- adverse drug reaction
- drug safety
- postmarketing surveillance
Adverse drug reactions (ADRs) were suggested to be the cause of over 100,000 deaths in 1994 in the United States.1 A study conducted in England found that between 1999 and 2008, ADRs were associated with 0.9% of all emergency hospital admissions and 26,399 deaths.2 A comprehensive meta-analysis indicated that ADRs may occur in 16.9% of hospitalized patients.3 Therefore, knowledge, minimization, and prevention of undesirable and harmful effects of medicinal products are important objectives of successful pharmacotherapy.
Pharmacists in both community and institutional settings commonly encounter reports of adverse events. (The distinctions among adverse event, adverse effect, and adverse reaction are described later in this article.) However, the extent of pharmacists’ participation in reporting adverse events is largely unknown. A survey of 377 Texas pharmacists found that 67.9% of them had never reported an adverse event and 65.7% had inadequate knowledge about the reporting process.4 Other studies have suggested a lack of awareness about and experience with adverse-event reporting in both professional pharmacists5 and pharmacy students.6–8
With clinical care being their main focus, pharmacists and other frontline healthcare professionals (HCPs) may not have the training or resources to investigate and report adverse events. The product manufacturers and regulatory authorities, on the other hand, are constrained by a lack of first-hand knowledge of the adverse events and depend on patients and HCPs to report adverse events. While each stakeholder (patient, pharmacist, HCP, manufacturer, and regulatory authority) may have different priorities and constraints regarding the handling of adverse-event information, the ultimate goal of such efforts should be timely recognition of harmful effects of medicinal products. It is therefore important for pharmacists and pharmacy students to understand the basic concepts of pharmacovigilance.
What is pharmacovigilance? Pharmacovigilance, as defined by the World Health Organization, comprises the science and activities relating to the detection, assessment, understanding, and prevention of adverse effects and other drug-related problems.9 Pharmacovigilance can be described as a systematic process encompassing these steps:
Collecting information about the nature, severity, clinical characteristics, and outcomes of adverse effects of medicinal products,
Documenting and analyzing the collected adverse-effects data to detect a causal link between the medicinal product and adverse effect,
Taking remedial actions to eliminate (or minimize) hazards posed by adverse effects of medicinal products, and
Monitoring the impact of any such remedial actions.
Scope of pharmacovigilance. The aim of pharmacovigilance processes is to collect information about various broad aspects of medicinal product safety. These aspects are listed in the guidance document on good pharmacovigilance practices of the Food and Drug Administration (FDA) and include the following10:
New and previously undocumented adverse events, especially if serious,
An increase in the severity of a previously documented adverse event,
Occurrence of serious adverse events thought to be extremely rare in the general population,
New product–drug, product–device, product–food, or product–dietary supplement interactions,
Identification of previously unrecognized at-risk populations,
Confusion about a drug’s name, labeling, packaging, or use,
Concerns arising from the way a drug is used (e.g., adverse events seen at higher-than-approved doses or in populations not recommended for treatment),
Inadequacies of a currently implemented risk-minimization action plan,
Cases of abuse, overdose, and attempted suicide involving medicinal products,
Use of medicinal products for unapproved indications,
Administration of medicinal products by incorrect routes, and
Lack of efficacy reports,
Various medicinal product–related safety issues not attributable to the pharmacologic properties of the product are also handled by pharmacovigilance processes. Upon learning about these issues, the regulatory authorities ask manufacturers to take remedial actions (e.g., product recalls). Some of these safety issues include (1) dosage form complaints (e.g., contamination, physical defects, abnormal odor or taste), (2) product packaging issues (e.g., broken seals, leaking bottles, short fill or count), (3) labeling problems (e.g., missing labels, missing lot numbers, missing expiry dates), and (4) counterfeit medicines.
Need for pharmacovigilance. The approval of any new medicinal product is based on efficacy and safety results obtained in clinical trials. However, various aspects of a medicinal product’s safety may not be known from clinical trials and until the product is used in routine clinical settings for the reasons described below.
Small sample sizes in clinical trials. Generally, the number of patients exposed to any new medicinal product in clinical trials is considerably smaller than the number of patients exposed to the product after commercialization. Therefore, rare adverse events may not be encountered in clinical trials due to the low number of patients exposed.
Importance of naturalistic settings. Clinical trial data are generally obtained in controlled settings where patients are carefully chosen and followed under strict conditions to assess particular efficacy endpoints. Real-life patients may also have more complex comorbidities and concomitant medication regimens. Some adverse effects (e.g., cancer) can only be detected after prolonged exposure to a medicinal product, which may not be feasible in clinical trials of limited duration.
