Pricing and patents of HIV/AIDS drugs in developing countries

Abstract

This article provides empirical evidence on the impact of patents on drug prices across developing countries. It uses sales data on human immunodeficiency virus (HIV)/acquired immuno deficiency syndrome (AIDS) drugs in a sample of 34 low- and middle-income countries between 1995 and mid-2000. The main findings are that patents do shift drug prices up, drug prices are correlated to per capita income levels and drug firms follow a skimming strategy when pricing new HIV/AIDS drugs. That is, there is across country and intertemporal price discrimination in the global drug markets.

Acknowledgements

I thank Jayashree Watal's leading role and support and Dani Rodrik's encouragement in the implementation of the CID project on studying the impact of patents on the access, pricing and availability of HIV/AIDS drugs in developing countries. I gratefully acknowledge the comments from Albert Saiz and many participants at Harvard, NBER, Carlos III and Vigo Seminars. I was a post-doctoral fellow at Harvard Economics Department while working on this article. I thank my host and Ariel Pakes kindly supervision. I gratefully acknowledge funding from the CID and unrestricted educational grants from Fundacion Ramon Areces (Madrid, Spain) and The Merck Foundation, the philanthropic arm of Merck & Co. Inc., Whitehouse Station (New Jersey, USA). The opinions stated are only those of the author and any error or omission remains the author's sole responsibility.

Notes

¹ Bagwell and Riordan (1991) analyse the case in which the firms signal high-quality new products with prices that are above full information profit maximizing prices. As information about the prices diffuses, the price distortion disappears.

² The empirical literature that studies specific drug markets shows that we should control dosage, efficacy, toxicity and side effects among other observed qualities: Berndt et al. (1993) study antihypertensive drugs; Berndt et al . (1995) and Berndt et al. (1999 and 2000) focus on anti-ulcer drugs; Berndt et al. (1996) analyse antidepressant drugs; and Cockburn and Anis (1998) arthritis drugs. We do not have enough data on differences in drug toxicity among ARV's although higher life-threatening toxicity has been related to the use of a type of ARV, the so-called nucleoside reverse transcriptase inhibitors (NRTI). Therefore, at least drug fixed effects take care of fixed differences in toxicity across drug types.

³ IMS provided us with aggregated sales data for two supranational entities: French West Africa, comprising aggregate sales in Benin, Cameroon, Democratic Republic of Congo, Ivory Coast, Gabon, Guinea and Senegal; and, Central America, including Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua and Panama. All economic indicators for those two supranational entities are population weighted averages of the national indicators. The data set includes annual data referring to the calendar years from 1995 to 1999 and also to the year from July 1999 to June 2000.

⁴ For each drug presentation, IMS reports data by year, country, molecule, firm, brand name, pharmaceutical form, strength and pack size. Sales revenues obtained from each package or presentation at the wholesale level are reported in current US$. Physical sales are reported in standard units (number of tablets, vials or teaspoons). IMS did not provide us with sales of the active ingredient in milligrams, nor in daily doses.

⁵ These were Argentina, Bangladesh, Brazil, Colombia, Egypt, Guatemala, Guinea, India, Morocco, Pakistan, Peru, Tunisia, Uruguay and Venezuela. Pakistan had a patent law in force, but an executive order disallowed pharmaceutical patents.

⁶ Colombia, Ecuador and Venezuela introduced product patents in 1996; Brazil in 1997; Argentina in 1999; Guinea, Guatemala and Morocco in 2000. We focus on product patents (exclusivity related to therapeutically active ingredient) rather than process patents (exclusivity related to the method of obtaining such an active ingredient). Process patents, like other type of patents on therapeutic uses, pharmaceutical forms and so on, are important but accessory ways of protecting the main and broader exclusivity right of the innovator, that of protecting the therapeutic active ingredient from being copied and sold.

⁷ When product patents are not available as of 1 January 1995, WTO members have to provide a system whereby drug patent applications can be filed (often referred to as a ‘mailbox’ system). ‘Mailbox’ applications do not have to be examined until the local patent law is passed. However, when a drug subject to a ‘mailbox application’ obtains marketing approval before the local patent office makes a decision on whether granting a patent right or not, the following special rule applies: An exclusive marketing right (EMR) of up to 5 years (or until the patent is granted or rejected, whichever is shorter) must be granted from the date of local marketing approval, provided that a patent has been filed for that drug and a patent and marketing approval obtained in another WTO member country after 1 January 1995.

⁸ An invention is considered to be new if it does not form part of the state of the art. The ‘state of the art’ is generally defined, as everything made available to the public by means of a written or oral description, by use, or in any other way, before the date of filing of the patent application. Under WTO rules, incorporating existing WIPO (World Intellectual Property Organization) conventions, for purposes of determining novelty, patent applicants may claim the priority of an earlier application made during the period of 12 months from the date of filing. It is theoretically possible to have a patent applicant not to claim priority from the date of an earlier filing and to claim that products, for which patent applications were filed elsewhere from say, mid-1993, are novel (since later than this date the application would then be published by another patent office after 18 months and so would no longer be novel as of or after January 1995), but we believe that this is unlikely to happen in practice.

⁹ The US Federal Food, Drug and Cosmetics Act required that drug firms provide patent information with all new drug applications. Taking into account this information, the FDA sets the exclusivity term during which an abbreviated new drug application is not granted (a generic is not approved). The Electronic Orange Book (FDA, 2000) publishes the number of the appropriate patents claimed by the firms when the drugs are subject to approval. Using the patent numbers, Balasubramaniam (2000) obtained each ARV key priority date from the US Patent and Trademark Office online database (http://www.uspto.gov). We thank Mike Palmedo from the Consumer Project on Technology for explaining to us how this data was gathered.

¹⁰ Product patents for drugs have been granted in all Central American countries since the 1950s, except in Guatemala where product patents were introduced in 2000. We set the patent regime indicator to be equal to 1, in all the country--drug pairs corresponding to Central America.

¹¹ As reported in the Red Book (1995, 1996, 1997, 1998, 1999, 2000), the PDR-Generics (1997) and PDR-CG (1998, 1999, 2000). We matched each local price with the US minimum wholesale list price for each particular drug and pharmaceutical form pair between 1995 and 2000.