Introduction
Activists have long claimed that patents make antiretroviral drugs unaffordable and inaccessible to people with AIDS in Africa; the pharmaceutical industry has stridently denied these claims. The debate has been one of the most intense and most public discussions of the TRIPs Agreement. This article examines the patent status of antiretroviral drugs in Africa and attempts to determine if it is, in fact, patent protection that limits access to treatment.

Setting and Assumptions
The authors consider the patent status of fifteen antiretroviral drugs patented by eight pharmaceutical firms in fifty-three African countries. To gather the necessary data, the authors sent written inquiries to the intellectual property divisions of the pharmaceutical companies in question. The companies’ responses were summarized and then returned for corrections and verification. This collection of data is one of the most interesting and useful elements of the study.

In setting up their model, the authors make the “generous” assumption that all countries offer pharmaceutical patents, an assumption that is currently inaccurate. This strong assumption strengthens the authors’ claim that patent protection has not impeded access to antiretroviral drugs. They calculate that for fifteen drugs and fifty-three countries there are potentially 15 x 53 = 795 patents. Of this number, 172, or 21.6 percent, actually exist.

Results
Because only 21.6 percent of the possible patents for antiretroviral drugs exist in these countries, Attaran and Gillespie-White conclude that patents do not limit access to antiretroviral drugs in Africa. They argue that other factors must be to blame: ubiquitous poverty, lack of political will, poor medical care and infrastructure, inefficient regulatory procedures, and high tariffs and taxes. The authors then suggest a number of possible solutions. These include a global purchasing facility to realize volume discounts, voluntary licensing by producers, which would allow third world countries to produce the drugs themselves, and an ensured supply financed by the international community. Lack of international financing, the authors conclude, bears the bulk of the blame.

It may be that these conclusions are exactly right, but the authors’ methodology raises serious questions about their validity. Attaran and Gillespie-White fail to account for a number of significant factors. They give equal weight to all patents in all countries when calculating the fraction of patent coverage. A more informative figure would adjust for infection rates, drug utilization, and income levels.

Specifically:
1. What portion of infected individuals actually lives in countries with extensive patent protection for antiretrovirals? If the majority of HIV/AIDS sufferers are found in countries where most of the drugs are on-patent, then patents may, in fact, be a serious obstacle to the treatment of these people.
2. What percentage of the drugs in use are on-patent? If 90 percent of all patients use one particular patented drug, the fact that none of the fourteen other antiretrovirals are not covered by patents would scarcely matter.
3. Is patenting more prevalent in countries with higher gross national income (GNI) per capita? That is, in the poorest countries, many patients may not be able to afford antiretroviral drugs even if they were not patented. Patents may only pose an obstacle to treatment in countries where many patients have sufficient income to afford unpatented drugs.

Some indication of the role these other factors play can be found in the table below. In this subset of countries where there is a total of at least six patents for antiretroviral drugs, the calculations for GNI per capita and the prevalence of infection—the percentage of the population infected with HIV—are the most interesting.

Country	Patent Rank1	Per Capita Income Rank2	Prevalence Rate Rank3	No. of Patents4	GNI per capita5	Prevalence Rate6
Botswana	6	3	1	6	$3,630	38.8%
Zimbabwe	2	21	2	8	$480	33.7%
Swaziland	6	11	3	6	$1,300	33.4%
Lesotho	6	18	4	6	$550	31%
Namibia	40	7	5	0	$1,960	22.5%
Zambia	6	29	6	6	$320	21.5%
South Africa	1	5	7	13	$2,900	20.1%
Gambia	3	27		7	$330
Sudan	3	28		7	$330
Uganda	3	33		7	$280
Ghana	6	30		6	$290
Malawi	6	44		6	$170

¹  Based on Table 1, “Patent Coverage in Africa for Antiretroviral Drugs, by Country” of the Attaran/Gillespie-White paper. 

²  Based on a rank ordering of the GNI per capita in 2001 (for the 53 countries in the Attaran/Gillespie-White paper), as reported by the World Development Indicators database of the World Bank, August 2002.  

³  Based on a rank ordering of the prevalence rates of African nations, as reported by UNAIDS.  Note that these figures represent the highest prevalence rates in Africa.  All other nations have rates lower than 20%.

⁴  Based on Table 1 of the Attaran/Gillespie-White paper. 

⁵  As reported by the World Development Indicators database of the World Bank, August 2002. 

⁶  As reported by UNAIDS.  These figures represent the highest prevalence rates in Africa.  All other nations have rates lower than 20%. 

---

On average, in the fifty-three African nations studied, there is an average of 3.2 patents per nation for antiretroviral drugs. In the six nations with the highest infection rates, the average more than doubles, rising to 7.5 patents. So, indeed, the prevalence of patenting is greater in the nations most affected by the disease (point number 1).

Though it is less straightforward, World Bank data on income levels also indicates that patents are more prevalent in countries that on average have higher income levels. The mean GNI per capita of the fifty-three nations considered is $330. For the subset of countries considered here, those with the greatest number of patents, nine of the twelve countries have a GNI per capita at or above this level. So patents pose the greatest obstacle to treatment in just those countries where many patients may be able to afford unpatented drugs (point number 3).

Unfortunately, there is no comprehensive data available about drug utilization. Lemivudine-zidovudine and Lamivudine are each on-patent in more than thirty countries. If these drugs make up a large share of the medications prescribed, then 21.6 percent is a very low estimate. Utilization data would shed a great deal more light on the question the authors are trying to answer.

All of these additional factors undermine the authors’ conclusion that “patents and patent law are not a major barrier to treatment access in and of themselves.”
-----------------
* Though there are no patents in Namibia, it was included because it appears on the UNAIDS list of nations with the highest prevalence of infection.