback, due to the availability of product and customer specific information, Hospira can then establish a specific provision for fees or rebates based on the specific terms of each agreement. Rebates under governmental programs are based on the estimated volume of products sold subject to these programs.
3
Hospira regularly analyzes the historical rebate trends and makes adjustments to recorded reserves for changes in trends and terms of rebate programs. At December 31, 2005 and 2004, accrued rebates of $83.5 million and $74.1 million, respectively, are included in other accrued liabilities. The methodology used to estimate and provide for rebates was consistent across all periods presented.
The following table is an analysis of chargebacks and rebates for 2005 and 2004. In each year, the provisions for chargebacks and rebates relating to prior period sales were not material.
(dollars in thousands) | | Wholesaler Chargebacks | | Rebates | |
Balance at January 1, 2004 | | $ | 72,224 | | $ | 72,677 | |
Provisions | | 726,870 | | 126,656 | |
Payments and credits | | (722,998 | ) | (125,218 | ) |
| | | | | |
Balance at December 31, 2004 | | $ | 76,096 | | $ | 74,115 | |
| | | | | |
Provisions | | 628,338 | | 130,951 | |
Payments and credits | | (640,250 | ) | (121,529 | ) |
| | | | | |
Balance at December 31, 2005 | | $ | 64,184 | | $ | 83,537 | |
Explanation.
We have reviewed our critical accounting policies and do not consider the allowances for cash discounts and returns to be critical accounting policies or estimates. In the Commission’s Interpretative Release No. 33-8350 (Commission Guidance Regarding Management’s Discussion and Analysis of Financial Condition and Results of Operations), critical accounting estimates are described as accounting estimates or assumptions where:
· the nature of the estimates or assumptions is material due to the levels of subjectivity and judgment necessary to account for highly uncertain matters or the susceptibility of such matters to change; and
· the impact of the estimates and assumptions on financial condition or operating performance is material.
Aggregate provisions for cash discounts are less than 2% of Hospira’s net sales. Discount terms are contractual in nature, and, based on our experience, estimation of discounts does not
4
require us to apply judgment over highly uncertain matters. Aggregate provisions for returns are less than ½ of 1% of Hospira’s net sales and do not have a significant impact on net sales or results of operations.
Part (a)
In considering your request that Hospira should disclose the effect that could result from using other reasonably likely assumptions than what was used to arrive at each accrual, such as a range of reasonably likely amounts or other type of sensitivity analysis, we note that numerous factors are involved in the determination of these provisions, which are based on multiple assumptions and estimates. Alternative assumptions for these factors could have competing effects on these provisions. The provisions are driven by end customer pricing, wholesaler inventory analysis and rebate programs applicable to a significant number of individual products, none of which are individually material to Hospira’s net sales but which have differing pricing, rebate and inventory stocking trends. Attempting to provide disclosure at a factor-by-factor and/or product-by-product level would require multiple variables to be increased or decreased and would result in a level of disclosure and a range of potential outcomes that would de-emphasize material information. In its disclosure above, Hospira has clearly outlined the key factors that drive its estimates, which Hospira believes best serves the objective of providing investors useful and material information to assess Hospira’s net sales and results of operations. We confirm that Hospira will disclose when changes in its methodology or assumptions used in developing its estimates and changes in specific components of its provisions have a material impact on its net sales and results of operations.
Part (b)
Hospira obtains some inventory data directly from its wholesalers, and has disclosed this fact above. In addition, as disclosed, rebates to direct customers are based on the sales to and specific contract terms with that customer and rebates to indirect customers are based on chargeback data received which provides specific customer and product information on which to determine the rebate provision.
Part (c)
On the roll forward of the chargeback and rebate accruals, we have not broken out the provisions related to prior year sales as the amounts are not material. We have, however, provided disclosure to that effect. A change in the provision for chargebacks relating to prior period sales could occur if either (a) end user contract prices change after a sale of a product to a wholesaler in the prior period but before sale of such product to the end user customer in the
5
current period or (b) the actual units sold subject to chargebacks are different than had been estimated. Based on Hospira’s experience, changes in the provision for chargebacks as a result of prior period sales have not been material, because (1) of the relatively short time frame between sale of the product to a wholesaler and the processing of the chargeback (typically 30 to 40 days), (2) of the insignificance of end customer contract pricing changes during that time and (3) historical trends in the volume of units subject to chargebacks have been relatively consistent. In future filings, Hospira will disclose if a change in the provision for chargebacks that results from prior period sales materially impacts its net sales in the current period.
A change in the provision for rebates relating to prior period sales could occur if the actual discount rate earned by the customer differs from the rate assumed in estimating the rebate. The discount rate earned by the customer is based upon the volume of their purchases during the contract rebate period. Hospira determines the estimated rebate based on historical experience with the customer, also taking into consideration other relevant current data, as available. Hospira’s experience has been that changes in the provision for rebates relating to prior period sales have not been material as the historical trends of customer purchases and associated discounts have been relatively predictable. In future filings, Hospira will disclose if a change in the provision for rebates that results from prior period sales materially impacts its net sales in the current period.
