Pharmacovigilance: A Comprehensive Guide (Part I)


Pharmacovigilance (PV or PhV), also known as drug safety, is the pharmacological science relating to the collection, detection, assessment, monitoring, and prevention of adverse effects with pharmaceutical products.

The etymological roots for the word “pharmacovigilance” are: pharmakon (Greek for drug) and vigilare (Latin for to keep watch). As such, pharmacovigilance heavily focuses on adverse drug reactions, or ADRs, which are defined as any response to a drug which is noxious and unintended, including lack of efficacy (the condition that this definition only applies with the doses normally used for the prophylaxis, diagnosis or therapy of disease, or for the modification of physiological disorder function was excluded with the latest amendment of the applicable legislation).

What is Pharmacovigilance?

Pharmacovigilance has been defined by the World Health Organisation (WHO) as “The science and activities relating to the detection, assessment, understanding and prevention of adverse effects or any other drug-related problem”.

WHO established its Programme for International Drug Monitoring in response to the thalidomide disaster detected in 1961. Together with the WHO Collaborating Centre for International Drug Monitoring, Uppsala, WHO promotes PV at the country level. At the end of 2010, 134 countries were part of the WHO PV Programme. The aims of PV are to enhance patient care and patient safety in relation to the use of medicines; and to support public health programmes by providing reliable, balanced information for the effective assessment of the risk-benefit profile of medicines.

Terms Commonly Used In Pharmacovigilance

Pharmacovigilance has its own unique terminology that is important to understand. Below are the most common terms used in Pharmacovigilance, although some are used by other disciplines within the pharmaceutical sciences as well.

  • Adverse drug reaction is a side effect (non intended reaction to the drug) occurring with a drug where a positive (direct) causal relationship between the event and the drug is thought or has been proven, to exist.
  • Adverse event (AE) is a side effect occurring with a drug. By definition, the causal relationship between the AE and the drug is unknown.
  • Benefits are commonly expressed as the proven therapeutic good of a product but should also include the patient’s subjective assessment of its effects.
  • Causal relationship is said to exist when a drug is thought to have caused or contributed to the occurrence of an adverse drug reaction.
  • Clinical trial(or study) refers to an organised program to determine the safety and/or efficacy of a drug (or drugs) in patients. The design of a clinical trial will depend on the drug and the phase of its development.
  • Control group is a group (or cohort) of individual patients that is used as a standard of comparison within a clinical trial. The control group may be taking a placebo (where no active drug is given) or where a different active drug is given as a comparator.
  • Dechallenge and rechallenge refer to a drug being stopped and restarted in a patient, respectively. A positive dechallenge has occurred, for example, when an adverse event abates or resolves completely following the drug’s discontinuation. A positive rechallenge has occurred when the adverse event re-occurs after the drug is restarted. Dechallenge and rechallenge play an important role in determining whether a causal relationship between an event and a drug exists.
  • Effectiveness is the extent to which a drug works under real world circumstances, i.e., clinical practice.
  • Efficacy is the extent to which a drug works under ideal circumstances, i.e., in clinical trials.
  • Event refers to an adverse event (AE).
  • Harm is the nature and extent of the actual damage that could be or has been caused.
  • Implied causality refers to spontaneously reported AE cases where the causality is always presumed to be positive unless the reporter states otherwise.
  • Individual Case Safety Report (ICSR) is an adverse event report for an individual patient.
  • Life-threatening refers to an adverse event that places a patient at the immediate risk of death.
  • Phase refers to the four phases of clinical research and development: I – small safety trials early on in a drug’s development; II – medium-sized trials for both safety and efficacy; III – large trials, which includes key (or so-called “pivotal”) trials; IV – large, post-marketing trials, typically for safety reasons. There are also intermediate phases designated by an “a” or “b”, e.g. Phase IIb.
  • Risk is the probability of harm being caused, usually expressed as a percent or ratio of the treated population.
  • Risk factor is an attribute of a patient that may predispose, or increase the risk, of that patient developing an event that may or may not be drug-related. For instance, obesity is considered a risk factor for a number of different diseases and, potentially, ADRs. Others would be high blood pressure, diabetes, possessing a specific mutated gene, for example, mutations in the BRCA1 and BRCA2 genes increase propensity to develop breast cancer.
  • Signal is a new safety finding within safety data that requires further investigation. There are three categories of signals: confirmed signals where the data indicate that there is a causal relationship between the drug and the AE; refuted (or false) signals where after investigation the data indicate that no causal relationship exists; and unconfirmed signals which require further investigation (more data) such as the conducting of a post-marketing trial to study the issue.
  • Temporal relationship is said to exist when an adverse event occurs when a patient is taking a given drug. Although a temporal relationship is absolutely necessary in order to establish a causal relationship between the drug and the AE, a temporal relationship does not necessarily in and of itself prove that the event was caused by the drug.
  • Triage refers to the process of placing a potential adverse event report into one of three categories: 1) non-serious case; 2) serious case; or 3) no case (minimum criteria for an AE case are not fulfilled).

