Why is medicine safety important for resilient health systems?
Huihui Wang, Patricio V Marquez, Albert Figueras
Background
In a health system, resilience is the ability to prepare, manage (absorb, adapt and transform), and learn from shocks. These shocks, as differing from predictable and enduring stresses, are sudden and extreme changes that impact on a health system, such as the COVID-19 pandemic. Medicines, vaccines, and other therapies, are critical countermeasures during a public health crisis, as well as in normal times. The global emergency response to the pandemic has required the introduction of various types of vaccines, the repurposing of several existing medicines, and the adoption of new antiviral therapies. Although indispensable for improving health outcomes, medicines and vaccines or their administration and use can produce adverse effects, requiring continuous vigilance to ensure that their health benefits outweigh the risks in different populations and situations.
The Significant Toll of Adverse Drug Events on Health Systems
As advised by the U.S. Federal Drug Administration (FDA), the benefits of medicines are the helpful effects a person gets when using them, such as lowering blood pressure, curing infection, or relieving pain. The risks of medicines are the chances that something unwanted or unexpected could happen to you when you use them---the possibility of a harmful interaction between the medicine and a food, beverage, dietary supplement (including vitamins and herbals), or another medicine; the combinations of any of these products could increase the chance that there may be interactions; the chance that the medicine may not work as expected; possibility that the medicine may cause additional problems.
Some of the risks of medicines are manifested as adverse drug events (ADEs) or a harm that occurs while a patient is taking a medicine, irrespective of whether the medicine is suspected to be the cause, and adverse drug reactions (ADRs) or an unwanted or harmful reaction experienced following the administration of a medicine or combination of medicines under normal conditions of use and is suspected to be related to or caused by the medicines, are a significant challenge to health systems, particularly because of the increasing complexity of therapeutics, an aging population, and multiple comorbidities. Although most ADRs are mild, serious ADRs sometimes lead to (i) clinical complications among patients who are already frail, (ii) longer hospital stays or longer work leaves for additional treatments to manage the ADRs, (iii) rising health care costs, and (iv) the occasional death of the patient.
The recent pandemic experience shows that while vaccination against COVID-19 provides clear public health benefits, vaccination also carries potential risks. Reports of myocarditis (inflammation of the heart muscle) and pericarditis (inflammation of the outer lining of the heart) after SARS-CoV-2 messenger RNA (mRNA) vaccination emerged. In both cases, the body’s immune system causes inflammation in response to an infection or some other trigger. Case reports, surveillance data, and other reports from the US, Israel, and Canada indicate an increased risk of myocarditis after vaccination with SARS-CoV-2mRNA vaccines, higher after the second dose, especially in younger men. Likewise, the results of a Nordic Cohort Study of 23 million people indicated that both first and second doses of mRNA vaccines were associated with increased risk of myocarditis and pericarditis. For individuals receiving 2 doses of the same vaccine, risk of myocarditis was highest among young males (aged 16-24 years) after the second dose. Most patients with myocarditis or pericarditis after COVID-19 vaccination responded well to medicine and rest and felt better quickly. Reports of death after COVID-19 vaccination are rare. The benefits of COVID-19 vaccination in protecting against severe COVID-19 disease, hospitalization, and death, continue to outweigh any potential risks.
