Use of distributed data networks

During the past 2 decades, a cornerstone of active post-marketing surveillance has been assessment of vaccine safety through large distributed networks based on linkage of large data collections. Distributed networks are independent groups with a common goal – in this case, investigation of concerns about vaccine safety. Such collaborations can achieve more power for studies than any single centre and can share methods and achieve greater generalizability. These collaborations have variously included common data models, protocols, analytics, case definitions, global scope and sustainability. Most were conceived for ad hoc issues and proved to be time-consuming and expensive. A platform that contains all the desirable features is necessary to address vaccine safety issues (real or perceived) in a timely, robust manner in globally relevant populations. In addition, with the globalization of vaccine manufacture, means are required to compare brands and schedules. Developing capacity for this type of research in LMICs also requires an established platform.

GACVS was informed about a new global vaccine data network (GVDN), the aim of which is to establish a network of collaborators with capacity in vaccine data linkage, supported by a central coordinating entity. The network would foster inclusion and mentor organizations that wish to link data in order to assess the effects of vaccines. An inaugural meeting was held in Annecy (France) in January 2018 to address the lack of data to properly understand, demonstrate and compare the risks versus the benefits of vaccines in various populations and individuals with different medical risk factors; the lack of adequate infrastructure and capability to monitor and evaluate vaccine outcomes in LMICs; delays in collating data for making decisions during global pandemic events; the lack of readily accessible, evidence-based material for use in communications to enhance public confidence; and lack of global coordination among assessments of vaccine safety and confidence in vaccines and communications.

Representatives from academia, research centres, industry and GAVI in 16 countries in all 6 WHO regions attended, providing extensive collective epidemiological, analytical and communications expertise. The potential estimated population under surveillance in the sites represented is 155 million people. The meeting reached agreement on a collaborative model for conducting studies of vaccine safety, efficacy and risk–benefit; a governance model that ensures full participation of sites and efficient development of study protocols; data models to protect individual privacy but allow collaborative agreements on common data models (standardization of data to allow pooling of results); and a pilot study to investigate the link between influenza vaccines and Guillain–Barré syndrome.

Current examples of distributed networks were presented. India has been building national and institutional capacity in vaccine pharmacovigilance. National networks can address poor access to electronic records in diverse settings, highlighting the potential of distributed networks in LMICs. India has established baseline rates of hospitalization for selected conditions and assessed potential associations with vaccination, such as for rotavirus vaccine and intussusception. India also collaborates in a multi-country collaboration on the safety of vaccines to be used in pregnancy in LMICs (vaccines against Zika virus, respiratory syncytial virus, group B streptococcus and hepatitis E) using the definitions of the Global Alignment of Immunization Safety Assessment in pregnancy to assess the minimal detectable risks for selected health outcomes. Data collection has begun at 24 sites, primarily in African counties and in India. A new initiative in Europe, Vac4EU,¹⁰ was presented to the Committee. The aim of the collaboration, which is based on the ADVANCE project, is to provide a sustainable platform for the European Region.¹¹

The Committee viewed long-term funding as a potential obstacle to maintaining sustainable infrastructure for a global vaccine data network. The role of local governments will be important in supporting the functioning of their own data sites, perhaps in partnership with the formal vaccine safety surveillance programmes within their national immunization programmes and through partnerships with active and passive monitoring communities. The advantages for countries would be access to the entire GVDN. The Committee also considered that the regulatory imperative for post-licensure surveillance could sustain support for a GVDN through a common pool. The requirement for this type of active surveillance in GAVI-eligible countries as a condition for receiving vaccines could drive capacity-building in those countries. Political will to support this type of network will be critical for a strong distributed network, as exemplified by India. With the development of new vaccines, new manufacturers, pandemic threats and increasing vaccine hesitancy, GACVS considers sustainable global collaboration in a vaccine data network to be an important, if not vital, component of future global vaccine pharmacovigilance, with prioritization of political and financial support.

¹⁰ Vaccine Monitoring Collaboration for Europe (http://www.vac4eu.org, accessed June 2019).

¹¹ Advance (http://www.advance-vaccines.eu/, accessed June 2019).

Report of GACVS meeting of 5-6 June 2019, published in the WHO Weekly Epidemiological Record of 12 July 2019