For Pharmacovigilance Basics – Read This:

Adverse event reporting

The activity that is most commonly associated with pharmacovigilance (PV), and which consumes a significant amount of resources for drug regulatory authorities (or similar government agencies) and drug safety departments in pharmaceutical companies, is that of adverse event reporting. Adverse event (AE) reporting involves the receipt, triage, data entering, assessment, distribution, reporting (if appropriate), and archiving of AE data and documentation.

The source of AE reports may include: spontaneous reports from healthcare professionals or patients (or other intermediaries); solicited reports from patient support programs; reports from clinical or post-marketing studies; reports from literature sources; reports from the media (including social media and websites); and reports reported to drug regulatory authorities themselves. For pharmaceutical companies, AE reporting is a regulatory requirement in most countries. AE reporting also provides data to these companies and drug regulatory authorities that play a key role in assessing the risk-benefit profile of a given drug. The following are several facets of AE reporting:

Individual Case Safety Report (ICSR)

One of the fundamental principles of adverse event reporting is the determination of what constitutes an Individual Case Safety Report (ICSR). During the triage phase of a potential adverse event report, it is important to determine if the “four elements” of a valid ICSR are present: (1) an identifiable patient, (2) an identifiable reporter, (3) a suspect drug, and (4) an adverse event.

If one or more of these four elements is missing, the case is not a valid ICSR. Although there are no exceptions to this rule there may be circumstances that may require a judgment call. For example, the term “identifiable” may not always be clear-cut. If a physician reports that he/she has a patient X taking drug Y who experienced Z (an AE), but refuses to provide any specifics about patient X, the report is still a valid case even though the patient is not specifically identified. This is because the reporter has first-hand information about the patient and is identifiable (i.e. a real person) to the physician. Identifiability is important so as not only to prevent duplicate reporting of the same case, but also to permit follow-up for additional information.

The concept of identifiability also applies to the other three elements. Although uncommon, it is not unheard of for fictitious adverse event “cases” to be reported to a company by an anonymous individual (or on behalf of an anonymous patient, disgruntled employee, or former employee) trying to damage the company’s reputation or a company’s product. In these and all other situations, the source of the report should be ascertained (if possible). But anonymous reporting is also important, as whistle blower protection is not granted in all countries. In general, the drug must also be specifically named. Note that in different countries and regions of the world, drugs are sold under various tradenames. In addition, there are a large number of generics which may be mistaken for the trade product. Finally, there is the problem of counterfeit drugs producing adverse events. If at all possible, it is best to try to obtain the sample which induced the adverse event, and send it to either the EMA, FDA or other government agency responsible for investigating AE reports.

If a reporter can’t recall the name of the drug they were taking when they experienced an adverse event, this would not be a valid case. This concept also applies to adverse events. If a patient states that they experienced “symptoms”, but cannot be more specific, such a report might technically be considered valid, but will be of very limited value to the pharmacovigilance department of the company or to drug regulatory authorities.

