Skip to main content

ICMRA meeting: COVID-19 Real-World Evidence and Observational studies

13 October 2020


Chairs: Kelly Robinson (Health Canada) and Xavier Kurz (European Medicines Agency)


1. Welcome and introduction

Xavier Kurz* (European Medicines Agency - EMA) opened the meeting by welcoming his co-chair Kelly Robinson (Health Canada - HC) and all the participants on the line. The aim of this 4th ICMRA call on COVID-19 Real-Wold Evidence observational studies was to share experience on initiatives ongoing in some of the regions, and to update on the progress made by the three technical workstreams initiated at the 19 May 2020 ICMRA meeting related to vaccines surveillance and vigilance, pregnancy and international cohorts. Kelly indicated that as the pandemic continues to evolve, the importance of studying COVID-19 remains extremely high. In many regions, we are observing a rise in cases and outbreaks, including in areas that previously brought case counts under control.

* Apologies from Peter Arlett, EMA

2. Sharing experiences of work on observational studies including sharing information on any initiatives

On Friday 9th October, the ICMRA secretariat launched a survey to enquire the ICMRA members on the initiatives relating to observational studies in COVID-19 ongoing or planned in their countries/regions, as well as on any learnings that could be shared with other regulators and on opportunities to leverage the work internationally.

Two agencies provided feedback during the teleconference:

  • The Danish Medicines Agency (DKMA): Jesper Kjaer (Head of Danish Medicines Agency’s Data Analytics Centre) presented the COVID cohort analyses and vaccine surveillance led by DKMA in collaboration with academic institutions, patient organisations and other national authorities. Eleven studies have been initiated so far based on existing national registries to investigate potential complications of existing treatments, such as NSAIDs and ACE/ARB. A steering group composed of representatives from various stakeholders with relevant expertise decides which analyses the expert group should carry out. All proposals approved are made public on DKMA’s website and in the EU-PAS Register to increase transparency. The analysis looking at the impact of NSAID on the occurrence and seriousness of Covid-19 infection by using several propulsion-based databases (Danish Microbiology databases, the Danish National Patient Registry, and the Danish National Prescription Registry) was particularly highlighted. Findings suggest that NSAIDS do not lead to a more severe coronavirus disease.
    The same registries will be used for vaccines surveillance. Phase 1 of this study will be to combine data from the Vaccines registry with side effect reports received by DKMA and the National Patient Register. This study will produce a general surveillance of vaccine rollout with daily statistics to rapidly detect side effects. Phase 2 of this study will incorporate other data sources e.g. national prescription registry, biochemical data and interaction with GPs, to allow for the early detection of side effects.
    Key messages to take away: data centralisation through existing population-based registries have contributed to the success of DK’s rapid preparedness for and response to the pandemic; importance of an independent steering group with representatives from all stakeholders to ensure appropriate expertise and reduce political implications.
  • The Japanese Agencies (PMDA) highlighted the major studies ongoing in Japan, namely the “Observational studies for Alvesco (Ciclesonide) and Avigan (favipiravir)”, and the “Registry study for Covid-19”. The interim reports are only available in Japanese, but PMDA remains at colleagues’ disposal for any questions they may have.

3. Update on the three technical workstreams initiated at the 19 May 2020 ICMRA meeting (slides attached below):

a) Vaccine surveillance and vigilance (MHRA)

Philip Bryan (Medicines and Healthcare Products Regulatory Agency - MHRA) presented the progress on the COVID Vaccine Pharmacovigilance Network co-chaired by the MHRA and the Therapeutic Goods Administration (TGA) in Australia which aims at sharing best practice and planning for pharmacovigilance of COVID-19 vaccine(s). The network comprises so far 14 countries + EMA and WHO and has raised interests from non-member organisations willing to engage in the activities including the IFPMA Covid-19 Vaccine Working Group on Pharmacovigilance (pharmaceutical industry), the vACcine Covid-19 monitoring ReadinESS (ACCESS) consortium led by Utrecht University, and the Global Vaccine Data Network (GVDN) led by New Zealand.

