Annually, thousands of individuals die and tens of thousands are hospitalized in association with suspected adverse drug reactions (ADRs) in Canada. We analyzed the reports from the Canada Vigilance Adverse Reaction online database and present a synopsis of the state of ADRs in Canada between 2009 and 2018. Our synopsis includes both cross-sectional and longitudinal insights into ADR demographics, outcomes, associated drugs and disease indications. In closing, we highlight five overarching issues uncovered in our analysis, which have potential implications for future policy formulation. Further in-depth exploration is required to shine some additional light on these issues.
An adverse drug reaction (ADR) is a noxious and unintended response to a pharmaceutical product that occurs during its recommended clinical use (Nebeker et al. 2004). ADR is a subtype of an adverse drug event where the causal relationship between a drug and an adverse event has been reasonably suspected or established (Shepherd et al. 2012). A serious ADR may potentially lead to prolonged hospitalization, permanent disability, a life-threatening condition or even death. ADR is one of the leading causes of morbidity and mortality in healthcare settings (Lazarou et al. 1999; Wester et al. 2008). The Canada Vigilance Adverse Reaction (CVAR) online database contains Canadian ADR records since 1965 (Government of Canada 2019b).
We analyzed the ADR reports from the database dated between 2009 and 2018. First, different data files containing report information, such as ADR outcomes, patient demographics, drug information and disease indication, were combined using unique Report ID identifiers. Next, duplicate reports were eliminated based on the list provided in the database. Finally, reports were stratified based on report year, patient gender, patient age, outcome, report source, drug ingredient, therapeutic class and disease indication. Population data were obtained from Statistics Canada (2019). All data analyses were performed using R. The database has some limitations, such as underreporting, missing data and data duplication. A detailed list of the limitations can be found at the CVAR database website.
Although the CVAR database has been tracking ADR data since 1965, two-thirds of the ADRs in the database were reported in the past 10 years. About 437,000 suspected ADRs were reported between 2009 and 2018 (Figure 1a). During this period, the number of reports per million population changed significantly from 531 (2009) to 2,173 (2018), indicating a more than fourfold increase in the reporting rate (Figure 1a).
The database contains post-market ADR reports that have been collected from various sources primarily by a spontaneous surveillance system. Source refers to the location through which a reporter sent the report. In the period of 2009 to 2018, there has been a substantial shift in ADR report sources in the database (Figure 1b). In 2009, market authorization holder, community and hospitals accounted for 67%, 26% and 6% of the report sources, respectively. Clinical studies were only recorded as a source for fewer than 100 reports. In 2018, clinical studies accounted for 46% of all the reports – a total of 37,000 records. Clinical studies include clinical trials in which the conditions experienced by patients significantly differ from the "real-world" condition (Kim et al. 2018). Because a substantial presence of the ADR reports from clinical studies could fundamentally alter our conclusions about the real-world ADRs, we performed a side-by-side analysis of the ADR reports with and without clinical studies. A 10-year average of most ADR indicators, such as the rate of serious outcomes, top ADR drugs and disease indications, did not exhibit any substantial difference between the two groups. In this report, we only present the results of our analysis of the entire ADR data set, which includes clinical studies.
A number of ADR patients suffered serious outcomes, including death, disability and unplanned in-patient hospitalization. According to the CVAR database, 8% of the above-reported 437,000 ADR patients died, 2% were disabled and 22% required in-patient hospitalization (Figure 1c). These are large numbers, underscoring the scale of the threat from this issue. To provide some context, according to Canadian motor vehicle collision statistics, 1,841 people died because of traffic accidents in 2017 (Government of Canada 2019a), which is the latest year for which the motor vehicle accident report was available from Transport Canada during the preparation of this article. In the same year, 4,200 people reportedly died from suspected ADRs, even though ADR is still considered an underreported phenomenon.