Insufficient safety data in special populations. Clinical trials are generally conducted in certain countries and limited patient populations and may not comprehensively obtain medicinal product safety information in special populations (ethnicity, pregnancy, pediatrics, geriatrics).
Broader aspects of product safety. A medicinal product’s abuse potential and effects of overdose may become known only after product commercialization. Similarly, medication errors may not be recognized until routine clinical use of the product. FDA’s advisories about confusing Brintellix (Takeda’s brand of vortioxetine) with Brilinta (AstraZeneca’s brand of ticagrelor)11 and medication errors involving Avycaz (Allergan’s brand of ceftazidime–avibactam)12 are two such recent examples. With the passage of time, information about interactions between a new medicinal product and older products may become evident. The burgeoning medical device industry also presents possibilities of novel adverse events involving product–device interactions (e.g., heparin-induced thrombocytopenia leading to left ventricular assist device thrombosis).13 With an increasing public tendency toward the use of herbal products and dietary supplements, there is an ever greater need for identifying their potentially harmful effects and interactions.14–16
Improved characterization of adverse effects. The availability of advanced diagnostic tools may facilitate better characterization of adverse effects and predisposed population groups. For example, the presence of the human leukocyte antigen B*1502 variant allele in Han Chinese (and other Asians) has been associated with carbamazepine-induced, serious, cutaneous ADRs.17 In December 2007, FDA recommended genotyping patients of Chinese ancestry for this allele before starting treatment with carbamazepine.18 The timing of this safety communication is remarkable when one considers the fact that carbamazepine was first approved by FDA in 1968.
Concepts and terminology. There is a diversity of terms used to describe the harmful effects of medicinal products. For example, the terms adverse effect, adverse drug reaction, side effect, and others are used interchangeably in clinical practice. There also exists marked variation in the definitions of medication-safety–related terms among various professional bodies and healthcare systems.19 Standardization of medication-safety–related terminology is an important goal of pharmacovigilance processes. Some of the definitions and concepts involved in pharmacovigilance are discussed below.
Adverse drug effects, events, and reactions. Adverse effects and adverse reactions may represent the same underlying phenomenon or derangement. Adverse effects should be viewed from the perspective of the medicinal product. Adverse reactions should be viewed from the perspective of the patient. In other words, medicinal products cause adverse effects; patients experience adverse reactions.
Aronson20 described the important distinction between an adverse effect and an adverse reaction as follows: An adverse (drug) effect is a potentially harmful effect suspected to be resulting from use of a medicinal product. For example, an elevated International Normalized Ratio (INR) and occult blood loss in a patient after starting an oral anticoagulant could be suspected to be adverse effects of the medication. Adverse effects may or may not be associated with a clinically appreciable symptom experienced by the patient (e.g., elevated INR or occult blood loss may not produce any noticeable harm to a patient).
An adverse (drug) reaction is an appreciably harmful or unpleasant reaction (an actual symptom experienced by a patient) suspected to be resulting from use of a medicinal product. For example, severe fatigue in a patient after starting an oral anticoagulant could be suspected to be an adverse drug reaction. One may suspect anemia (due to occult blood loss caused by the medicinal agent) to be the cause of the patient’s fatigue.
Adverse events include all undesirable events occurring after the use of a medicinal product that may not necessarily be ascribed to the product. For example, consider a patient diagnosed with stage III breast cancer 1 week after starting an oral anticoagulant. An association between the oral anticoagulant and advanced breast cancer is not plausible. Therefore, while breast cancer is an adverse event, it is not an adverse effect of the drug.
It is important to note that FDA encourages pharmacists and HCPs to report all serious adverse events. Only after diligent assessment and evaluation may the role of a medicinal product in causing the adverse event become suspect.
Seriousness and severity. The MedWatch program defines a serious adverse event as one that is fatal (or life threatening), leads to hospitalization, causes a significant persistent disability, or results in a congenital anomaly, or is considered serious by the reporter.21 Severity, in comparison with seriousness, is a measure of intensity of an adverse event. For example, an itch can be mild, moderate, or severe. However, considering the above definition of seriousness, an itch is a nonserious adverse event.
Expectedness. The regulatory terminology defines an unexpected adverse event as one the nature or severity of which is not consistent with the available product information (e.g., package insert).22 For example, cough is a well-known adverse reaction to angiotensin-converting enzyme (ACE) inhibitors and is described in their respective package inserts. Therefore, a patient report of cough after starting an ACE inhibitor does not qualify as an unexpected event.
