Impacts of intellectual property provisions in trade treaties on access to medicine in low and middle income countries: a systematic review

Intellectual Property (IP) provisions in free trade agreements (FTAs) ensure protection for the creation or invention of artistic works and goods, the creation or invention of which sometimes requires, as in the case of medicines, high sunk cost in the form of investment in research and development (R&D). Developing a new medicine requires large investment with high uncertainty. These R&D costs occur after a product patent is granted, which is typically very early in clinical development. IP provisions restrict the use and marketing of such goods and provide exclusive rights to the investors/creators to offset their sunk cost during clinical development [1]. This is to encourage more research and development (R&D) investment by the private sector to develop and invent new products [2]. Consequently, new or improved medicines are protected by patent and other IP provisions.

However, this protection creates a monopoly market for these medicines. Since the demand for medicines is generally price and income inelastic, this allows the owner of the patented medicine to charge a very high price [3]. As a result, there is growing concern among health care and development practitioners that IP provisions in trade agreements may have serious consequences on at least the affordability and/or availability of medicines in low and middle-income countries [4‐7]. Affordability and availability of medicines are key dimensions of “access”.

The Agreement on Trade-Related Aspects of Intellectual Property Rights (or the TRIPS Agreement) set the standards for intellectual property protection in the world. It came into force on 1 January 1995 and is binding on all members of the World Trade Organization (WTO) [8]. The TRIPS Agreement sets minimum standards in the international rules governing patents, including patents on medicines [8]. Countries that are members of the WTO agree to these minimum standards in the way they enact and implement their patent laws. In recent years, many countries have been coming under pressure to enact or implement additional conditions in their patent laws than may negatively impact access to medicines – these are commonly known as ‘TRIPS-plus’ provisions [9].

There are TRIPS IP requirements and TRIPS-plus provisions that may negatively impact access to medicines, a list of which collectively might include: (1) relaxed standards of patentability, including patents on new uses, modifications of active pharmaceutical ingredients, new formulations/dosages [10] (2) patent term extensions to compensate for delays in patenting and registration decisions [11]; (3) limiting or eliminating patent oppositions [12]; (4) data/marketing exclusivity [12]; (5) patent/registration linkage [10]; (6) TRIPS-plus restrictions on compulsory and government use licenses [10]; (7) enhanced IP enforcement and remedies [10‐12].

Providing a protected monopoly market to pharmaceutical products in countries could adversely affect access to originator medicines as well as to less expensive generic equivalents. Given this theoretical expectation of negative effects of stronger IP protection- e.g., TRIPS-plus, on access to medicines, a number of studies have been carried out to attempt to quantify the size of the effect. These studies are either ex-ante or ex-post in nature. Ex-ante studies use structural models and simulations to predict the likely impact of IP provisions on access to medicines, whereas ex-post studies utilize empirical data to measure the size of the effect. Some authors have suggested that the ex-ante studies invariably predict a robust negative effect of stronger IP regime on the affordability in the form higher prices or costs of medicines and availability in the form of lower consumption of medicines, whereas ex-post studies find mixed results from relatively mild negative to some positive effects [9].

Recently, Gleeson et al. [13] examined four trade and investment treaties to identify a channel of potential impacts of the specific treaty language on access to medicines and discussed studies that support their proposed analytical framework of pathways. They discussed the impacts on access to medicine mostly with respect to high-income countries and included mostly qualitative studies [13]. Our review is a complement to Gleeson et al. [13] as we focus more on quantitative empirical studies, and especially critically appraising the methodologies of these studies. Hence, the objectives of our study are to systematically review the literature for quantitative evidence that explore how the IP provisions in bilateral or multilateral FTAs affect the access to medicines in low and middle-income countries. Here, we have conducted a systematic literature review to analyze differences in methodologies of the studies, to summarize the range of impacts of IP protection on access to medicines and to assess the limitations of the studies. To this end, this systematic review attempts to answer the following questions:

We followed the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines [14].

The search strategy identified 1344 unique abstracts to review (Fig. 1 and Additional file 1). After removal of duplicates, 744 titles remained. After the first stage review of the abstracts,118 studies were selected, the rest being excluded as not being relevant. At the second stage, we identified 38 studies (see References) that fulfilled eligibility criteria 2 and 3. Finally, three authors (DI, WAK, VW) independently reviewed all 38 studies selected from the second stage and out of the 38 studies, 14 studies are selected unanimously. Out of the 14 studies, 7 studies are ex-ante and 7 are ex-post.

The selection process of articles at different stages is shown below in Fig. 1.

Additional file 3 summarizes the rationale for exclusion of studies. Additional file 4 provides the authors’ checklist for determining study limitations.