The age distribution of the patients exhibited the expected trend. Older patients are more abundant in the reports – patients between ages 51 and 70 years represent 40% of all the reports (Figure 2a, blue and orange bars). When normalized by the population of the age group (Statistics Canada 2018), both 70- to 79-year-old men and women exhibit the highest rate of ADRs. The sex distribution of the ADR outcomes is quite revealing (Figure 2b). The ratio of female patients to male patients in the ADR reports is 1.6 to 1 (Figure 2b, "All reports"). This trend is partially accounted for by the use of prescription medicines by the two sexes. According to a Canadian Institute for Health Information (CIHI) report on public drug spending, the ratio of the total number of female users to male users is 1.22 to 1 (CIHI 2019a). However, when it comes to serious ADR outcomes such as death, the sex overrepresentation reverses (Figure 2b, "Death"). Male and female patients accounted for 47% and 43% of the ADR-associated deaths from 2009 to 2018, respectively. The total is not 100% because a number of reports do not contain any sex information. Based on these numbers, we conclude that male ADR patients have a 75% higher risk of ADR-associated death compared to their female counterparts. It remains to be determined what percentage of these deaths are avoidable deaths, how we can reduce the risk of fatal ADR outcomes and if those risk-reduction strategies should differ by sex.
Number of drugs
About half of all ADR reports list only one drug (Figure 3a). Although this may encourage one to quickly establish a causal relationship between an adverse reaction and the associated pharmaceutical product, it is important to remember that the information provided in the ADR reports is inadequate for this purpose. Confirming such a relationship requires the knowledge of many additional factors and circumstances, such as strength of the association, temporality and plausibility (Agbabiaka et al. 2008). Nevertheless, a strong ADR association makes a good case for a follow-up in-depth analysis.
Overall, we observed that the average number of drugs listed per report increased between 2009 (3.26 drugs/report) and 2018 (3.82 drugs/report). Reports with serious adverse outcomes such as hospitalization and disability have a higher average number of drugs listed per report compared to that for all the reports. For the 10-year period, the average for the number of drugs listed per report for all reports is 3.38. The average rises to 4.58 and 4.18 when calculated exclusively for disability and hospitalization, respectively. In contrast to this trend, the average number of drugs listed per report is 3.06 for the patients who suffered death. The use of multiple drug products indicates a higher chance of unwanted drug–drug interactions and reduced patient morbidity. Additional research is needed to untangle the contribution of these two factors toward ADR outcomes.
Among the thousands of different drug ingredients listed in the ADR reports, biologics rank high in the list (Figure 3b). Together, two tumour necrosis factor-α (TNF-α) inhibitors, infliximab and adalimumab, are indicated in 20% of all the reports and account for 7% of all drug occurrences (i.e., the number of times a particular drug product is listed in the database) between 2009 and 2018. TNF-α inhibitors are typically used for treating autoimmune or immune-mediated diseases, such as rheumatoid arthritis and inflammatory bowel disease (Lis et al. 2014). These two products are followed by methotrexate, which accounts for 3% of all drug occurrences in the same period. Methotrexate is used to treat certain types of cancers and autoimmune diseases. Among the top ADR drug ingredients, methotrexate has experienced a fivefold growth, from 1% in 2009 to 5% in 2018. In 2018, methotrexate tops all other drug ingredients for drug occurrence. For the Anatomical Therapeutic Chemical (ATC) class (ATC/DDD Index 2020), immunosuppressants (L04) account for 17% of all drug occurrences, followed by analgesic drugs (N02, 5%) and drugs for acid-related disorders (A02, 3%; Figure 3c). Immunosuppressants are defined as agents that completely or partly suppress one or more factors in the immunosystem. Infliximab, adalimumab and methotrexate belong to the immunosuppressants therapeutic class (L04). Analgesics belonging to the N02 therapeutic class are typically pain-relief medicines, and the A02 therapeutic class includes antacids and antiflatulents.
Our analysis revealed that 80% of the reports contain only one disease indication. In all, 99% of the reports contain four or fewer disease indications. Rheumatoid arthritis tops the list of indications, accounting for 12% of all indications reported (Figure 3d). However, this has not always been the case. In 2009, rheumatoid arthritis only represented 4% of all indications reported, indicating a threefold increase in 10 years (Figure 3d). Crohn's disease is the second most abundant ADR disease indication, representing 6% of all the indications reported in 2018.
