Data from the Pharmaceutical Security Institute indicate that poor-quality medicines were found in 124 countries in 2011, with the problem more severe in low- and mid-income countries than in developed countries (IOM 2013). While much attention has been focused on intellectual property rights protection (notably issues surrounding the WTO’s TRIPS1 agreement), poor-quality samples were more prevalent in cheap, generic drugs than in expensive, innovator-branded drugs when we tested drug samples from 18 low-to-mid-income countries (Bate et al. 2011). Moreover, many people in developing countries have to rely on cheaper, generic drugs for most diseases, and the percentage of imports in drug supply is as high as 70% in African countries (UNAIDS 2013). From the public-health perspective, international trade is arguably more important on the low end than on the high end of drug quality.
There is a widespread perception amongst pharmaceutical experts that some Indian manufacturers take advantage of the poor regulatory environment in certain countries to sell them products of inferior quality. When Ranbaxy, a major Indian generic drug manufacturer, was charged with selling adulterated drugs, fabricating data, and committing fraud, Dinesh Thakur, then director and global head of research information and ultimately the company’s whistleblower, described how “Ranbaxy took its greatest liberties in markets where regulation was weakest and the risk of discovery was lowest” (Eban 2013). In a recent National Bureau of Economic Research working paper (Bate et al. 2014b), we tested to see if this perception is validated by the data.
To conduct this analysis, we used covert shoppers in five cities inside India and 17 low- to middle-income countries outside of India to buy 1,470 samples of antibiotics and tuberculosis medicines that were ostensibly ‘made in India’, as per the labelling on the package. We focused on these drugs because broad-spectrum antibiotics and specialised tuberculosis medicines in solid oral form are among the most commonly used in developing countries. All these medicines were assessed using semi-quantitative thin-layer chromatography, which tests for the presence of the active pharmaceutical ingredient (API) in the sample and compares it to a reference standard. In our study, we refer to a drug sample as falsified if it contains zero correct API, and as substandard if it contains less than 80% of the correct API.
Africa seems to get substandard drugs
We found that 10.9% of samples failed basic quality tests. Within these, both antibiotics and tuberculosis drugs had more substandard than falsified products, suggesting that poor drug quality might be a result of negligence rather than outright crime. Moreover, drugs purchased from Africa were more likely to fail the thin-layer chromatography test than the same type of drugs in the Indian domestic or non-Africa groups. There are also interesting differences between the passing rate of registered and non-registered drugs. The passing rates of registered drugs were similar across Africa, India, and non-Africa, averaging more than 90%. However, among non-registered products, the passing rate was less than 50% in Africa and 67% in India, and the majority of failures were driven by substandard drugs. In short, the same manufacturers’ drugs (as labelled on the package) are of lower quality in Africa than in India and non-Africa.
There are several possible parties responsible for the poor-quality drugs.
- First, some Indian manufacturers might intentionally export inferior products to Africa.
This could happen because African countries are typically poorer, have a less educated population, and do not function well in regulating drug quality (Seiter 2010).
- Second, counterfeiters might focus on Africa because the risk of being caught is lower there.
- Third, wholesale distributors might obtain the same good-quality drugs from India, but do a worse job in storing and distributing drugs in Africa.
This could happen either because the cost of proper storage is too high in Africa or because distributors cut corners intentionally.
Poor distribution is unlikely to be the main culprit
While poor distribution undoubtedly occurs in some settings (Bate 2012), it is unlikely to reduce API from 100% to zero. Hence, poor distribution cannot explain falsified products. Moreover, our previous paper analysed a larger dataset of ciprofloxacin samples including both drugs approved by a stringent regulatory authority (SRA) and other types (Bate et al. 2014a), where SRA approval refers to production approved by at least one Western country with a stringent quality standard (e.g. the US). That paper found that SRA-approved ciprofloxacin drugs, if containing any correct API, always passed the basic quality test regardless of whether they were purchased from Africa or elsewhere.2 This suggests that degradation should not be the main factor driving poor drug quality in Africa. Additionally, a study undertaken in Ghana found poor performance of ‘Indian-made’ products, but no product quality problems for European manufactured medicines sampled (USP 2013), suggesting degradation is not a major factor (at least in formal pharmacies). For this reason, we ignore the role of distributors and focus on the potential identity of manufacturer.
