- Jeffrey K Aronson
Centre for Evidence Based Medicine, Nuffield Department of Primary Care Health Sciences, University of Oxford, Oxford, UK
- Twitter @JKAronson
Defining pharmacovigilance
The earliest use of the term “pharmacovigilance” of which I am aware comes from a French document, published by the World Health Organization in 1969, reporting the proceedings of a symposium that had been held in 1968.1 That the term was relatively new is suggested by the context in which it appears. The relevant paragraph reads:
“Parmi les réactions adverses aux médicaments qui sont peu fréquentes chez l’homme, il en est beaucoup qui sont impossibles à prévoir d’aprés les résultats de 1’expérimentation toxicologique sur l’animal ou de l’essai clinique contrôlé si les populations exposées diffèrent par l’âge, le sexe, l’état morbide, etc. de celles qui ont fait l’objet de l’essai clinique ou si les spécifications des médicaments ont été modifiées postérieurement à cet essai. Ainsi est née l’idée de la surveillance des médicaments du point de vue des réactions adverses (pharmacovigilance) qui a pour but principal de déceler aussi rapidement que possible les médicaments qui peuvent produire des effets indésirables.”
I translate this as follows:
“Many uncommon adverse reactions to medicinal products are impossible to predict from toxicological experiments in animals or to detect in controlled clinical trials if the exposed populations in a trial differed by age, sex, morbidities, etc. from others in the trial, or if the specifications of the medicinal products were changed after the trial. Thus was born the idea of drug surveillance for adverse reactions (pharmacovigilance), whose main purpose is to detect adverse drug reactions as soon as possible.”
Two forms of pharmacovigilance were identified in the report. Here is the original text in which they were mentioned:
“On entend par pharmacovigilance la notification, l’enregistrement et l’évaluation systématique des réactions adverses aux médicaments délivrés avec ou sans ordonnance. Les renseignements sur ces réactions peuvent étre obtenus, soit par des notifications volontaires de médecins praticiens et d’hôpitaux à des centres préalablement désignés (pharmacovigilance spontanée), soit par l’application de techniques épidémiologiques permettant de recueillir systématiquement des informations à certaines sources: hópitaux, échantillons représentatifs du corps médical, etc. (pharmacovigilance intensive).”
…and here is my translation:
“Pharmacovigilance is the notification, recording and systematic evaluation of adverse reactions to medicinal products supplied with or without prescription. Information on these reactions may be obtained either by voluntary notification from medical practitioners and hospitals to pre-designated centres (spontaneous pharmacovigilance), or by applying epidemiological techniques to systematically collect information from certain sources: hospitals, representative samples of the medical profession, etc. (intensive pharmacovigilance).”
This description makes it clear that the two types of pharmacovigilance mentioned here, “pharmacovigilance spontanée” and “pharmacovigilance intensive,” are what might today be called “passive” and “active” pharmacovigilance respectively.
I have previously listed the main activities involved in pharmacovigilance and have derived from them an extensional definition.2 Extensional definitions are so called because they incorporate an extended list of the main components or features of the definiendum, the thing to be defined. Such definitions may be represented in diagrams, showing the processes in relation to each other, or they may be couched verbally, in which case, if many processes are involved, the definitions tend to be rather long, as my definition of pharmacovigilance was, albeit highly explicit.
However, it is then sometimes possible to craft an acceptable shorter, intensional, definition from the longer extensional one, with the most important features stated explicitly, but retaining expressions that imply the other features. I have now done that with my definition of pharmacovigilance, and here is the result:
pharmacovigilance, n. A scientific activity related to the benefits and harms of any intended therapeutic intervention, primarily involving accurate identification and quantification, mechanistic analysis, continued surveillance, communication, and prevention of its adverse effects, adverse reactions to it, and adverse interactions with other interventions [ancient Greek ϕαρμακο-, combining form, from ϕάρμακον drug, medicine, poison, and Latin vigilare to be watchful or alert, via French vigilance.]
This reduces the definition from 149 words to 41.
Defining terms
The longer definition I proposed begins by describing pharmacovigilance as a surveillance activity and the shorter includes continued surveillance as an important aspect.
Ideally, a definition should not include words that need further definition. However, no definition truly fulfils that requirement, since it is always possible that a reader, for example one whose first language is not, in this case, English, will not understand one or more words used in the definition. The best one can do is to use terms that are sufficiently familiar to be generally understood without further explanation. Definitions of technical terms are more likely to include terms that fail this desideratum. Furthermore, even if the general sense of a technical term is understood, needing no further explanation, it will always be necessary to provide a formal definition of it. Even words such as “and” and “the” are defined in general dictionaries of the English language.
In the case of the word “surveillance” one would expect general readers to understand what is implied by its use in the definition of pharmacovigilance, without needing further explanation. Nevertheless, the term needs its own definition.