The CVAR online database is a great source of many important pharmacovigilance insights. However, it is important that we remain aware of its limitations. According to the disclaimer provided with the database:
… accumulated case reports should not be used as a basis for determining the incidence of a reaction or estimating risk for a particular product as neither the total number of reactions occurring, nor the number of patients exposed to the health product is known (Health Canada 2011).
Nonetheless, these reports collectively serve as a real-world indicator of the ADRs in Canada. In this section, we highlight five overarching future research questions uncovered by our analysis of the CVAR database, which have the potential to inform policy decision making and improve safe delivery of pharmaceuticals to the public.
What are the main drivers of the recent increase in the number of ADR reports?
The observed fourfold increase in the number of reports per million population between 2009 and 2018 could be driven by many factors, such as better reporting due to policy intervention, per capita increase in the use of pharmaceutical products, changes in the suggested use of pharmaceutical products, substitution of the patented medicines by generics and biosimilars and inappropriate use. Information available in the CVAR database is not adequate to uncover the factors responsible for the average 17% year-to-year increase in the number of ADR reports per million during 2009 to 2018. Such information may be of extreme importance for future policies aiming to reduce the ADR risk. Nevertheless, we share some of our speculations here. Future in-depth studies are needed to judge the merits of them.
ADR reporting requirements have significantly evolved over the past couple of decades, increasing the share of mandatory reports in the database from different regulated parties as required by relevant regulations (Al Dweik et al. 2016). In 2013, Health Canada released the finalized guidance document for clinical trial sponsors (Health Canada 2013). Along with that, Health Canada rolled out its new electronic ADR reporting system (Health Canada 2013). Together, these two initiatives likely boosted the number of reports from clinical trials in the database. ADR reports from clinical studies, which include clinical trials, jumped from 3% to 7% from 2013 to 2014 and from 7% to 35% in the following year. In the database, hospitals represent only a small fraction of the ADR report sources, and this remained quite constant over the period of our analysis. In June 2019, Health Canada issued a mandatory ADR requirement for hospitals (Health Canada 2019). The impact of this regulatory requirement remains to be seen in the CVAR database in the coming years.
Increased use of biologics
Between 2009 and 2018, the sales of patented biologics increased by threefold from $2 billion to $6.7 billion (Lungu 2019). Biologics are typically associated with a higher risk of adverse reactions compared to the chemically synthesized drugs (Singh et al. 2011). Therefore, it is reasonable to think that some of the recent increase in ADR reports is due to an expanding user base of the biologic drugs. Additional in-depth analysis of the CVAR database is needed to find out the contribution of the use of biologics to the increased ADR rate. Here, it is also important to explore the role of biosimilars contributing to the ADR reports. The information about the real-world ADR rates of biosimilars compared to their biologic counterparts is a critical component in promoting a higher biosimilar adoption.
As described earlier, the average number of drugs listed per report increased from 3.26 to 3.82 from 2009 to 2018. Serious ADR outcomes tend to indicate a higher number of drugs listed per report. We speculate that an increasing use of multiple drugs in combination contributed to the observed increase in ADR reports. A detailed analysis of drug–drug combinations listed in the ADR reports will shine more light on this matter.
What is the overall economic burden of ADRs in Canada?
As Canada has been contemplating the rolling out of a universal pharmacare program, it is more important than ever to know the economic burden of ADRs to the health system and to society. However, we rarely come across articles that estimate a combined economic burden of all levels of ADRs. For example, Wu et al. (2012) computed the economic burden of ADRs among elderly patients in Ontario emergency departments. Although these estimates are useful, they do not present the fuller picture. Thus far, our search did not find any estimate of the total economic burden of ADRs in Canada. A 1999 article reported that the economic burden of ADRs in the United States had been estimated as at least US$30 billion and could be as much as US$130 billion in a worst-case scenario (White et al. 1999). Inflated to 2019 US dollars, it is about $46 billion and $200 billion, respectively. The CVAR database presents a unique opportunity to estimate the overall economic burden of ADRs in Canada. For example, according to CIHI, the average healthcare cost of ADR-related hospitalization in 2017 was $4,130, which implies an annual cost of more than $65 million for ADR-related hospitalization in 2018 (CIHI 2019b). Direct hospital cost is just one of the many components contributing to the economic burden of ADRs. Our report paves the path for a future estimate of the overall ADR-associated economic burden in Canada.
