They need to ascertain the incidence of the suspected reaction and whether it is actually caused by the drug in question – or by another agent or process. This page provides a brief introduction to some of the challenges within pharmacovigilance in arriving at such accurate assessments, when considering information received via spontaneous reporting within the EU and elsewhere.
There may be spontaneous reports arriving at the pharmaceutical company, or regulators or at other agencies. The causality link must be estimated and it must be clearly understood whether patients were taking any other medications at the time. The time between starting the suspect medication and the onset of the reaction should also be established. A shorter onset would often lead to greater suspicion; a longer one could mean that causes other than the medicine are more likely, although it depends on the nature of the adverse event. The health status of the patient is also important, particularly if they are at risk of developing new pathologies, or if a disease from which they were suffering is prone to fluctuate in severity. Along with a number of other diagnostic criteria, an assessment can be made grading causality as definite, probable, possible, or unlikely, or anywhere along the spectrum between definitely related and definitely unrelated. It would be desirable – but unfortunately not possible – to be able to estimate the rate of occurrence to inform future clinical decisions from a drug safety perspective. This is because of the unknown extent to which adverse reactions are under-reported – more of which below.
Each report arrives as unsolicited data, provided by doctors, nurses, pharmacists, or by patients themselves. The quality of data is very variable – there may be scant detail, including omissions on health status, the issue of polypharmacy and other important information. It may be that a busy professional is simply unable to supply comprehensive records, or that issues of patient confidentiality prevent them from doing so. With reports generated by patients themselves, there may be sensitive issues involved which mean they are reluctant to disclose certain types of information, or they may not have understood some of the detail. Whether originating from professionals or patients, it can then sometimes be difficult to follow up reports for any number of reasons.
This describes the correlation between the launch of a new medicinal product onto the market and a high level of voluntary reporting of suspected adverse reactions. It has generally the case that the volume of reports is higher within the first two years of prescribing for a new drug. The volume tends to follow an upward curve from the first 6 months to 2 years, followed by a gradual decrease to a steady rate: this is known as the Webber Effect.
The public awareness of a particular drug is heightened if it becomes ‘common knowledge’ that it has been taken by a celebrity and should there be any suspicion of an adverse reaction to the drug this will be subject to media exposure. An example might be where a celebrity has publically claimed to have had an ‘allergic reaction’ to a drug which led to their hospitalisation, More commonly, a “drug scare” can result from media activity if there are suspicions of inappropriate behaviour by the pharmaceutical industry or some shortcoming of the regulatory process. A warning from the regulatory authority or a public health agency can sometimes trigger such concerns. Typically the result is a temporal response, with increased reports corresponding to the release of information in the media about the medicinal product- whether or not that information is an accurate reflection of the actual safety profile of that product.
Whilst Pharmacovigilance professionals would naturally wish to estimate incidence rates for adverse drug reactions in reality this is not possible based on spontaneous reporting. This is because the reporting rate is an underestimate of the true occurrence rate of an adverse reaction: it is the case that there are more happening than are ever reported. In the UK, for example, researchers have estimated that 90% of all serious adverse reactions are not reported, and the figure rises to between 96 and 98 % of all non-serious reactions. It has also been calculated that just 1% of all serious adverse reactions may be reported in the USA. The extent of under-reporting also varies according to the type of adverse reaction. Moreover, it is often difficult to obtain accurate information about the extent of usage of a medicine, so that the population exposed may be uncertain.
Millions of patients may be taking a drug for which there are a just a handful of reports raising the suspicion of a serious adverse reaction and some of the reports could be of poor quality and others could have been caused by something different! Thus, it should be understood that the main – considerable – value of collecting and analysing spontaneous reports is to raise the possibility of a signal of a previously unknown adverse reaction to a medicine, or some change in the nature of known adverse reactions (such as greater severity than previously suspected). Once such a signal has been generated, it is necessary for them to be evaluated and then confirmed by other means.
So however many limitations spontaneous reports placed on drug safety work, it is nonetheless essential that these systems are in place. It would not be feasible to mount sufficiently large clinical trials in order to accurately reflect real life prescribing. In the meantime, an experienced pharmacovigilance services provider offers the necessary expertise to process the reports to the highest possible standards, aiming to protect the public and facilitate accurate, timely and compliant reporting to the regulatory authorities on behalf of pharmaceutical companies.
+44 (0)1483 307920
(+1) 781 703 5540
United Kingdom (Head Office)