Adverse events following immunization: is this time for the use of WHO causality assessment?

Abstract

In recent years, public health authorities in industrialized countries have noted an increase in the numbers of parents choosing not to have their children vaccinated and in the activities of ‘antivaccination’ movements. Doubts about vaccine safety and lack of surveillance of adverse events following immunization (AEFI) are the most frequent themes proposed by antivaccination movements. This editorial aims to critically analyze the use of AEFI assessment procedures among national health authorities and public health researchers. In fact, the WHO recommended and published a systematic and standardized causality assessment process for serious AEFI, providing a method for individual causality assessment to be used by staff of national immunization programs, regulatory authorities and pharmacovigilance or surveillance departments. The last update was published in March 2013 but to date, an Internet search reveals no information or reports on AEFI surveillance that uses the WHO AEFI causality assessment.

Keywords:

• adverse events following immunization
• causality assessment
• Fluad case
• postmarketing surveillance
• WHO algorithm

In recent years, several public health authorities of the industrialized countries noted an increased numbers of parents choosing to have their children not vaccinated [1]. Trying to explain why this is occurring, public health experts refer to the activities of an antivaccination ‘movements’. Frequently, these movements denounce that vaccines are not safe and that national authorities do not carry out postmarketing surveillance activities about the incidence of adverse event following immunization (AEFI) [2].

In June 2005, the Global Advisory Committee on Vaccine Safety suggested WHO to consider the need for improved monitoring and analysis of vaccine-related adverse events globally, to attribute the causality to AEFI, especially those considered severe, for ensuring vaccine safety [3]. For this reason, during the same year, WHO published an aide-mémoire to a systematic and standardized causality assessment process for serious AEFI, providing a method for individual causality assessment to be used by staff of national immunization programs, regulatory authorities and pharmacovigilance or surveillance departments. The manual was developed with significant input by key vaccine safety stakeholders, and WHO recommended that national authorities spread the AEFI causality assessment procedures among public health and pharmacovigilance services [4].

In the following 7 years, several limitations had been identified during its use in the field, including the need for more detailed guidance on the elements required to perform the assessment of causality, confusion over the terms used to classify the likelihood of association of the event to the vaccine and the incomplete use of parameters for establishing causal association [3].

In March 2013, a group of the Global Advisory Committee on Vaccine Safety updated the protocol for the causality assessment of an AEFI and published a user manual as a guide to a systematic, standardized global causality assessment process for serious AEFI. Following this protocol, the AEFI’s association to the vaccine could be categorized as ‘consistent’, ‘not consistent’, ‘undetermined’ or ‘unclassifiable’ [5].

In the view of a standardization of AEFI assessment, the Brighton Collaboration, an independent global vaccine safety research network for healthcare professionals, in 2002 developed a standardized case definitions for an exhaustive number of AEFI complemented by guidelines to facilitate their use; Brighton definitions, periodically updated, were widely broadcast and globally accepted [6].

Despite WHO recommendations and Brighton efforts, in biomedical literature, there are few examples of postmarketing AEFI surveillance studies. A research on PubMed database using as key words ‘adverse events following immunization post-marketing’ identified only 30 articles: the oldest was published in 2000, the most recent in December 2014, 9 of the 30 were published last year. Nineteen of them are research articles; in particular, 16 are studies on adverse events of only one vaccine (vaccines against Japanese encephalitis, influenza, MMR, HPV, rotavirus, anthrax and diphtheria, tetanus, and acellular pertussis). The others are reviews (n = 8) and editorials (n = 3).

We repeated the search by adding ‘causality assessment’ to the key words and the number of the findings dramatically decreased. If the key word ‘adverse events following immunization post-marketing causality assessment’ is entered in PubMed search tool, only two articles are shown [7,8]. Neither of them described the use of the WHO AEFI causality assessment methodology, and one of the two was published before 2005.

Then, trying to expand the research also to clinical trials, we removed ‘post-marketing’ from the key words, and the number of studies does not greatly increase. We found 64 articles published between 1989 and 2014. Twenty-one were published in 2013 and 2014.

The only one that cites the WHO AEFI methodology is the article that explains the new WHO algorithm for AEFI classification [9].

Our search strategy could be too narrow because authors often refer simply to ‘adverse event following vaccine X’ or assess a specific adverse event, which is named or discuss the ‘safety’ of vaccines, then we could fail in considering some scientific contributions. Although, this approach reflects a lack of interest on the methodology of assessing AEFI, also from national and international scientific organizations.

As well as in biomedical literature, also in the Internet, there are few reports on AEFI and often they are in national languages; most of them are reports of Public Health or Drug Safety Agencies [10–12]. Until now we did not find in the Internet any information or reports on the AEFI that use the WHO AEFI causality assessment.

The protocol does not seem to be used neither in scientific research, neither in the clinical practice of National Agencies.

There were several possible reasons for lack of use of the WHO manual in the literature. For example, causality assessments on an AEFI occurring in an individual may not lend itself to a peer-reviewed publication. Also, data from National Agencies could be not accessible for researchers, because of national law about pharmacovigilance or privacy.

In November 2014, the Italian Medicines Agency suspended, as a precautionary measure, the administration of two batches of Fluad flu vaccine after three postvaccination deaths. The deaths occurred in people aged 87, 79 and 68 years and were reported through the National Network of Pharmacovigilance as possible adverse events following the vaccination. The National Authority performed investigations on the ‘suspected’ vaccine batches to assess their safety and their investigations concluded that the Fluad flu vaccines were safe. In the following weeks, the number of deaths reported as suspected adverse reactions increased, probably as a result of the massive media coverage, which has also likely strengthened the public’s hesitancy to be vaccinated [13].

None of the Italian Medicines Agency official communication reported the use of WHO AEFI causality assessment in the evaluation of these deaths, which was really not used by the Agency; according to data of National Pharmacovigilance Network (one of the authors is member of this Network), the deaths occurred in people who had several comorbidities so that the association between vaccine and deaths could be judged as ‘inconsistent’. Nevertheless, the new could cause fear and panic among the population, with the risk of a dramatic fail in the vaccination campaign and of an increase of influenza incidence.

The ‘Italian Fluad case’ revealed the lack of the use of a consistent, shared and evidence-based methodology in the AEFI surveillance. Probably, National Authority preferred other perspectives, as a paradoxical ‘principle of precaution’ that has been applied to improbable vaccine-related deaths and not to the prevention of vaccine preventable deaths. Finally, the Italian Fluad case was discussed also in recent article published in BMJ, but this article (written by very notable public health scientists) did not refer to the WHO AEFI assessment protocol [13].

Thus, this is the time for the scientific approach on the ‘AEFI questions’, based on the extensive use of WHO causality assessment protocol, both in the national routine surveillance and in the postmarketing ad hoc survey.

WHO guidelines while potentially useful are not the exclusive means of examining potential causal relationships between vaccine administration and adverse reactions; national and international authorities must promote active surveillance system, in particular, for newly introduced vaccines, as performed in Italy after the introduction of HPV vaccines [14]. Anyway, WHO approach is more economic and consistent.

WHO must encourage national health authorities to more broadly use this protocol; if we want to maintain the public’s confidence in vaccines, the availability of standardized data about AEFI is crucial in the perspective of the public health accountability (e.g., to monitor rare AEFI).

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.