Why are men overrepresented in ADR-associated deaths?
As reported earlier, male patients have a 75% higher risk of ADR-associated deaths. Several factors may influence the outcome of an ADR (Yousaf et al. 2015). We can roughly group them into four categories: physiological (e.g., patient's general physiological condition and age), structural (e.g., access to healthcare system and quality of the healthcare providers), social (e.g., ability to pay for medical care and potential loss of wages as a result of receiving medical care) and psychological (e.g., willingness to seek medical help in a timely manner). Some of these factors are unavoidable (e.g., physiological) or beyond the control of an individual (e.g., structural). In contrast, psychological factors are modifiable at the individual level and, therefore, offer a faster way to minimize fatal ADR outcomes by changing behaviour. Based on previous reports on men's medical help-seeking behaviour, we speculate that psychological factors are significantly responsible for male overrepresentation in fatal ADR outcomes (Courtenay 2000; Galdas et al. 2010; George and Fleming, 2004). More than 33,000 people reportedly died in association with ADR between 2009 and 2018. Understanding the factors that could potentially reduce avoidable ADR-associated deaths may be useful in developing policies to mitigate male overrepresentation in serious ADR outcomes.
How does a relatively high rate of ADR impact the cost-effectiveness of TNF-α inhibitors?
The top two ADR drugs infliximab and adalimumab are TNF-α inhibitors, which belong to the class of drugs called biologics. Biologic drugs are produced from living organisms. Typically, biologics are complex mixtures that are not easily identifiable or characterizable. Biologics are different from traditional drugs in terms of chemistry, mode of action, metabolism and immunogenicity and are reportedly associated with a higher rate of ADRs (Aubin et al. 2013; Scherer et al. 2010). Based on the publicly available reports, infliximab and adalimumab had around 80,000 users in 2017 and accounted for $1.8 billion of prescription drug spending in Canada (Patented Medicine Prices Review Board 2018). The annual average cost of treatment for infliximab and adalimumab in 2017 was $29,500 and $16,400, respectively, when using brand-name products (Patented Medicine Prices Review Board 2018). Given the significance of their market share in overall prescription drug spending in Canada, it is critical to know how the high ADR rates associated with these two drugs impact their cost-effectiveness. ADRs influence the cost-effectiveness ratio, such as cost per quality-adjusted life-year (QALY), in a double-negative way. ADR-associated morbidity and mortality reduces the QALY gain in the denominator, whereas ADR-related healthcare and opportunity costs increase the cost in the numerator. A high ADR rate is likely to alter the cost-effectiveness of a drug in a significant way, especially when the expected QALY gain is low and the cost is high.
What changes resulted in the recent increased association between methotrexate and ADRs?
Methotrexate accounted for more than 40,000 ADR reports between 2009 and 2018. Rheumatoid arthritis accounts for about half of all disease indicators for methotrexate. Methotrexate is also used for treating multiple types of cancers and psoriasis. Although the drug has been around for the past 50 years, its high rate of association with ADRs is relatively recent. In 2009, methotrexate only accounted for 1% of all ADR–drug associations. In 2018, it accounted for 5% of all ADR–drug associations, indicating a fivefold increase. However, a direct causal role of methotrexate is harder to establish. In only 5% of the associated ADR reports, methotrexate is listed as a single agent. On the other hand, 23% of the methotrexate-associated ADRs have more than 10 drugs listed. Drug–drug interactions and the morbidity associated with various disease states make it a complex challenge to decipher the observed association of methotrexate with a greater number of ADR reports. Further in-depth analysis of the ADR reports is required to understand the source of methotrexate's rapid gain in associating with ADR reports starting around 2015.
Despite its limitations, the CVAR database is a unique source of real-world ADR reports in Canada. Our descriptive analysis of the reports offers an overview of the Canadian ADR landscape. Based on our analysis, we present five future research questions. Because a sizable fraction of all ADRs are avoidable, this study paves the way to additional in-depth exploration that will potentially contribute to reducing ADR-related morbidity and mortality, delivering better health outcomes and lowering healthcare costs.