The group initiated a survey among its ICMRA members on the vigilance readiness activities in view of creating a repository of all these initiatives that will be made widely visible to facilitate sharing of knowledge and reduce duplication. The scope of activities is large and includes the sharing of:

  • existing sources of data available to support pharmacovigilance, including spontaneous reporting systems and the use of electronic healthcare records for signal detection/strengthening,
  • ad-hoc/dedicated active surveillance programmes and the use of bespoke IT solutions,
  • approaches and methodologies for signal detection and strengthening
  • priorities and case definitions of AESIs, and exchanging information on background rates of AESIs to inform surveillance
  • information on emerging safety signals and planned communications.

Countries/Regions are likely to use their own methods to monitor vaccines and overlap of activities is unavoidable. However, the idea here is to promote awareness of the various national / regional approaches so regulators may identify opportunities for harmonisation on some aspects of vaccines surveillance, which could potentially lead to the agreement on common procedures and foster international collaboration (e.g. pooling of data for research purposes).

DKMA inquired if this group will set up similar analytical programs across regulators instead of pooling data. At this point this has not been determined, but is currently happening with third party organisations, and highlights the need for information sharing to gain a common understanding of current initiatives.

Another question was raised on the possible outcomes of the interactions with pharmaceutical industry associations. Discussions with industry are only starting and any suggestions from the ICMRA members on the scope of collaboration, expected outcomes and types of recommendations are welcome.

b) Pregnancy observational research (EMA)

Corinne De Vries (EMA) presented the ongoing work on Covid-19 infection in pregnancy. Back in July 2020, EMA contracted the EU PE & PV Research Network led by University Medical Center Utrecht (UMCU) in the Netherlands to conduct studies in Europe & prepare for meta-analyses on the impact of COVID-19 infection and medicines in pregnancy.

The project is called CONSIGN (Covid-19 infectiON and medicineS In pregnancy) and aims at guiding evidence based decision-making about COVID-19 vaccine indications, vaccination policies, and treatment options for pregnant women. The consortium is composed of 3 main networks including Conception (Innovative Medicines Initiative -funded project with 88 partners), INOSS (International Network of Obstetric Survey Systems - Multinational collaboration of organisations conducting prospective population-based studies of serious illnesses in pregnancy and childbirth) and COVI-PREG (Multinational collaboration of organisations conducting prospective population-based studies of serious illnesses in pregnancy and childbirth using tools developed for the Zika crisis).

Case control designed studies will be performed within the INOSS and COVI-PREG networks to estimate the impact of specific medicines on pregnancy outcomes in pregnant women with COVID-19 through primary data collection. In parallel, studies based on secondary use of real-world data collected in health care databases will be performed in 9 population-based electronic health and medical birth registers in 8 European countries to estimate drug use, outcomes of COVID-19 and pregnancy outcomes related to medicines. Some of the databases used in the 3 different workstreams overlap, but this will allow for data validation. The 1st protocol on the retrospective studies will be delivered in October 2020 and the final analyses are due to be delivered in July 2021.

The consortium closely works with partners outside Europe (e.g. Sentinel, CNODES, other ICMRA groups, as well as INOSS and COVI-Preg groups outside of Europe), giving the opportunities to collect data from other regions/continents and to increase the numbers of patients for analyses.

In parallel, an ICMRA technical experts working group on pregnancy (Health Canada, WHO, FDA, MEB, AIFA, and MHRA) led by EMA has been established to foster international collaboration on global meta-analyses. The idea is to share the CONSIGN protocols with all countries wishing to participate and adapt it to their data sources in order to reach meaningful numbers for global-scale analyses focusing on the electronic health records/population-based cohorts. An inventory of relevant COVID-in-pregnancy ongoing and planned initiatives, existing data sources and networks was created based on the input from the ICMRA technical group members and is available upon request. These could potentially be consulted if more details on specific cases/signals are needed in the future.

The ultimate goals of all this work, is to get a better understanding of the natural history of COVID-19 disease in pregnancy & impact on mother and baby and prepare for response to questions regarding treatments currently used in COVID-19 affected pregnancy. Provided we are successful in reaching these objectives, we will have managed to build a solid infrastructure that can then be re-used beyond the COVID pandemic in the context of benefit-risk monitoring of medicines in pregnant women and appraisal of proposals for risk management planning.