Regulatory authorities are interested in receiving reports of serious and unexpected adverse events on an urgent and priority basis. It must be stressed that pharmacists should use their clinical judgment while assessing adverse events, and the above definitions should not hinder them from reporting events that they believe to be significant and worth reporting.
Causality. When the adverse event is a likely consequence of a known pharmacologic property of the medicinal product, it is plausible to assign causality to the product (e.g., a drug with known anticholinergic properties can plausibly cause dry mouth). In some cases, prior reports of similar adverse events associated with the suspect product may facilitate causality determination. The effects of discontinuation of the suspect product (dechallenge) and reintroduction (rechallenge) may also provide information about the product’s involvement. However, such methods of causality determination may not be applicable for a new medicinal product lacking much information about its pharmacologic or safety profile.
Probability scales (e.g., Naranjo et al.23 adverse drug reaction scale, causality categories of the Uppsala Monitoring Centre24) can objectively ascertain the likelihood of a causal relationship between a medicinal product and an adverse event. A free version of the Naranjo et al. algorithm is available as an online calculator.25
Harmonization of event terms. Descriptions of adverse events can vary greatly depending on the source and reporter. Pharmacists commonly encounter colloquial terms with unclear and ambiguous meaning (e.g., “almost died,” “feeling funny,” “skin on fire,” “zoned out”). Also, different reporters may use different words to report the same adverse event (e.g., headache versus “head was exploding”).
Use of standardized adverse-event terminology can facilitate easy summarization, analysis, and exchange of safety information. FDA and pharmaceutical companies routinely conduct analysis of cumulative safety data. Such an analysis may be difficult if the events were reported using colloquial language. One resource commonly used by pharmaceutical manufacturers to harmonize adverse-event terms is the Medical Dictionary for Regulatory Activities.26
The pharmacovigilance process. The basic components of the pharmacovigilance process are summarized below.
Data generation. The major sources of spontaneously generated adverse-event data are reports from patients, HCPs, and manufacturers. It is increasingly common to see patients reporting adverse events on social media websites. Other sources of adverse-event data include literature articles, pregnancy and disease registries, ongoing clinical trials, and epidemiologic studies.22
Case evaluation and reporting. Product manufacturers typically evaluate each case for seriousness, expectedness, and causality. Manufacturers may also contact the patient (or HCP) to get additional information. After analyzing available information, manufacturers may determine if the case is urgently reportable to the regulatory authorities. Manufacturers are also required to maintain a database of all reports of adverse events regardless of their seriousness, expectedness, or causality.
Adverse events are reported to FDA on a MedWatch form.21 A tutorial to familiarize HCPs with the MedWatch program is available online.27 Medication errors may also be reported to the National Medication Errors Reporting Program of the Institute for Safe Medication Practices.28 When reporting an adverse event, as a general rule of thumb, effort should be made to supply the following 4 minimum pieces of information:
Patient details (age and sex). These details are important in order to have an identifiable patient. Therefore, it is not appropriate to report that “several people” have experienced an adverse reaction after using a medicinal product. The MedWatch program does not require the patient’s name or address. Other information (patient’s medical history, concomitant medications) also aids in the thorough evaluation of a case.
Information about the reporter (or an identifiable source). This information helps the regulatory authorities and manufacturer to contact the reporter for further information.
Information about the suspect product. This includes the product name, dose, route, and dates of administration. The manufacturer’s name, if known, should also be reported.
The description of the adverse event should include the following:
Chronology of the adverse event in relation to use of the medicinal product,
Accurate description of the adverse event (observed signs, symptoms, and associated diagnostic findings),
Results of discontinuation of the suspect product (dechallenge) and reintroduction (rechallenge), if available,
Details of corrective actions taken can help with seriousness assessment; management of serious adverse events may require significant interventions (e.g., surgery, hospitalization), while a nonserious adverse event (e.g., cough) may not require any intervention,
The outcome of the adverse event (i.e., resolved or not resolved), and
A summary narrative to capture case details that cannot be accommodated in the specific fields of the MedWatch form or a colloquial description of the adverse event.