Coding of adverse events

Adverse event coding is the process by which information from an AE reporter, called the “verbatim”, is coded using standardized terminology from a medical coding dictionary, such as MedDRA (the most commonly used medical coding dictionary). The purpose of medical coding is to convert adverse event information into terminology that can be readily identified and analyzed. For instance, Patient 1 may report that they had experienced “a very bad headache that felt like their head was being hit by a hammer” [Verbatim 1] when taking Drug X. Or, Patient 2 may report that they had experienced a “slight, throbbing headache that occurred daily at about two in the afternoon” [Verbatim 2] while taking Drug Y. Neither Verbatim 1 nor Verbatim 2 will exactly match a code in the MedDRA coding dictionary. However, both quotes describe different manifestations of a headache. As a result, in this example both quotes would be coded as PT Headache (PT = Preferred Term in MedDRA).

Seriousness determination

There are a set of criteria within pharmacovigilance that are used to distinguish a serious adverse event from a non-serious one. An adverse event is considered serious if it meets one or more of the following criteria:

  1. Results in death, or is life-threatening;
  2. Requires inpatient hospitalization or prolongation of existing hospitalization;
  3. Results in persistent or significant disability or incapacity;
  4. Results in a congenital anomaly (birth defect); or
  5. Is otherwise “medically significant” (i.e., that it does not meet preceding criteria, but is considered serious because treatment/intervention would be required to prevent one of the preceding criteria.)

Aside from death, each of these categories is subject to some interpretation. Life-threatening, as it used in the drug safety world, specifically refers to an adverse event that places the patient at an immediate risk of death, such as cardiac or respiratory arrest. By this definition, events such as myocardial infarction (MI), which would be hypothetically life-threatening, would not be considered life-threatening unless the patient went into cardiac arrest following the MI. Defining what constitutes hospitalization can be problematic as well. Although typically straightforward, it’s possible for a hospitalization to occur even if the events being treated are not serious. By the same token, serious events may be treated without hospitalization, such as the treatment of anaphylaxis may be successfully performed with epinephrine. Significant disability and incapacity, as a concept, is also subject to debate. While permanent disability following a stroke would no doubt be serious, would “complete blindness for 30 seconds” be considered “significant disability”? For birth defects, the seriousness of the event is usually not in dispute so much as the attribution of the event to the drug. Finally, “medically significant events” is a category that includes events that may be always serious, or sometimes serious, but will not fulfill any of the other criteria. Events such as cancer might always be considered serious, whereas liver disease, depending on its CTCAE (Common Terminology Criteria for Adverse Events) grade—Grades 1 or 2 are generally considered non-serious and Grades 3-5 serious—may be considered non-serious.

Expedited reporting

Expedited reporting refers to ICSRs (individual case safety reports) that involve a serious and unlisted event (an event not described in the drug’s labeling) that is considered related to the use of the drug. (Spontaneous reports are typically considered to have a positive causality, whereas a clinical trial case will typically be assessed for causality by the clinical trial investigator and/or the license holder.) In most countries, the timeframe for reporting expedited cases is 7/15 calendar days from the time a drug company receives notification (referred to as “Day 0”) of such a case. Within clinical trials such a case is referred to as a SUSAR (a Suspected Unexpected Serious Adverse Reaction). If the SUSAR involves an event that is life-threatening or fatal, it may be subject to a 7-day “clock”. Cases that do not involve a serious, unlisted event may be subject to non-expedited or periodic reporting.