The burden of ADRs on health systems is clearly appreciated by looking at service utilization metrics. For example, the prevalence of hospital emergency department visits for ADRs in the United States was estimated at 4 per 1,000 visits in 2013 and 2014, and anticoagulants, antibiotics, diabetes agents, and opioid analgesics were the most common drug classes implicated. Recent work at the CDC has estimated that more than 1 million individuals are seen in hospital emergency departments for ADRs each year in the United States; more than one-quarter of these patients must be hospitalized for further treatment. Similarly, a European Commission report documented that 3 percent–10 percent of hospital admissions between 2012 and 2014 were estimated to have been associated with ADRs, totaling about 2.5 million–8.4 million annually, and 2.1 percent–6.5 percent of hospitalized patients experienced an ADR, corresponding to 1.8 million–5.5 million annually. The onset of at least one ADR during hospitalization was associated with a median prolongation in the hospital stay of four days, which is similar to the findings of another study (Nobili et al. 2011). A study using a census of hospital admission data from New South Wales, Australia, between July 2000 and June 2012 assessed the changes in incidence and their impact on length of stay, readmission, and in-hospital mortality. The authors conclude that the relative incidence of serious ADRs, at best, did not decrease between 2000 and 2012 and, in some cases, substantially increased. Additionally, serious ADRs were associated with a considerably longer hospital stay and a significant increase in the risk of readmission or in-hospital mortality for certain types of ADR.
Besides the clinical consequences of ADRs for the individual, ADRs are expensive. ADRs raise health care resource utilization, including for emergency department visits. Likewise, ADR-related hospitalization is a significant factor in rising direct medical costs. Indeed, the findings of a recent study in the Republic of Korea suggest that drug-related emergency department visits increase the burden on health insurance systems and the out-of-pocket costs to patients, mostly because of higher hospitalization costs. The findings of the study show that the mean cost per ADR rose by 26.1 percent during the six-month follow-up compared with the cost during the six months before the emergency department visit and that preventable ADRs accounted for approximately 19.9 percent of the cost increase among all ADR cases.
How can Pharmacovigilance Contribute to Build Resilience in Health Systems?
Pharmacovigilance (PV) refers to methods and activities relating to the detection, assessment, understanding, and prevention of ADRs or any other drug-related problem. It provides valuable information for assessing possible safety risks, preventing serious risks, and, ultimately, ensuring patient safety.
Building the capacity in countries and at the regional level to conduct PV, or the thorough monitoring of the safety in the use of medicines, is a critical ‘public good’ investment to ensure that the medicines work correctly and that their health benefits outweigh their risks. PV, as post-marketing medicine surveillance, is crucial to quantify previously recognized adverse reactions, to identify unrecognized adverse events, to evaluate the effectiveness of medicines in real-world situations, and thanks to this knowledge, to decrease mortality and morbidity associated with adverse events. The scope of PV therefore covers product quality, including substandard products; medication errors, including therapeutic ineffectiveness; and previously known or unknown ADRs. The key stakeholders involved in PV are patients, healthcare professionals, governments and pharmaceutical companies.
As discussed below, PV can provide critical support for the performance of different functions in a health system.
Support for national drug policy and regulation
The provision of good quality, safe and effective medicines and their appropriate use is the responsibility of national governments. For all medicines, there is a trade-off between the benefits and the potential for harm. PV plays a specialized and critical role in ensuring ongoing safety of medicinal products. In particular, post marketing safety drug monitoring, includes detection of rare ADRs not detected in clinical trials, drug interactions and other effects caused by the use of medicines in everyday care, assessing the contribution of excipients and preservatives to the safety profile, systems for comparing safety profiles of similar medicines, and measuring the environmental burden of medicines through the surveillance of the effects of drug residues such as antibiotics, psychoactive drugs and hormones on human health and livestock.
A health system that includes pharmacovigilance promotes the safety of medications by minimizing the occurrence of ADRs and provides a warning network of various healthcare providers, regulators, manufacturers and consumers to take remedial actions in a timely and orderly manner, and preventive measures to avoid ADRs in future patients by improving how medicines are prescribed and used. As shown by the experience of OECD countries, regulatory agencies have use evidence derived from routinely collected data to confirm or counter suspected safety concerns, and depending on the results, such evidence has led to market withdrawal, safety notifications and labelling changes, modified indications, or confirmation of the initial terms of marketing authorization.