A sample narrative may have the following general outline:
This report is regarding a (age)-year-old (fe)male patient who was taking (suspect medication). Medical history of the patient included (conditions if known). Patient’s concomitant medications included (medications and doses if known). On (date), the patient started taking (suspect medication and dose) for (indication). (Time) after starting (suspect medication), the patient experienced (adverse event including any verbatim description by patient). The patient was (or not) hospitalized. Results of (laboratory tests if performed) showed (results if known). The patient underwent (procedures or interventions if any). The (suspect medication) was (dis)continued. After (time), patient’s (adverse event) resolved (or did not resolve). The patient was (or not) followed up after (time). Follow up evaluation showed (status of adverse event, laboratory tests if any).
Patient privacy. Concerns about the privacy of patients may hinder the reporting of adverse events by pharmacists. It is important to note that the Health Insurance Portability and Accountability Act’s privacy rule permits covered entities (e.g., pharmacists) to disclose protected health information when the intention is to identify threats to public health and safety.29 Only the minimally required protected health information should be disclosed to facilitate reporting and follow-up of adverse event cases.
Aggregate data analysis. The preceding discussion focused on evaluating and reporting individual adverse event cases. Individual reports may indicate some level of association between a medicinal product and an event. However, confirming a causal relationship between a medicinal product and an adverse event requires analysis of the cumulatively collected information on an ongoing basis. FDA maintains a record of received cases into the FDA Adverse Event Reporting System (FAERS). In 2012 and 2013, more than 1 million cases of adverse events were reported to FDA.30 Such a staggering number of reports, along with the varied quality and quantity of information, suggests the need for aggregate data analysis to detect potential safety signals. For an overview of FDA’s handling of adverse event cases and detection of safety signals, the readers are directed to an article by Weaver et al.31
Pharmaceutical manufacturers routinely query their databases to conduct signal detection activities. Quantitative data mining of administrative claims databases and searches of hospital electronic medical records are other methods of identifying medicinal product safety concerns and are discussed elsewhere.32
Dissemination of safety data and remedial actions. FDA publishes new potential safety signals identified from the FAERS on its website.33,34 FDA may also require the product manufacturer to analyze previously collected safety data, make changes to product labeling, provide a medication guide when the product is dispensed, send warning letters to HCPs about safety issues, restrict the prescribing and dispensing of a drug to specialized pharmacies or physicians, require monitoring of patients who use the product, or withdraw the product from the market.
Some of the above preemptive and corrective actions are part of various risk evaluation and mitigation strategies (REMS) programs.35 A well-known REMS program is the iPLEDGE program to prevent pregnancy in female patients receiving treatment with isotretinoin.36 An example of a REMS program in a hospital setting is the Assisting Providers and Cancer Patients with Risk Information for the Safe use of Erythropoiesis-Stimulating Agents (ESA APPRISE) program.37
Pharmacovigilance for nonprescription products. The Dietary Supplement and Nonprescription Drug Consumer Protection Act of 2006 called for mandatory postmarketing safety-reporting requirements for these products.38 The processes for assessing and reporting these adverse events are generally similar to the process described above.
Medical device adverse-event reporting. Adverse-event reports involving medical devices are also reportable to FDA.39 An important and distinct aspect of medical device reporting is that serious adverse events and problems related to the device itself (e.g., device failure and malfunction) are reportable. In addition, FDA is also interested in receiving reports of errors in using the device due to inadequate design, manufacture, or labeling of the device.
Vaccine adverse-event reporting. Analogous to the pharmacovigilance process for drugs and dietary supplements, reporting of vaccine-related adverse events falls under the purview of the Vaccine Adverse Event Reporting System (VAERS) program.40 This program aims to collect and analyze vaccine safety data in order to identify vaccine-related safety issues. HCPs are required to report events listed in the reportable events table.41 Analysis of VAERS data by the Centers for Disease Control and Prevention and FDA may lead to corrective actions including safety communications and vaccine recalls.42
Role of pharmacists. Vigilance is key. Pharmacists’ specialized knowledge of drug therapy can help to detect potential adverse reactions, medication errors, and drug interactions. In certain areas of drug therapy, pharmacists may be the only ones to recognize potential safety concerns. For example, without pharmacists’ initiative, product quality issues may never be identified for review and remedial actions. Selected pharmacovigilance resources are listed in the appendix.