Economic incentives facing manufacturers and counterfeiters
To make an intelligent guess of whether the true manufacturer is the labelled Indian firm or a counterfeiter, we argue that the incentives are different for firms registering products, those that do not, and outright counterfeiters. Since product registration is costly, registered products enjoy higher prices on the market, and selling unregistered products is illegal, registered firms should have more incentive to produce good quality products. For counterfeiters, the main tradeoff is the higher profit of counterfeiting certain products versus the potentially higher risk of being caught. Economic incentives suggest that counterfeiters are most likely to counterfeit registered products and produce the worst quality (falsified) drug, as long as drug quality is not observable to consumers, the penalty for counterfeiting is independent of drug quality, and the chance of being caught counterfeiting is about the same for counterfeiting registered and unregistered products.
These arguments imply that the substandard drugs in our data are unlikely to be driven by counterfeiters or firms registering products. They are most likely driven by Indian firms that do not bother to register products in the countries they sell to. On the Pharmalot blog of the Wall Street Journal, a consultant working for the Indian generics industry suggested that we have not proven ‘intent’ of companies to sell inferior medicines, and it is possible that Indian firms sell to wholesalers (either in India or Africa) who then take the responsibility for (non-)registration. In that scenario, the Indian firms doing so can claim they do not intend to sell these products in a particular country. But the problem for them is that they are selling worse quality products to various middlemen and not worrying about where they end up. This does not eliminate the manufacturers’ responsibility, and suggests that poor-quality products would end up in places with poor oversight.
To further explain why drug quality is worse in Africa, we collected city- or country-specific data on GDP per capita, adult literacy rates, price regulations, maximum penalties for counterfeiting, the Rule of Law Index, and the International Property Rights Index. Our results suggest that GDP per capita plays an important role in explaining quality differences across country groups, though other characteristics such as literacy rates and price regulations also seem to matter.
To conclude, our sample of ‘Indian-made’ medicines reveals two patterns:
- First, drug quality is poorer among drugs purchased inside African countries than among those purchased inside India or middle-income countries.
- Second, substandard drugs that contain insufficient API and are non-registered in the African destination are the biggest driver of this quality difference.
These findings are based on crude API assessments of a limited number of drug samples (1,470) across a limited number of Indian manufacturers (17), thus their generalizability is subject to future research.
That being said, these findings support what has been known anecdotally for years, that some Indian drug companies and/or their distributors segment the global medicine market into portions that are served by different quality medicines. Although this paper has focused on Indian-produced medicines, India is by no means the only large exporter of drugs. Further research into the drug quality of China and other export countries would be useful to understand how widespread the problem may be.
Bate, R (2012), Phake: The Deadly World of Falsified and Substandard Medicines, American Enterprise Institute Press.
Bate, R, G Z Jin, and A Mathur (2011), “Does Price Reveal Poor-Quality Drugs? Evidence from 17 Countries”, Journal of Health Economics 30(6): 1150–1163.
Bate, R, G Z Jin, and A Mathur (2014a), “Falsified or Substandard? Assessing Price and Non-Price Signals of Drug Quality”, accepted by the Journal of Economics & Management Strategy. Also available as NBER Working Paper 18073.
Bate, R, G Z Jin, A Mathur, and A Attaran (2014b), “Poor Quality Drugs and Global Trade: A Pilot Study”, NBER Working Paper 20469.
Eban, K (2013), “Dirty Medicine”, Fortune, 15 May.
Institute of Medicine (IOM) (2013), Countering the Problem of Falsified and Substandard Drugs, National Academy of Sciences.
Seiter, A (2010), A Practical Approach to Pharmaceutical Policy, World Bank Press.
UNAIDS (2013), “African Leaders call for greater industrialization of an emerging Africa”, 26 March.
United States Pharmacopeial Convention (USP) (2013), “Post-Market Quality Surveillance Project: Maternal Healthcare Products (Oxytocin and Ergometrine) on the Ghanaian Market”.
1 TRIPS stands for Trade-Related Aspects of Intellectual Property Rights, an international agreement on the standard and enforcement of intellectual property rights.
2 As reported in Bate et al. (2014a), there are 89 SRA-approved ciprofloxacin drugs in our data. 88 of them passed the basic quality test and one was found to be falsified.