Support for the delivery of clinical services to patients
PV is also important for clinical practice because it can have a large impact on healthcare quality by generating safety information on the use of medicines to guide the work of health care providers and protect the patients. Evidence from routinely collected data has been used to drive changes in clinical guidelines, for example on medicines used in the management of attention deficit hyperactivity disorder (ADHD) in Australia, given the concern that the practice of given children aged 5 years and younger antipsychotics for behavior modification was inappropriate, on statin use in prevention of cardiac events in patients with ischemic heart disease in Israel, and in the United Kingdom, where evidence on the safety of pertussis vaccine in pregnant women supported the continuation of the vaccination program. Or, conversely, the detection of poor adherence to guidelines for the management of diabetes because some patients showed an increase in ADR in Cameroon.
During the first waves of the pandemic, the absence of medicines and vaccines for treatment and prevention of COVID-19 led to a rush to repurpose drugs already approved for other indications. Medicines such as remdesivir, hydroxychloroquine, ivermectin, and azithromycin were used off-label—that is, the unapproved use of an approved drug—for the treatment of COVID-19 patients, even if the underlying scientific evidence on benefits was low in quality and mostly based on in vitro studies. After some adverse effects from the administration of these drugs were reported among COVID-19 patients, regulatory agencies issued warnings against their use to prevent serious problems among patients. This was the case of azithromycin, an antibiotic that has been widely used for the treatment of COVID-19 patients, but the known proarrhythmogenic activity (producing or tending to produce cardiac arrhythmia) of which may be exacerbated if the drug is used in combination with other drugs proposed for COVID-19 treatment, for example, hydroxychloroquine. Or the case of kidney damage associated with the use of remdesivir reported to the U.S. Food and Drug Administration Adverse Event Reporting System (FAERS).
PV also has the potential to strengthen current antimicrobial stewardship strategies. Antimicrobial overuse, underuse, or use for inappropriate infections pose a particularly difficult challenge to health systems, as they contribute to the increase in antimicrobial resistance, which is associated with a number of clinical complications, unnecessary prolongation of hospital stays, and, in some cases, the deaths of patients. Antimicrobial resistance requires urgent multidisciplinary solutions, and PV databases can serve as sources of data on suspected resistance and inappropriate use. More specifically, PV data can signal issues that may help provide a bigger picture to prescribers who have a choice to make during each consultation.
Support to public health programs for disease control
Likewise, monitoring medicine safety is important to public health programs for disease control at the population level. Many TB programs have introduced and institutionalized active drug safety monitoring and management platforms for drug-resistant TB. In Vietnam, as the management of multidrug-resistant tuberculosis (MDR-TB) is a significant public health challenge due to the complexity and long duration of the MDR-TB treatment, active TB drug-safety monitoring among patients has shown to be useful to understand the safety of MDR-TB treatment and explore the risk factors for toxicity, helping reduce the inconvenience, discomfort, and toxicity of such regimens, and, importantly, increase adherence and likelihood of successful treatment completion. The introduction of novel medicines and regimens for antiretroviral (ARV) treatment for HIV/AIDS also requires comprehensive surveillance systems for adverse events and adverse drug reactions. Because HIV programs must develop active ARV toxicity monitoring systems to ensure safe global scale-up of newer regimens, such as tenofovir-lamivudine-dolutegravir, building on existing drug monitoring infrastructure has been proposed to actively monitor ARV regimens as a synergistic TB/HIV collaborative activity and to narrow active toxicity monitoring gaps. PV has also been of great importance for malaria control programs in view of the increasing resistance to existing antimalarial medicines that led countries to switch to using combinations of various artemisinin derivatives as their first- and second-line treatments for malaria.
Support for selecting and funding effective medicines
The prices and reimbursement conditions of new medicines are often determined just after market entry, based on evidence of risks and benefits generated in pre-registration clinical trials. However, as shown on a OECD report, the use of these medicines in routine clinical practice sometimes leads to the emergence of unanticipated outcomes (rare or delayed adverse effects not detectable in clinical trials; variable clinical results) and very often reveals a gap between efficacy (benefits assessed in clinical trials) and effectiveness (benefits observed in clinical practice). Retrospective observational studies, using appropriate analytical methods, can therefore help in assessing the value of medicines in use in healthcare systems. In countries such as Australia, Estonia, Finland, and France, observational studies based on routinely collected data have influenced decisions on coverage conditions or prices.