Proactively investigate and report. The requirement for reporting drug-related adverse events in hospitals is stipulated by the Joint Commission’s hospital accreditation manual.43 ASHP has also provided guidelines for ADR monitoring and reporting programs.44 Consequently, many hospitals have implemented systems and processes for the reporting of medicinal product–related adverse events. Community pharmacies, on the other hand, may not have such systems, and the task of reporting adverse events may even be disruptive to their work-flow. Regardless of the presence (or absence) of such systems and processes, the underreporting of adverse events has been well recognized.45,46 Sullivan and Spooner47 suggested a lack of resources as the likely reason for underreporting of ADRs in a teaching hospital. By involving pharmacy students; creating a collaboration among pharmacists, nurses, and physicians; and changing a hospital’s existing ADR data collection form, Sullivan and Spooner reported an increase of more than sevenfold in the number of ADRs documented annually in their hospital. Such a proactive approach may also enable the discovery of institution-specific trends that may be of interest to hospital administrators as well as the pharmacy and therapeutics committee.
It must be emphasized that even a single report can contribute valuable safety information. For example, after receiving only 3 reports of medication errors involving ceftazidime–avibactam, FDA ordered changes in how the product labeling expressed the amounts of drugs in the combination.12 The original product labeling for Avycaz displayed the strength of each component separately (2 g of ceftazidime and 0.5 g of avibactam per vial). FDA noted that other previously approved β-lactam/β-lactamase antibacterial drug products expressed the strength on their labels as the sum of the two active ingredients (e.g., 1.5 g of ampicillin–sulbactam per vial). Therefore, due to pharmacists’ familiarity with this convention for expressing the product strength, there existed a potential for medication errors involving Avycaz. The revised label for Avycaz expresses the strength as the sum of the 2 active ingredients (i.e., 2.5 g of ceftazidime–avibactam per vial). Another tenet that needs to be ingrained is that an absolute confirmation of causality, which may be difficult to prove, is not required to report an adverse event. In situations of time constraints, pharmacists, at the very minimum, should bring the adverse event or safety issue to the attention of the product manufacturer. Pharmaceutical companies employ staff who are well trained in pharmacovigilance and are better equipped to investigate and report safety issues.
Act as a resource for other HCPs. A lack of institutionwide policies and infrastructure to detect, investigate, and report adverse events may place pharmacists in the unique and central position as expert resources and educators for other HCPs who may not be fully aware of the importance of and procedures for reporting adverse events. Pharmacists can help other HCPs with correctly completing and submitting a MedWatch report. The importance of reporting complete information is emphasized by the following statistic: one third of drug quality–problem reports received by FDA in 2008 were incomplete, precluding adequate follow-up by the agency.48 Pharmacists may also act as disseminators of medicinal product safety updates to other HCPs.
Provide patient education. Pharmacists should place a special emphasis on educating patients about avoiding, recognizing, and managing ADRs, therapeutic duplications, and unnecessary medication use. By obtaining a comprehensive medication history, pharmacists may be able to pinpoint the potential causes of adverse events. Without such efforts, the pharmacist may recommend another nonprescription medication, which adds to the patient’s medication burden and may lead to further adverse reactions and drug interactions. The pharmacist should also update a patient’s medical records to include any adverse reactions experienced by the patient.
Summary. Pharmacovigilance is a continuous process requiring active participation of patients, pharmacists, other HCPs, medicinal product manufacturers, and regulatory authorities. Educating pharmacy students and professional pharmacists about pharmacovigilance on an ongoing basis is likely to improve their participation in this process. The onus to integrate the detection and reporting of adverse events into routine clinical practice and education is also on pharmacy businesses, pharmacy colleges, and hospitals and health systems.
Karyn Sullivan, Pharm.D., M.P.H., is gratefully acknowledged for reviewing this manuscript.
Appendix Pharmacovigilance resources for pharmacists
Educational and information resources
Council for International Organizations of Medical Sciences. Current challenges in pharmacovigilance: pragmatic approaches (www.cioms.ch/index.php/publications/print-ablev3)
MedWatch tutorial (www.accessdata.fda.gov/scripts/medwatchlearn)
Medicinal products requiring risk evaluation and mitigation strategies (www.accessdata.fda.gov/scripts/cder/rems)
Medicinal products requiring medication guides (www.fda.gov/drugs/drugsafety/ucm085729.htm)
Adverse-event reporting portals
Medicinal product safety communications
Homepage for MedWatch program (with links to current safety findings related to drugs, dietary supplements, and medical devices) (www.fda.gov/safety/medwatch)
ISMP newsletters (www.ismp.org/newsletters)
ISMP QuarterWatch (analysis of MedWatch reports submitted to the Food and Drug Administration) (www.ismp.org/quarterwatch)
The author has declared no potential conflicts of interest.
- Copyright © 2017 by the American Society of Health-System Pharmacists, Inc. All rights reserved.