Studies selected in our systematic review used a variety of methods to disentangle the effects of IP provisions on access to medicines.

Our overall results show that there are effectively only two broad IP categories for which different quantitative studies have attempted to estimate their impact on access to medicines. These are: a) the TRIPS Agreement, with implementation into national IP laws [2, 16, 18, 23‐27], and b) TRIPS-plus provisions which include patent term extensions [19, 20] and data exclusivity or other commercial exclusivity provisions [17, 19, 21, 22, 26, 28]. Results of these studies shows that extending the patent term or ensuring data exclusivity has a larger negative effect on access to medicines compared with the IP benchmark set by the TRIPS Agreement [19, 20]. On the other hand, in comparing data exclusivity to patent term extension in Brazil, Chaves et al. [19] estimated larger expenditure on HIV and Hepatitis C medicine under data exclusivity than under patent term extension.

We identified several limitations of both ex-ante and ex-post studies with respect to methodologies and data used in those papers. The main limitations of each paper regarding data and methodologies are shown in the Tables 4 and 5. Here we discuss limitations of studies in details.

Our systematic literature review makes several contributions:

First, the studies we have reviewed show that changes in IP policy due to the implementation of trade agreements are associated with changes in price, medicines expenditure and sales, consumer welfare, and ultimately the affordability, of medicines. The direction and magnitude of the effects differ between ex-ante and ex-post studies. Regarding prices and costs of medicines, ex-ante studies predict that prices and costs (primarily public expenditure) of medicines could increase several hundred percent due to the impact of various IP provisions such as increased patent enforcement, TRIP-plus and other provisions in various multilateral and bilateral agreements. These ex-ante studies confirm what the theory would say [35] i.e., that stronger IP monopoly rights would tend to eliminate competition and thus incur societal costs which are higher prices for IP products.

On the other hand, empirical ex-post studies found at most a moderate increase in prices and costs of medicines due to the imposition of similarly heightened IP rules. There is, however, some consensus between ex-ante and ex-post studies that TRIPS-plus provisions relating to clinical data protection, rather than the imposition of more stringent patent rules, would cause a larger increase in prices and costs of medicines and lead to lower access to medicines. We note that extending the patent term may have an additionally important, but as-yet undifferentiated, impact since most data protection provisions are confined within the period of existing patent protection and are not additive to patent extensions. Second, the reported impacts of IP changes due to trade agreements on access to medicines seem clearly multifactorial. Duggan et al. [24] found an insignificant increase in medicine prices after patent law reform and argued that this might be because the existing generic producers are ‘grandfathered’ and continue to produce the generic medicines even after patent enforcement. This is because TRIPS does not require retroactive IP protection on pre-1994 medicines. Kyle and Qian [26] found that the existence of a patented molecule does not always block generic imitation, nor does the lack of patents always deter an originator from making a product available. They also pointed out that effects of IP may well be different depending on the size of the local generic sector, e.g., the impact in India with its large and robust generic medicine sector may be different as compared most other low and middle income countries. They asserted that the “… existence of IPs is neither necessary nor sufficient …” for the launch of pharmaceutical innovations at the country level. This suggests substantial heterogeneity in the effects of IPs, both across countries and across medicines.

Third, ex-ante studies that use structural models are often better able than ex-post studies to draw causal effects of IP policy changes on access to medicines. But ex-ante studies are based on stringent model assumptions and provide only counterfactual estimates. On the other hand, ex-post studies attempt to measure the actual effects of IP protection on access to medicines, but in most of the ex-post studies, the empirical models are not well identified and hence only a weak causal inference can be established. Shadlen et al. [30] emphasize the temporal impact of changes in IP provisions. The authors suggest that, depending on when countries first began allowing drugs to be patented, TRIPS-Plus provisions will have different effects.

To access the real effects of IP policy changes due to trade agreements on access to medicines, which approach, ex-ante or ex-post, would be more accurate? It is clear from our findings, that both methods have advantages and limitations and, on balance, what does seem clear is that both types of studies predict, for the most part, an increase in price and a decrease in consumer welfare with imposition of IP in trade agreements. The main differences between these studies are in the magnitude of the changes. The fact that such a difference in magnitude exists may well be due to the assumptions in ex-ante models and limitations of ex-post studies, but there are likely omitted and unmeasurable institutional variables in the health policy ecosystem that contribute as well.