Support for building trust in the health system.
PV, by helping ensure that risks in medicine use are anticipated and managed, provide regulators with the necessary information to amend the recommendations on the use of the medicines, improve communication between the health professionals and the public, and educate health professionals to understand the effectiveness or risk of medicines that they prescribe. In particular, effective communication by governments and health care providers, as well as the pharmaceutical industry, on the benefits and risks of medicines, is vital to address disinformation that hampers public health efforts, concerns and fears among the public about adverse effects of medicines, and ultimately, to save lives. As shown by the COVID-19 vaccination experience, the European Medicines Agency (EMA), on the basis of its review of thromboembolic events associated with the administration of the AstraZeneca COVID-19 vaccine, advised national governments in the first part of 2021 to resume vaccination and communicate to the population that the benefits of the AstraZeneca vaccine in preventing COVID-19, with its associated risk of hospitalization and death, outweighed the risks of reported adverse effects. As a result, the rollout of the AstraZeneca vaccine continued as part of vaccination campaigns in Europe and elsewhere to control the pandemic and supported the rapid reactivation of social life and economic activity.
Support for the implementation of Eco-Pharmacovigilance
The occurrence of pharmaceuticals in the natural environment has been frequently reported around the world. As noted in a recent article, as a kind of biologically active compounds specially designed to be effective even at very low concentration levels, pharmaceuticals in the environment could have adverse impacts to the health of human beings or other non-targeted organisms due to long-term exposures. To minimize the pharmaceutical pollution is has been proposed that drawing on the experience of PV, EPV could be developed. More specifically, it is suggested that common methods and activities used in PV including spontaneous reporting, intensive monitoring, database studies, and their potential applicability to the environment, could be adapted and applied to support the future development of EPV.
Looking Ahead
The pharmaceutical landscape is constantly evolving with an increased volume and complexity of new drugs and therapeutics becoming available on the market on an annual basis. This raises concerns regarding drug safety and adverse events. Artificial Intelligence (AI), a branch of computer science that enables computers and other devices to implement tasks that would otherwise require high level of human intelligence and time, stands to transform PV in the future through analyzing and processing large amounts of data that would normally be out of the scope of a human’s lifetime.
The development of advanced methodologies, including machine learning techniques and the availability of large amounts of electronic health care data, augurs well for the growing importance of PV in a health system, as it offers the opportunity for optimizing drug benefit-risk profile evaluation in real world settings. Indeed, systematic clinical data mining can accelerate the speed at which ADE signals can be detected, further contributing to building resilience in health systems. Such an active drug safety surveillance system would allow drugs to be monitored longitudinally over their entire life cycle, providing regulatory authorities with timely access to new information with which to evaluate a drug’s risk profile, enhancing the safety of medicine use and contributing to good health outcomes for the benefit of society as a whole.
More specifically, the automation and machine learning models can enhance PV processes and provide a more efficient way to analyze information relevant to patient safety, including in the prediction of side effects and ADR. The results of a recent systematic review illustrate how machine learning is already being used for the identification of ADEs and ADRs (57.6%), processing safety reports (21.2%), extraction of drug-drug interactions (7.6%), identification of populations at high risk for drug toxicity or guidance for personalized care (7.6%), prediction of side effects (3.0%), simulation of clinical trials (1.5%), and integration of prediction uncertainties into diagnostic classifiers to increase patient safety (1.5%).
In conclusion, it should be clear that in supporting future efforts to strengthen PV as part of building resilient health systems in countries, stakeholders and other actors will be helping to realize a basic tenet associated with the millennia-old Hippocratic oath, “First, do no harm,” which is at the core of public health and medical practice.