Part (d)
In considering your request that Hospira provide a results of operation discussion on the amount and reasons for changes to the reductions in gross sales, we note that, to help investors understand the period-to-period comparisons in Hospira’s net sales, we disclose the effect that material volume and pricing changes have on net sales. These changes often directly impact the factors considered in determining the chargeback and rebate estimates. For example, a launch of a new product that has 180 days of market exclusivity will often drive higher volume and end user contract prices in the period launched, which would be disclosed in the discussion of Hospira’s net sales. It would not necessarily provide useful or meaningful disclosure to repeat that the product launch also drove higher provisions for chargebacks and/or volume-based rebates. However, if a change in a provision has a material impact on net sales, for example because of a change in assumption or methodology resulting from a change in industry trend or practice, the impact will be disclosed in the context of the discussion in changes in net sales. Additionally, we note that while, in the case of chargebacks, a difference between wholesaler prices and end user contract prices is a primary driver of the amount of the provision for chargebacks and can significantly change the amount of provisions relative to gross sales from year to year, these differences are not necessarily indicative of the prices Hospira charges end customers for its products and the overall level of Hospira’s net sales. For example, a
6
decrease in wholesaler prices would result in a lower chargeback provision but would have no impact on end customer contract prices and, accordingly, no impact on net sales. Disclosure of the changes in wholesaler prices and corresponding chargeback provision changes would not provide useful or meaningful information to investors as there is no resulting change in net sales.
Comment 2. Please provide us your analysis under FAS 13 that demonstrates how the length of your warranty obligation effects the classification of your leases as either operating or sales-type for drug delivery pumps.
HOSPIRA RESPONSE
In assessing the appropriate revenue recognition for its lease arrangements (i.e. operating vs. sales-type), Hospira performs an analysis of its lease arrangements in accordance with FAS 13. Under FAS 13, if a lease meets at least one of the criteria in paragraph 7 and both of the criteria in paragraph 8, the lease is classified as sales-type. Otherwise, the lease is classified as operating.
Hospira’s leases generally meet at least one of the criteria in paragraph 7. For example, Hospira’s lease term generally averages five years. The estimated economic life of many of our pumps is five years. As such, for these leases, the criterion in paragraph 7(c) is met since the lease term is equal to 75% or more of the estimated economic life of the pump.
Paragraph 8(a) requires that collectibility of the minimum lease payments is reasonably predictable. Hospira’s historical experience with our customers provides a reasonable basis for assessing collectibility and determining an appropriate reserve for potential uncollectibility. Based on our experience, the criterion in paragraph 8(a) is met for all leases. Therefore, the classification depends on whether the criterion in paragraph 8(b) is met, which is that there are no important uncertainties surrounding the amount of unreimbursed costs yet to be incurred by the lessor under the lease.
Paragraph 8(b) indicates that important uncertainties might include commitments by the lessor to guarantee performance of the leased property in a manner more extensive than the typical product warranty. Some of Hospira’s lease arrangements contain a contractual provision whereby Hospira is responsible for the repair and maintenance of the pump over the entire lease term. Hospira considers this provision to be an important uncertainty as to the amount of unreimbursed costs yet to be incurred because the cost to repair a pump over the extended term is not considered incidental — there is a substantial amount of labor, materials and re-certification required when a pump is returned for service. Hospira considers these costs to be
7
an “important uncertainty yet to be incurred”. As a result, these arrangements do not meet the requirements of paragraph 8(b) and are therefore accounted for as operating leases.
In contrast, other of Hospira’s arrangements include a one year warranty, which we believe is generally standard in the industry. Beyond the one year term, the customer either pays Hospira an incremental amount for extended warranty or takes on the responsibility for maintaining the pump. Historically, costs incurred under the one year warranty have not been significant. Hence, for these arrangements the criterion in paragraph 8(b) is met and these leases are accounted for as sales-type leases.
Comment 3. Please provide us, in disclosure-type format, your accounting policy for recognizing revenue for “placements under contracts that include associated disposable set purchases” for drug delivery pumps.
HOSPIRA RESPONSE
In response to your comment, we propose the following revised disclosure:
Hospira has contractual arrangements with certain customers whereby it places drug delivery pumps at customer sites, and the customers agree to purchase minimum levels of disposable products (sets) that are used with the pumps. These arrangements do not include any upfront fees or payments. The contractual arrangements generally set forth fixed prices for the purchases of the disposable products, where the prices for the disposables do not change over the term of the arrangement, other than, in some cases, for changes in Consumer Price Index provisions. Title for the pumps is retained by Hospira throughout these arrangements, and the related asset is depreciated over its estimated useful life on a straight-line basis. In these placement arrangements, revenue is recognized as the disposable products are delivered, in accordance with SFAS No. 48, Revenue Recognition when Right of Return Exists and SAB No. 104, Revenue Recognition.
Comment 4. Please provide us, in disclosure-type format, your accounting policy for your warranty obligation in connection with sales and lease of your products.
HOSPIRA RESPONSE
Hospira does not consider its accounting policy for warranty obligations on drug delivery pumps to be a significant accounting policy as the impact of the estimates and assumptions made by Hospira in establishing its warranty reserve is not material to Hospira’s financial condition or operating performance. In addition, Hospira has considered FIN45 related to its
8
reserve for warranty costs. Because Hospira’s reserve for warranty costs is not material, detailed disclosure in the notes to the consolidated financial statements is not required.
Hospira provides for warranty only on outright sales and sales-type leases. In other contractual arrangements such as operating leases and contracts that include associated disposable set purchases, Hospira retains ownership of the assets throughout the term of the agreement and, accordingly, depreciates those assets. Hospira retains all obligations of ownership, including responsibility for the repair and maintenance of the pumps over the term of the agreement. Historically, the majority of Hospira’s pumps have been placed under these types of arrangements. As such, Hospira’s warranty reserve has not been material. As Hospira transitions to more sales-type lease arrangements and outright sales, we will continue to monitor the reserve and provide the disclosure requested if it becomes material.
* * *
9