Fourth, our literature review found that the impact of IP provisions in various trade agreements manifests itself through the healthcare/pharmaceutical ecosystem. Thus, this has implications for developing better empirical models to measure the effects on key outcome variables. For example, Jung and Kwon [25] assert that IP exerts an influence on medicine utilization only in countries above a certain income level. They did not observe any significant effect of IP on access to medicines in low-income countries where Gross Domestic Product (GDP) per capita is below 1000 US dollars. They also found that those who live in rural areas and have health insurance were more likely to report that they could not access their prescribed medicines compared to those who live in urban areas. Shaffer and Brenner [28] noted that CAFTA’s data exclusivity and patent rules were implemented in Guatemala through domestic law so one might ask whether differences in domestic implementation of IP provisions have an impact on the effect size on access to medicines. Kessomboon et al. [20] suggested that the strategies to address the negative consequences of an FTA that affect access to medicines would be based on several elements of the pharmaceutical system: medicine selection, procurement, distribution, and medicine use. Overall, we found a scarcity of studies analyzing the effects of changes in IP on different elements of the pharmaceutical system. Indeed, all of the ex-post studies on TRIP-plus provisions were done in Jordan.

Fifth, our literature review identified important research gaps that should be addressed: Is there a differential impact of IP provisions on different medicines for similar conditions, what is the impact of such provisions on essentially interchangeable medicines such as the insulins, what is the impact on local medicine production, on medicine quality, on affordability for various socio-economic groups, on medicine procurement, on medicine dispensing, on patient choice, and on prescriber choice in both public and private sectors? Indeed, our results suggest that we are unable, at this time, to unpack the main IP drivers that impact access to medicines. Further, the quantitative literature we have reviewed simply cannot say much about “cross country variation” in the effects of IP provisions on access to medicines. This is a clear research gap and should be subject to future research.

Clearly, trade treaties will manifest their impacts on a complex healthcare ecosystem. Sustainable Development Goal (SDG) 3.8 emphasizes the need for “access to safe, effective, quality and affordable medicines” [36]. Assessing the effect of IP provisions in trade treaties- by whatever methodology and setting aside methodological implications and limitations- should include a study of domestic implementation, access, availability and affordability; safety, efficacy and quality; rational use of medicines; procurement; and local production capacity.

Finally, several approaches and data sources would be relevant in this regard and should be driven by the research question, irrespective of methodology. The complex impact of IP and trade provisions on “access to safe, effective, quality and affordable medicines” calls forth relationships between medicines and health financing, human resources, health information and service delivery [31]. Studies looking at the impact of trade rules on populations’ access to medicines should no longer be addressed mainly through ‘siloed’ approaches that focus primarily on price. On one hand, mixed methods approaches can, in principle, offset the limitations of quantitative and qualitative studies by allowing for both exploration and analysis in the same study. Quantitative research is weak in understanding the context and qualitative research does not often lend itself to statistical analysis and generalization. For example, one possible next step in either ex-ante or ex-post studies is to use granular household level data, particularly in those many low and middle income countries where patients pay out-of-pocket for medicines.

For ex-ante studies, this will allow researchers to estimate the demand elasticities of medicines and hence, this will help predict the changes in prices and quantities, and effects on social welfare more accurately. For ex-post studies, granular medicine level data and rigorous empirical strategies could potentially isolate the causal effect of IP policy change on access to medicines directly at the patient level. This can be done in combination with qualitative research on perceptions of relevant stakeholders to changes in IP provisions and access to medicines. One disadvantage of the simplicity of ex-ante models is that it makes assumptions that fail to mirror the complexities of the relationships between variables in the real world, for example, the price differential of a product before and after going off patent, and the constant price elasticity of demand. It is usually difficult to estimate realistic values for key variables.

On the other hand, the challenges of dynamic complexity in the public health ecosystem may be effectively addressed with the modeling methodology of system dynamics. The methodology involves development of causal diagrams and policy-oriented computer simulation models that are unique to each problem setting [37]. The International System Dynamics Society was established in 1983, and within the society a special interest group on health issues was organized in 2003 [37]. System dynamics uses computerized models in which alternative policies and scenarios can be tested in a systematic way that answers both “what if” and “why.”

Our review has several limitations that should be taken into considerations. We limited our search to seven search engines, which may have resulted in missing relevant studies. We also did not perform a meta-analysis for data synthesis because of the variation in outcome variables chosen.

Many people lack access to medicines, particularly in low and middle income countries, even without any IP protection laws. Imposing IP protection laws or strengthening these laws as a result of trade agreements may further reduce the access to medicines. The magnitude of the effect on different outcome variables such as price, medicines expenditure and consumer welfare differ depending on a host of factors, most importantly domestic policies in place to counteract the potential negative effects on access. More studies are necessary to fill the gap in understanding the mechanisms through which changes in IP affect medicines access and which outcomes relevant to access are most effected by which type of changes in the IP.