A question was raised as to what methods will be employed to ensure there will be no duplication of patients included in the analyses considering that some databases overlap across the 3 main networks. Corinne clarified that this has been identified as a potential issue which the consortium has been asked to address in their protocol that will be submitted shortly.

Contacts in case ICMRA members are interested to join the technical group on pregnancy: and

c) Building international cohorts (HC)

Gaya Jayaraman (HC) provided an update on the ICMRA Building International Cohorts of COVID-19 Patients technical sub-working group ongoing activities. This group was established to share expertise and increase study power and data quality in order to meet regulatory requirements and address existing knowledge gaps. It is led by HC, with membership from FDA, EMA and AEMPS.

Over its last 3 meetings, the group shared activities planned or ongoing in their regions in view of identifying gaps and/or build on ongoing work at international level. The members agreed upon criteria for the identification and prioritisation of specific research questions which would benefit from international collaboration. Three topics were selected, namely the COVID-19 Case Definitions study led by HC, the steroid use study led by the EMA and the coagulopathy study led by FDA. This work will hopefully lead to the development of protocols that will leverage real world data sources to address key knowledge gaps.

The goal is to implement common protocols and/or define analytic approached to address heterogeneity by Winter 2020. To date, the building international cohorts working group has the benefit of shared learnings on the process for regulatory collaboration on key scientific prioritises, both within and beyond the context of the pandemic, both scientifically and operationally.

  • Steroid use in COVID-19 outpatients and inpatients (Xavier Kurz): IQVIA (in collaboration with Erasmus University and University of Oxford) has been contracted by EMA to run the study on databases converted or being converted into the OMOP data model. The primary objective is to describe utilisation of systemic corticosteroids (e.g., dexamethasone, prednisolone, methylprednisolone or hydrocortisone) for treatment of COVID-19 in two types of setting: hospitalized (in hospital care) and ambulatory (any care received outside the hospital) within 90 days following COVID-19 diagnosis. The draft protocol was received in September and has since been reviewed by EMA, FDA and HC. Four major questions and a large number of minor questions have been raised and communicated to IQVIA. The updated protocol is awaited by 30 October.
  • Coagulopathy study (Silvia Perez-Vilar): FDA is planning to run the study in Sentinel using a retrospective cohort design. The primary outcomes are the incidence of arterial thrombotic events (acute myocardial infarction or acute ischemic/ embolic stroke) and venous thrombo-embolic events (acute deep vein thrombosis or pulmonary embolism). The study protocol is being finalised and a feasibility data assessment to help identify data partners is ongoing. The final results are expected in August 2021.
  • EMA announced its plan to contract a third party to conduct a study in Europe with similar objectives, study design and variables as far as possible, given the differences existing between the databases of the Sentinel system and those available in Europe, in view of pooling the data across continents for future meta-analyses.

Questions and answers:

  • The FDA was asked whether a CDM will be used for the coagulopathy study. This was confirmed, as a CDM is already implemented in Sentinel, but FDA specified that they will include as many cohorts a possible in the analysis.
  • As the protocols for the studies on coagulopathy and steroid use are expected to be finalised soon, a question was raised as to the possibility to share the documents with the ICMRA members. This is indeed the intent, so the protocols can be implemented in all countries/regions wishing to participate.
  • Query on need for agreement of confidentiality of information for participating in the research initiatives: This aspect has been discussed within the technical group on pregnancy. EMA will follow up with Korean colleagues who asked the question.
  • Finally, the importance of communication and transparency on the progress of the different workstreams was highlighted. In addition to publishing the minutes and press releases on the ICMRA website, other communication channels could be considered in order to reach as many regulatory agencies and stakeholders as possible.

4. Summary and next steps:

The co-chairs closed the meeting by acknowledging the impressive work achieved so far and emphasised the importance of the ICMRA collaboration, especially given the COVID-19 pandemic. The forum allows for regular updates on the progress of each technical groups, promoting transparency and knowledge sharing.

The intervention from DKMA showed the opportunity of the existing registries to effectively establish a wide range of studies on a large population.

Colleagues were encouraged to complete the survey circulated by the ICMRA secretariat circulated on Friday October 9 in order to collect information from various regulatory agencies worldwide.

The next meeting will be scheduled in approximately 6 weeks’ time by the ICMRA secretariat which was thanked for all its support.