Improving the PPI sample for prescription pharmaceuticals - Producer Price Index

Monthly Labor Review, Oct, 1997 by Gregory G. Kelly

Frequent new product introductions make developing price indexes for this industry challenging; BLS updates its methodology to capture new products and make its sample more representative

The prescription pharmaceutical industry, with its frequent product innovation and tightly regulated markets, poses unique challenges to the development of price indexes. This article describes the industry and the features that make it unique. It also discusses some of the problems confronting analysts as they develop price indexes for the industry, and the solutions implemented by the Producer Price Index (PPI) program of the Bureau of Labor Statistics to more accurately measure price change for prescription pharmaceuticals.(1)

About the industry

The pharmaceutical industry is characterized by frequent product innovation, with manufacturers continually developing and marketing new products. Some of these newly-developed products compete with existing products; others are completely new. The development of prescription pharmaceuticals requires costly and time-consuming research. After a product has been developed, it must undergo the rigorous approval process of the Food and Drug Administration (FDA). This process, which currently takes about 18 months, tests the product for safety as well as its efficacy in treating specific conditions.

To allow time to recoup the considerable investment costs associated with developing new products, manufacturers of FDA-approved new products are granted a period of patent/exclusivity protection. During this period, the company gaining approval has exclusive rights to the product's formulation. The new product competes against other products used to treat the same condition, but with different formulations and characteristics. When the protection period expires, other companies can gain approval to market "generic" versions of the product.(2)

Two new laws enacted since the early 1980s accelerate the approval process for new pharmaceutical products, both patented and generic. The Hatch-Waxman Act, passed in 1984, reduces the time between the patent expiration of a predecessor product and the approval of bioequivalent generic competitors.(3) The Prescription Drug User Fee Act, passed in 1992, significantly decreases the approval time for newly developed drugs. Both new drugs and generic versions of existing drugs now reach the market sooner and competition has increased within the therapeutic classes of drugs. Increased competition may help control inflation in the industry. Throughout the 1980s, prices for prescription pharmaceuticals increased much more rapidly than overall prices. Since 1992, however, inflation in the prescription pharmaceutical industry has slowed to less than onehalf its 1980-92 average. (See chart 1.)

[Chart 1 ILLUSTRATION OMITTED]

Price behavior and product age

Ongoing BLS research shows that prices of both new protected and generic drug products increase less rapidly (or even decline) in the 2 years following the products' introduction, than those of drugs on the market for more than 2 years. In one internal analysis, BLS researchers found that prices for drugs on the market for 2 or fewer years had decreased by 15.9 percent, while prices for the sample as a whole had increased by 3 percent. The same analysis also found that new protected products had a lower rate of inflation than the sample as a whole, while the rate for generic drugs declined.(4) (See table 1.)

Table 1. Percent change in the producer price index for prescription pharmaceuticals by product age class, 1993-96

 Product                         December 1993   December 1995
age class                             to              to
(in years)                       December 1994   December 1996

[is less than or equal to] 2         -15.9            0.7
[is greater than] 2 to
 [is less than or equal to] 4          1.3           -3.1
[is greater than] 4 to
 [is less than or equal to] 6          4.1            4.7
[is greater than] 6 to
 [is less than or equal to] 9          2.6            1.5
[is greater than] 9 to
 [is less than or equal to] 24         6.7            3.4
[is greater than] 24                   3.8            4.4
Total                                  3.0            2.9

The sampling universe was stratified into 34 separate therapeutic classes. Products were selected from each stratum in two stages, with probability of selection proportionate to dollar sales. In the first stage of sample selection, the product was selected. In the second stage of sampling, the specific dosage form and strength were selected. No restrictions were placed on the number of quotes per company. Specific product presentation and transaction variables were selected at the establishment in a third stage, with probability of selection again proportional to dollar value. The final sample consisted of 965 items from 124 manufacturers. Because the number of quotes per manufacturer was not restricted, many companies were asked to supply the PPI with more than 20 quotes, and some were asked for more than 30. BLS was able to obtain cooperation in repricing 511 items.