The health consequences of falsified medicines‐ A study of the published literature

To analyse and present the literature describing the health consequences of falsified medicines, focusing on mortality and morbidity, as well as the scale of the issue, the geographic extent, the medicines affected, and the harm caused at both the individual and population levels.


Introduction
Falsified medicines may lead to avoidable morbidity, mortality, drug resistance, early death or treatment failure, as well as loss of faith in health systems, especially in lowincome and middle-income countries, and therefore a reliable supply of good-quality medicines is essential for public health [1][2][3][4]. However, it has been difficult to quantify the impact on patients' health, because of fragmented and incomplete reporting of incidents and the consequences of falsified medicines range from no effect at all to enabling disease progression [5], or to lethal toxicity, as in the case of diethylene glycol-containing cough syrup [6,7]. Also, inadequate doses of anti-infectives may lead to drug resistance [8,9]. With the exponential increase in internet connectivity, those engaged in distribution of falsified medical products have gained access to a global market place [10], and the growth of a culture of self-diagnosis and self-prescribing has led to the emergence of thousands of unregulated websites providing unsupervised access to medical products with no guarantee of authenticity [11]. Thus, what was once considered a problem suffered by developing and low-income countries has now become an issue for all, although low-and middle-income countries and areas of conflict or unrest are still most vulnerable to falsified medical products owing to inadequate health-related regulatory systems.
The purpose of this research is to study the literature describing the health consequences of falsified medicines, focusing on mortality and morbidity, as well as the scale of the issue, the geographic extent, the medicines affected and the harm caused at both the individual and population levels. We do not attempt to present an exhaustive analysis of every health consequence of substandard or falsified drugs, but aim to give an overview of the actual impact of these products on public health. It should be noted that this article deals only with falsified drugs and not with substandard drugs resulting from inadequate quality control during manufacture or distribution, although such drugs are also likely to have a substantial impact on public health.

Methods
This study is based on searches performed in PubMed. Four authors independently performed the literature search, selection of relevant articles and data extraction. Initially, we searched in several databases using different combinations of keywords to identify relevant articles, aiming to optimize keyword and source selection. Keywords were extracted from the abstracts of the identified articles using IBM SPSS Text Analytics (IBM Japan, Tokyo, Japan), and we finally selected '(counterfeit OR fake OR bogus OR falsified OR spurious) AND (medicine OR drug)'. Searches of Google Scholar, Scopus, Web of Science and PubMed using these keywords showed that the PubMed database search gave the most comprehensive result, whereas searches of the other databases missed some articles. Therefore, PubMed was selected as the preferred database for the present purpose.
Based on the above findings, a comprehensive search was conducted in all articles listed in PubMed up to Feb. 2017; the earliest identified publication dated from 1972. From the hits, we selected English language articles describing damage to patients' health due to falsified medicines. Among the 2006 articles hit in the final search using the selected key words, full-length English language articles amounted to 1791. These articles were manually searched to find those that contained a description of casualties related to falsified medicines; finally, 81 publications were selected as meeting this criterion. Publication year was not taken into consideration. Data from publications were included and tabulated if they provided examples of serious health hazards, adverse reactions, injury or deaths. When possible, primary sources (reporting the outbreak or case studies) have been cited and the references were divided into primary and secondary sources. There were some articles among the searched articles in which the terms false, deliberate contamination or adulteration were used to describe apparently falsified drugs. Incidents described in these articles were judged as falsified medicine incidents and data from these articles were included.

Statistical analysis
Results are summarised as descriptive statistics and expressed as number and percent. Graphs were prepared using Microsoft Excel 2010. IBM SPSS Text Analytics (IBM Japan, Tokyo, Japan), 2013 version was used for the keyword selection.

Results and discussion
We identified 81 articles describing 48 incidents related to falsified drugs (Table 1), resulting in approximately 7200 casualties including 3604 deaths. Among the articles identified, several are not listed in Table 1, and the data are not included in the figures, because these reports were considered incorrect or unreliable, as described below. For example, there have been several reports that in 2001 alone, a total of 192 000 people died due to fake medicines in China [5,[12][13][14][15]. However, the original article by Cockburn et al [16]., incorporated translational mistakes by the San Francisco Examiner, which were traced to the Shenzhen Evening News in Fackler's article [17] and corrected later by Cockburn et al [18]. There was another report in which diethylene glycol contamination of paracetamol syrup was claimed to have been responsible for the death of the same number of people in China [19], but it seems possible that this refers to the same incident. Other reports mentioned the death of 700 000 people due to fake antimalarial and tuberculosis drugs, but gave no details as to the country involved, or the year [8,9].