Surveillance activities in pharmacovigilance
“Surveillance” comes from Old French, sur (over)+veiller (to watch). The original Indo-European root of veiller is WEG, from which the Latin words vigor and vigilare (to watch or stay awake) derive, with their connotations of watchfulness and liveliness.3 And “veiller” comes from “vigilare.” Thus, surveillance has the same origins as vigilance and implies watchfulness and speedy action.
The concept of surveillance in pharmacovigilance and pharmacoepidemiology evolved from the concept of surveillance in epidemiology. Here are its essential features as a part of pharmacovigilance, as previously described3:
● its purposes are to identify, evaluate, understand, and communicate previously unknown effects of healthcare products, or new aspects of known effects, in order that they may be harnessed (if beneficial) or prevented or mitigated (if harmful);
● it is not itself interventional, although it can adduce evidence from all kinds of studies, including interventional ones;
● it involves collection, management, analysis, and interpretation of data from any relevant source;
● it involves actions to be taken after signal detection, including evaluation and communication;
● it should contribute to the classification of adverse reactions and their prevention or mitigation and/or to the harnessing of beneficial effects.
Active and passive surveillance
The idea of active surveillance arose in studies of the epidemiology of tuberculosis in the 1890s.45 However, the term “surveillance” was not used in pharmacovigilance until the mid-1960s,67 and the term “active surveillance” was introduced soon after. In a 1969 editorial in the British Medical Journal it was contrasted with spontaneous reporting of adverse drug reactions, but was not defined.8
Papers on active surveillance of toxic shock syndrome in relation to tampons9 or barrier contraceptives10111213 appeared in the 1980s. Active surveillance for drug effects was mentioned in a review of the beneficial effects of antiepileptic drugs in 1986,14 and active surveillance for adverse effects in a Bulgarian paper in 1987.15
The first occasion on which the term “active surveillance” was used in relation to a specific therapeutic drug was in 1994, in a description of adverse events in older patients taking digitalis.16 It was later used to describe active surveillance for adverse reactions to vaccines,17 appetite suppressants,18 missed doses of medications,19 general adverse drug reactions in hospital,20 and the safety of cyproterone acetate.21
The term “passive surveillance” was first used in the context of drug therapy or vaccines in the 1980s,22 in most cases in relation to infectious diseases; in about 18% it described surveillance of the adverse effects of vaccines.
Postmarketing surveillance
Although pharmacovigilance is an activity that can be carried out at any time during drug development, the term “surveillance” is most often used in connection with the postmarketing phase. The term “postmarketing surveillance,” whose activities are also known as phase 4 studies, first appeared in the 1960s to describe analysis of data accumulated typically for the purpose of surveillance for adverse effects after a company has been given a marketing authorisation for a product.2324 It generally encompasses the processes of signal detection and signal evaluation.
Defining surveillance in pharmacovigilance
The different qualifiers that have been appended to the term “surveillance” have been defined in many different ways and in relation to many different circumstances. Indeed, the cases to which the term has been applied have been so diverse that other terms have occasionally been suggested to differentiate them. Apart from “active” and “passive” surveillance, these include terms such as “stimulated surveillance,” “enhanced passive surveillance,”2526 and “solicited reactions.”27
When Manfred Hauben, Andrew Bate, and I essayed a definition of “surveillance” in 2012,3 we suggested that it would be less confusing to use a single definition of “surveillance” and, instead of adding qualifiers, the precise surveillance method used should be detailed in any description of a specific system of surveillance.
In proposing a definition we took into account the following features:
● surveillance should be distinguished from monitoring, in that it involves populations while monitoring involves individuals;
● surveillance can be performed at any time during the lifetime of a healthcare product (a medicinal product or device) and is not restricted to a single period of time;
● although itself non-interventional, it can adduce any types of evidence (interventional, observational, or anecdotal, potentially at different times);
● it encompasses data collection, management, analysis, and interpretation (including the evaluation of signals, although not all evaluation studies are necessarily part of surveillance);
● it includes the planning of actions to be taken after signal detection, including initial evaluation of signals and communication;
● it should contribute to the classification of adverse reactions and their prevention or mitigation and/or to the harnessing of beneficial effects.
We also proposed that there was no need to take into account the supposedly passive and active elements that are involved.
This is the definition that we proposed:
surveillance, n. a form of non-interventional public health research, consisting of a set of processes for the continued systematic collection, compilation, interrogation, analysis, and interpretation of data on benefits and harms (including relevant spontaneous reports, electronic medical records, and experimental data) [French sur- over+veiller, from Latin vigilare, to watch]
As a codicil, we noted that the purposes of surveillance are to identify, evaluate, understand, and communicate previously unknown effects of healthcare products, or new aspects of known effects, in order to harness such effects (if beneficial) or prevent or mitigate them (if harmful).
Footnotes
-
Competing interests: none declared.
-
Provenance and peer review: not commissioned, not externally peer reviewed.