Clinical trial reporting

Also known as SAE (serious adverse event) reporting from clinical trials, safety information from clinical studies is used to establish a drug’s safety profile in humans and is a key component that drug regulatory authorities consider in the decision-making as to whether to grant or deny market authorization (market approval) for a drug. SAE reporting occurs as a result of study patients (subjects) who experience serious adverse events during the conducting of clinical trials. (Non-serious adverse events are also captured separately.) SAE information, which may also include relevant information from the patient’s medical background, are reviewed and assessed for causality by the study investigator. This information is forwarded to a sponsoring entity (typically a pharmaceutical company) that is responsible for the reporting of this information, as appropriate, to drug regulatory authorities.

Spontaneous reporting

Spontaneous reports are termed spontaneous as they take place during the clinician’s normal diagnostic appraisal of a patient, when the clinician is drawing the conclusion that the drug may be implicated in the causality of the event. Spontaneous reporting system relies on vigilant physicians and other healthcare professionals who not only generate a suspicion of an ADR, but also report it. It is an important source of regulatory actions such as taking a drug off the market or a label change due to safety problems. Spontaneous reporting is the core data-generating system of international pharmacovigilance, relying on healthcare professionals (and in some countries consumers) to identify and report any adverse events to their national pharmacovigilance center, health authority (such as EMA or FDA), or to the drug manufacturer itself. Spontaneous reports are, by definition, submitted voluntarily although under certain circumstances these reports may be encouraged, or “stimulated”, by media reports or articles published in medical or scientific publications, or by product lawsuits. In many parts of the world adverse event reports are submitted electronically using a defined message standard.

One of the major weaknesses of spontaneous reporting is that of under-reporting, where, unlike in clinical trials, less than 100% of those adverse events occurring are reported. Further complicating the assessment of adverse events, AE reporting behavior varies greatly between countries and in relation to the seriousness of the events, but in general probably less than 10% (some studies suggest less than 5%) of all adverse events that occur are actually reported. The rule-of-thumb is that on a scale of 0 to 10, with 0 being least likely to be reported and 10 being the most likely to be reported, an uncomplicated non-serious event such as a mild headache will be closer to a “0” on this scale, whereas a life-threatening or fatal event will be closer to a “10” in terms of its likelihood of being reported. In view of this, medical personnel may not always see AE reporting as a priority, especially if the symptoms are not serious. And even if the symptoms are serious, the symptoms may not be recognized as a possible side effect of a particular drug or combination thereof. In addition, medical personnel may not feel compelled to report events that are viewed as expected. This is why reports from patients themselves are of high value. The confirmation of these events by a healthcare professional is typically considered to increase the value of these reports. Hence it is important not only for the patient to report the AE to his health care provider (who may neglect to report the AE), but also report the AE to both the biopharmaceutical company and the FDA, EMA, … This is especially important when one has obtained one’s pharmaceutical from a compounding pharmacy.

As such, spontaneous reports are a crucial element in the worldwide enterprise of pharmacovigilance and form the core of the World Health Organization Database, which includes around 4.6 million reports (January 2009), growing annually by about 250,000.

Aggregate reporting

Aggregate reporting, also known as periodic reporting, plays a key role in the safety assessment of drugs. Aggregate reporting involves the compilation of safety data for a drug over a prolonged period of time (months or years), as opposed to single-case reporting which, by definition, involves only individual AE reports. The advantage of aggregate reporting is that it provides a broader view of the safety profile of a drug. Worldwide, the most important aggregate report is the Periodic Safety Update Report (PSUR) and Development Safety Update Report (DSUR). This is a document that is submitted to drug regulatory agencies in Europe, the US and Japan (ICH countries), as well as other countries around the world. The PSUR was updated in 2012 and is now referred to in many countries as the Periodic Benefit Risk Evaluation report (PBRER). As the title suggests, the PBRER’s focus is on the benefit-risk profile of the drug, which includes a review of relevant safety data compiled for a drug product since its development.

Other reporting methods

Some countries legally oblige spontaneous reporting by physicians. In most countries, manufacturers are required to submit, through its Qualified Person for Pharmacovigilance (QPPV), all of the reports they receive from healthcare providers to the national authority. Others have intensive, focused programmes concentrating on new drugs, or on controversial drugs, or on the prescribing habits of groups of doctors, or involving pharmacists in reporting. All of these generate potentially useful information. Such intensive schemes, however, tend to be the exception.