Geographical distribution of incidents
Among the total reported incidents involving health damage due to falsified medicines (n = 48), 27 (56.3%) occurred in developing countries and 21 (43.7%) in developed countries, as classified according to the World Bank country and lending groups country classification [20]. The distribution of these incidents among regions is shown in Figure 1.
The 2006 estimate of falsified medicines by WHO indicated that the prevalence of falsified medicines ranged from less than 1% in developed countries to over 10% developing countries [3,4,21]. But, it is noteworthy that our results show that the difference between developing and developed countries is quite small in terms of the number of incidents where falsified pharmaceuticals actually impacted on human health. What was once regarded a problem mostly affecting developing and low-income countries now seems a serious issue for all [22,23]. Surprisingly, the USA alone accounted for around 25% of the total incidents reported. Since 2001, at least 10 drug falsification incidents affecting patients were reported, although the US-FDA suggests that drug falsifying occurs less frequently in the U.S. than in other countries due to

Therapeutic categories of falsified drugs
Among the categories of falsified drugs that caused health damage, sedatives, hypnotics, narcotics and drugs for sexual dysfunction were the most common in both In some cases, different numbers of injuries in the same incident were given in different articles. In cases where multiple articles give the same number, that number is shown; otherwise the maximum reported number of injuries for each incident is shown. Some cases were regarded as duplicates, because health damage, occurrence year, country, drugs and cause were the same. Some cases are not included in the table because the reports gave very few details and could not be verified, and they were considered incorrect or unreliable (see text for details).
developing and developed countries. However, patients from developing or under-developed countries were mainly affected by falsified antipyretics, analgesics, and antitussives, such as paracetamol elixir, cough syrup, or teething mixture containing diethylene glycol [26,73].
Overall, the number of incidents by drug therapeutic category is illustrated in Figure 2.

Annual trends in number of incidents and number of persons affected by falsified medicines
According to the Pharmaceutical Security Institute (PSI), the global incidence of drug falsifying has increased by 51% between 2011 and 2015, with 2015 seeing the highest level of falsifying to date, a 38% increase compared with 2014. The Institute documented 3002 incidents of pharmaceutical crime during 2015 alone [94]. Regarding   Table 1 were excluded, as the year of occurrence was not stated.
health damage from falsified drugs, Figure 3 shows the numbers of incidents during various time periods. There is no clear trend to suggest that the number of incidents or the numbers of persons affected has been decreasing over time, although the number of victims is currently much smaller than in the 1990s ( Table 1). The largest number of people seemed to have been affected during the period of 1992 to 1996 (more than 2500), followed by 1997 to 2001 (more than 1500) and 2012 to 2016 (more than 1000). Only 200 people were affected between 2002 and 2006. These figures exclude cases where insufficient information was available in published reports. However, it is possible that reporting of incidents of health damage due to falsified medicines has become more comprehensive.

Nature of drug falsification
The identified incidents involved many kinds of falsified drug products. Some did not contain any active ingredient (meningitis vaccine containing no active ingredient or just salt water) [8,14,53], or included the active ingredient in harmful amounts (traditional antidiabetic medicine containing six times the normal dose of glibenclamide) [5,74,80]. Others involved a completely different active ingredient or incorrect formulation, or contained unacceptably high levels of impurities (cyanide-laced paracetamol or zolpidem and acetaminophen laced with metal) [16,61,67,70]. Looking at the health damage caused by falsified drugs at Table 1, the most common category (nine cases) covered antipyretic, analgesics, (acetaminophen, paracetamol) antitussive medication (cough syrup, paracetamol, dextromethorphan), and in eight of them, diethylene glycol was present as a contaminant (either deliberate or accidental) in the drug. Mass poisoning incidents with diethylene glycol have occurred in a number of countries over a long period. Medicines containing an incorrect amount of API or a totally different kind of API appeared in at least 12 cases. Another common finding was endotoxin and microbial contamination of ophthalmic products or infusions.

Limitations of the study
This study has several limitations. Firstly, it covers only English language articles in the PubMed database. In some of the reported cases, there was no information about specific drug involvement, exact number of patients harmed, year of occurrence or the location of the incident. Also, for many of the incidents, there was no primary source of data but only a review(s) summarising the incident details. In particular, several incidents that were claimed to involve hundreds of thousands of people were very poorly described. It should also be borne in mind that the results might be biased by differences in the effectiveness of reporting systems among countries; for example, less serious incidents in under-developed countries might not have been reported. In addition, different reports sometimes gave conflicting information about the same incident. Thus, we cannot estimate the true extent of the problem.

Conclusion
It is clear from the results of this study that falsified medicines impact both directly and indirectly on global public health and remain a serious problem. A wide range of medicines has been falsified in a variety of ways, and our findings may be helpful to identify particular causes for concern, such as deliberate or accidental contamination with ethylene glycol. Although many of the reports identified in our study only provided seriously inadequate or even conflicting data, we believe this study of reported drug falsification incidents involving health damage will be useful to focus attention on the potential scale of the problem, and may provide a basis for further studies in the future. Recognition of the problem, coordination of responses and active engagement of key stakeholders will be essential in combating transnational pharmaceutical crime, and reducing the human cost of falsified medicines.