About the Author(s)
Tuhin S. Maity, PhD, is currently a doctoral student at the DeGroote School of Business, McMaster University. Tuhin has previously earned a doctoral degree in chemistry and held various research positions. He can be reached at firstname.lastname@example.org
Christopher J. Longo, PhD, is an associate professor in health policy and management at DeGroote School of Business, McMaster University. Christopher has over 20 years of industry experience in clinical research, economic evaluation and market access strategies for pharmaceuticals.
Agbabiaka, T.B., J. Savovi? and E. Ernst. 2008. Methods for Causality Assessment of Adverse Drug Reactions: A Systematic Review. Drug Safety 31(1): 21–37. doi:10.2165/00002018-200831010-00003.
Al Dweik, R., S. Yaya, D. Stacey and D. Kohen. 2016. Spontaneous Adverse Drug Reaction Reporting by Patients in Canada: A Multi-Method Study—Study Protocol. SpringerPlus 5: 213. doi:10.1186/s40064-016-1838-9.
ATC/DDD Index 2020. ATC/DDD Index. Retrieved January 22, 2020. <https://www.whocc.no/atc_ddd_index/>.
Aubin, F., F. Carbonnel and D. Wendling. 2013. The Complexity of Adverse Side-Effects to Biological Agents. Journal of Crohn's and Colitis 7(4): 257–62. doi:10.1016/j.crohns.2012.06.024.
Canadian Institute for Health Information (CIHI). 2019a. National Health Expenditure Trends, 1975 to 2019. Retrieved October 30, 2019. <https://www.cihi.ca/sites/default/files/document/nhex-trends-narrative-report-2019-en-web.pdf>.
Canadian Institute for Health Information (CIHI). 2019b. Patient Cost Estimator. Retrieved January 24, 2020. <www.cihi.ca/en/patient-cost-estimator>.
Courtenay, W.H. 2000. Constructions of Masculinity and Their Influence on Men's Well-Being: A Theory of Gender and Health. Social Science and Medicine 50(10): 1385–1401. doi:10.1016/S0277-9536(99)00390-1.
Galdas, P.M., J.L. Johnson, M.E. Percy and P.A. Ratner. 2010. Help Seeking for Cardiac Symptoms: beyond the Masculine–Feminine Binary. Social Science and Medicine 71(1): 18–24. doi:10.1016/j.socscimed.2010.03.006.
George, A. and P. Fleming. 2004. Factors Affecting Men's Help-Seeking in the Early Detection of Prostate Cancer: Implications for Health Promotion. Journal of Men's Health and Gender 1(4): 345–52. doi:10.1016/j.jmhg.2004.10.009.
Government of Canada. 2019a, February 27. Canadian Motor Vehicle Traffic Collision Statistics: 2017. Retrieved October 30, 2019. <https://www.tc.gc.ca/eng/motorvehiclesafety/canadian-motor-vehicle-traffic-collision-statistics-2017.html>.
Government of Canada. 2019b, May. Data Extracts from the Canada Vigilance Adverse Reaction Online Database. Retrieved May 9, 2019. <https://www.canada.ca/en/health-canada/services/drugs-health-products/medeffect-canada/adverse-reaction-database/canada-vigilance-online-database-data-extract.html>.
Health Canada. 2011. Procedure – The Release to the Public of Information Obtained from Adverse Reaction and Medical Device Incident Reports. Retrieved October 30, 2019. <https://www.canada.ca/content/dam/hc-sc/migration/hc-sc/dhp-mps/alt_formats/pdf/pubs/medeff/guide/2011-procedure/_2011-procedure-eng.pdf>.
Health Canada. 2013, May 29. Guidance Document for Clinical Trial Sponsors: Clinical Trial Applications. Retrieved November 13, 2019. <https://www.canada.ca/en/health-canada/services/drugs-health-products/drug-products/applications-submissions/guidance-documents/clinical-trials/clinical-trial-sponsors-applications.html>.
Health Canada. 2019. Mandatory Reporting of Serious Adverse Drug Reactions and Medical Device Incidents by Hospitals: Guidance Document. Retrieved November 13, 2019. <www.canada.ca/en/health-canada/services/drugs-health-products/medeffect-canada/adverse-reaction-reporting/mandatory-hospital-reporting/drugs-devices/guidance.html>.