Risk management

Risk management is the discipline within pharmacovigilance that is responsible for signal detection and the monitoring of the risk-benefit profile of drugs. Other key activities within the area of risk management are that of the compilation of risk management plans (RMPs) and aggregate reports such as the Periodic Safety Update Report (PSUR), Periodic Benefit-Risk Evaluation Report (PBRER), and the Development Safety Update Report (DSUR).

Causality assessment

One of the most important, and challenging, problems in pharmacovigilance is that of the determination of causality. Causality refers to the relationship of a given adverse event to a specific drug. Causality determination (or assessment) is often difficult because of the lack of clear-cut or reliable data. While one may assume that a positive temporal relationship might “prove” a positive causal relationship, this is not always the case. Indeed, a “bee sting” AE—where the AE can clearly be attributed to a specific cause—is by far the exception rather than the rule. This is due to the complexity of human physiology as well as that of disease and illnesses. By this reckoning, in order to determine causality between an adverse event and a drug, one must first exclude the possibility that there were other possible causes or contributing factors. If the patient is on a number of medications, it may be the combination of these drugs which causes the AE, and not any one individually. There have been a number of recent high-profile cases where the AE led to the death of an individual. The individual(s) were not overdosed with any one of the many medications they were taking, but the combination there appeared to cause the AE. Hence it is important to include in your/one’s AE report, not only the drug being reported, but also all other drugs the patient was also taking.

For instance, if a patient were to start Drug X and then three days later were to develop an AE, one might be tempted to attribute blame Drug X. However, before that can be done, the patient’s medical history would need to be reviewed to look for possible risk factors for the AE. In other words, did the AE occur with the drug or because of the drug? This is because a patient on any drug may develop or be diagnosed with a condition that could not have possibly been caused by the drug. This is especially true for diseases, such as cancer, which develop over an extended period of time, being diagnosed in a patient who has been taken a drug for a relatively short period of time. On the other hand, certain adverse events, such as blood clots (thrombosis), can occur with certain drugs with only short-term exposure. Nevertheless, the determination of risk factors is an important step of confirming or ruling-out a causal relationship between an event and a drug.

Often the only way to confirm the existence of a causal relationship of an event to a drug is to conduct an observational study where the incidence of the event in a patient population taking the drug is compared to a control group. This may be necessary to determine if the background incidence of an event is less than that found in a group taking a drug. If the incidence of an event is statistically significantly higher in the “active” group versus the placebo group (or other control group), it is possible that a causal relationship may exist to a drug, unless other confounding factors may exist.

Signal detection

Signal detection (SD) involves a range of techniques (CIOMS VIII). The WHO defines a safety signal as: “Reported information on a possible causal relationship between an adverse event and a drug, the relationship being unknown or incompletely documented previously”. Usually more than a single report is required to generate a signal, depending upon the event and quality of the information available.

Data mining pharmacovigilance databases is one approach that has become increasingly popular with the availability of extensive data sources and inexpensive computing resources. The data sources (databases) may be owned by a pharmaceutical company, a drug regulatory authority, or a large healthcare provider. Individual Case Safety Reports (ICSRs) in these databases are retrieved and converted into structured format, and statistical methods (usually a mathematical algorithm) are applied to calculate statistical measures of association. If the statistical measure crosses an arbitrarily set threshold, a signal is declared for a given drug associated with a given adverse event. All signals deemed worthy of investigation, require further analysis using all available data in an attempt to confirm or refute the signal. If the analysis is inconclusive, additional data may be needed such as a post-marketing observational trial.

SD is an essential part of drug use and safety surveillance. Ideally, the goal of SD is to identify ADRs that were previously considered unexpected and to be able to provide guidance in the product’s labeling as to how to minimize the risk of using the drug in a given patient population.

Risk management plans

A risk management plan (RMP) is a documented plan that describes the risks (adverse drug reactions and potential adverse reactions) associated with the use of a drug and how they are being handled (warning on drug label or on packet inserts of possible side effects which if observed should cause the patient to inform/see his physician and/or pharmacist and/or the manufacturer of the drug and/or the FDA, EMA)). The overall goal of an RMP is to assure a positive risk-benefit profile once the drug is (has been) marketed. The document is required to be submitted, in a specified format, with all new market authorization requests within the European Union (EU). Although not necessarily required, RMPs may also be submitted in countries outside the EU. The risks described in an RMP fall into one of three categories: identified risks, potential risks, and unknown risks. Also described within an RMP are the measures that the Market Authorization Holder, usually a pharmaceutical company, will undertake to minimize the risks associated with the use of the drug. These measures are usually focused on the product’s labeling and healthcare professionals. Indeed, the risks that are documented in a pre-authorization RMP will inevitably become part of the product’s post-marketing labeling. Since a drug, once authorized, may be used in ways not originally studied in clinical trials, this potential “off-label use”, and its associated risks, is also described within the RMP. RMPs can be very lengthy documents, running in some cases hundreds of pages and, in rare instances, up to a thousand pages long.

In the US, under certain circumstances, the FDA may require a company to submit a document called a Risk Evaluation and Mitigation Strategies (REMS) for a drug that has a specific risk that FDA believes requires mitigation. While not as comprehensive as an RMP, a REMS can require a sponsor to perform certain activities or to follow a protocol, referred to as Elements to Assure Safe Use (ETASU), to assure that a positive risk-benefit profile for the drug is maintained for the circumstances under which the product is marketed.

Risk/benefit profile of drug

Pharmaceutical companies are required by law in most countries to perform clinical trials, testing new drugs on people before they are made generally available. This occurs after a drug has been pre-screened for toxicity, sometimes using animals for testing. The manufacturers or their agents usually select a representative sample of patients for whom the drug is designed – at most a few thousand – along with a comparable control group. The control group may receive a placebo and/or another drug, often a so-called “gold standard” that is “best” drug marketed for the disease.

The purpose of clinical trials is to determine:

  • if a drug works and how well it works
  • if it has any harmful effects, and
  • if it does more good than harm, and how much more? If it has a potential for harm, how probable and how serious is the harm?

Clinical trials do, in general, tell a good deal about how well a drug works. They provide information that should be reliable for larger populations with the same characteristics as the trial group – age, gender, state of health, ethnic origin, and so on though target clinical populations are typically very different from trial populations with respect to such characteristics.

The variables in a clinical trial are specified and controlled, but a clinical trial can never tell you the whole story of the effects of a drug in all situations. In fact, nothing could tell you the whole story, but a clinical trial must tell you enough; “enough” being determined by legislation and by contemporary judgements about the acceptable balance of benefit and harm. Ultimately, when a drug is marketed it may be used in patient populations that were not studied during clinical trials (children, the elderly, pregnant women, patients with co-morbidities not found in the clinical trial population, etc.) and a different set of warnings, precautions or contraindications (where the drug should not be used at all) for the product’s labeling may be necessary in order to maintain a positive risk/benefit profile in all known populations using the drug.


Pharmacoepidemiology is the study of the incidence of adverse drug reactions in patient populations using drug agents.

Pharmacogenetics and pharmacogenomics

Although often used interchangeably, there are subtle differences between the two disciplines. Pharmacogenetics is generally regarded as the study or clinical testing of genetic variation that gives rise to differing responses to drugs, including adverse drug reactions. It is hoped that pharmacogenetics will eventually provide information as to which genetic profiles in patients will place those patients at greatest risk, or provide the greatest benefit, for using a particular drug or drugs. Pharmacogenomics, on the other hand, is the broader application of genomic technologies to new drug discovery and further characterization of older drugs.

To be continued…