Kim, H.-S., S. Lee and J.H. Kim. 2018. Real-World Evidence versus Randomized Controlled Trial: Clinical Research Based on Electronic Medical Records. Journal of Korean Medical Science 33(34): e213. doi:10.3346/jkms.2018.33.e213.
Lazarou, J., B.H. Pomeranz and P.N. Corey. 1999. Incidence of Adverse Drug Reactions in Hospitalized Patients: A Meta-Analysis of Prospective Studies. Survey of Anesthesiology 43(1): 53–54. doi:10.1001/jama.279.15.1200.
Lis, K., O. Kuzawi?ska and E. Ba?kowiec-Iskra. 2014. Tumor Necrosis Factor Inhibitors–State of Knowledge. Archives of Medical Science 10(6): 1175–85. doi:10.5114/aoms.2014.47827.
Lungu, E. 2019, April. Biosimilars in Canada: Current Environment and Future Opportunity. Retrieved October 30, 2019. <http://www.pmprb-cepmb.gc.ca/CMFiles/News%20and%20Events/Speeches/biosimilars-april2019-en.pdf>.
Nebeker, J.R., P. Barach and M.H. Samore. 2004. Clarifying Adverse Drug Events: A Clinician's Guide to Terminology, Documentation, and Reporting. Annals of Internal Medicine 140(10): 795–801. doi:10.7326/0003-4819-140-10-200405180-00009.
Patented Medicine Prices Review Board. 2018. Patented Medicine Prices Review Board: Annual Report 2017. Retrieved October 31, 2019. <https://www.pmprb-cepmb.gc.ca/view.asp?ccid=1380>.
Scherer, K., D. Spoerl and A.J. Bircher. 2010. Adverse Drug Reactions to Biologics. Journal of the German Society of Dermatology 8(6): 411–26. doi:10.1111/j.1610-0387.2010.07339.x.
Shepherd, G., P. Mohorn, K. Yacoub and D. Williams. May. 2012. Adverse Drug Reaction Deaths Reported in United States Vital Statistics, 1999–2006. Annals of Pharmacotherapy 46(2): 169–75. doi:10.1345/aph.1P592.
Singh, J.A., G.A. Wells, R. Christensen, E.T. Ghogomu, L.J. Maxwell, J.K. MacDonald et al. 2011. Adverse Effects of Biologics: A Network Meta-Analysis and Cochrane Overview. Cochrane Database of Systematic Reviews 16(2): CD008794. doi:10.1002/14651858.CD008794.pub2.
Statistics Canada. 2018. Life Expectancy, at Birth and at Age 65, by Sex, Five-Year Average, Canada and Inuit Regions. Retrieved October 30, 2019. <https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1310040301>.
Statistics Canada. 2019. Population Estimates on July 1st, by Age and Sex. Retrieved November 13, 2019. <https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1710000501>.
Wester, K., A.K. Jönsson, O. Spigset, H. Druid and S. Hägg. 2008. Incidence of Fatal Adverse Drug Reactions: A Population Based Study. British Journal of Clinical Pharmacology 65(4): 573–79. doi:10.1111/j.1365-2125.2007.03064.x.
White, T.J., A. Arakelian and J.P. Rho. 1999. Counting the Costs of Drug-Related Adverse Events. PharmacoEconomics 15(5): 445–58. doi:10.2165/00019053-199915050-00003.
Wu, C., C.M. Bell and W.P. Wodchis. 2012. Incidence and Economic Burden of Adverse Drug Reactions among Elderly Patients in Ontario Emergency Departments: A Retrospective Study. Drug Safety 35(9): 769–81. doi:10.1007/BF03261973.
Yousaf, O., E.A. Grunfeld and M.S. Hunter. 2015. A Systematic Review of the Factors Associated with Delays in Medical and Psychological Help-Seeking among Men. Health Psychology Review 9(2): 264–76. doi:10.1080/17437199.2013.840954.
Be the first to comment on this!
This article is for subscribers only. To view the entire article
Note: Please enter a display name. Your email address will not be publically displayed