A pharmacovigilance paradise?

The second listing, let’s call it the ICSR listing, has two versions:

The daily snapshot of in-process articles: this will present all articles where the presence or absence of ICSRs has not yet been confirmed.

The completed archive: this will present all articles where there are one or more ICSRs, or where there are no ICSRs, or where there may be ICSRs but no verification was possible.

If multiple ICSRs are identified in the same article then multiple lines will be present, each with a single World Wide ICSR ID.

The criteria for Inclusion or Exclusion of the information in the article for ICSRs is encoded, as well as seriousness, potential or confirmed causality, concomitant drugs, and a plethora of other information.

Note: This is just the bibliographic part of the MLM, in a single row, of an Excel spreadsheet. For the Safety Management systems, all confirmed cases detected by the MLM also exist as E2B(R2) XML files ready to be downloaded and imported for further processing, follow up and analysis.

Or a pharmacovigilance nightmare?

The harvest of references for pharmacovigilance purposes is a result of two factors: the databases used the searches applied to each database.

This EMA harvest will differ from the daily/weekly harvest of references by MAHs (and CROs on behalf of MAHs) by a variety of reasons:

Some MAHs use only PubMed; other MAHs use very well-constructed or very poorly constructed searches on PubMed and/or on more content rich databases like Embase.

All MAHs are expected to scan local literature that EMA will never scan. Of course, no two MAHs will scan the same sets of local literature, as their national operations will always have different capacities for literature screening.

Anyway, the inevitable results is: for each substance on the MLM initiative, each MAH has a universe of bibliographic references and articles to review and the EMA has another universe.

EMA’s Medical Literature Monitoring - Industry Insights | Elsevier R&D Solutions


  • MAH is fully responsible for executing literature screening for ICSRs (Yellow and Green area).
  • MHA must report ICSRs detected on literature from Yellow area.
  • MAH must not report ICSRs from Blue area.
  • MAH must incorporate the cases detected by EMA from the Blue area in their safety management systems. Integrating the literature references and articles in the literature screening systems may have advantages in the long term.
  • MAH must report the ICSRs that have escaped EMA (by screening the Green area) as QC incidents, not as ICSRs.

So, it is vitally important for MAHs to cross-reference the articles that entered the EMA system and find ways to harmonize the information with those that are entering the MAHs systems, because the MAH must not submit duplicate ICSRs.

The harmonization effort required seems daunting, and it probably is. I’m sure data engineers in charge of keeping the information domains of their Safety and Pharmacovigilance operations feel that they have full justification for budget increases.

Harmonization starts on literature screening efforts, attempting to achieve harmonization between the EMA MLM data and each MAH literature screening pools. There is a lot you can learn and gain with MLM, and it main even happen that one day you will find you can’t live without it. Let me give you some pointers:

Tip 1: As soon as possible, find a way to detect your Yellow area, and screen it while EMA is screening the Green area. As a MAH, you have access to the reserved areas of EudraVigilance, and can harvest that information daily. If your PV bibliographic harvest is weekly set it to run on the day following the day EMA runs their own. Then compare immediately your harvest with the published data on the articles spreadsheet out of the EMA reserved area Eudra Vigilance site. You now have your Yellow area!

Tip 2: There is no automated way to collect the spreadsheets; lobby the EMA for one to be created. In the meantime, get an intern to do it!

Tip 3: Pay regular attention to the Blue area, i.e., those references and sources that never crossed your radar. Do the ICSRs matter? What does your Risk and Signal Detection have to say about them? Should you be switching or extending sources? Is there information relevant for PSURs in those information artifacts… Enjoy the data discovery that EMA is doing for you.

Tip 4: Although MLM targets a very narrow set of generic medicinal products, no drug exists isolated in its own universe. Either as a concomitant drug, as part of a drug-drug interaction or as an optional suspect drug, all MAHs will, every once in a while, be referenced in the MLM ICSR List.

I believe everyone in PV, from literature surveillance Officer, Manager of PV Operations up to Senior VP of Safety, should be looking into ways to capitalize on the fact that EMA is funding a new source of signal detection. This data will be closely scrutinized by EMA’s Pharmacovigilance Risk Assessment Committee (PRAC). If you can get access, be sure to closely scrutinize it… you can, now, know a good deal of what PRAC knows, and I’ve been told that it is a Senior VP of Safety’s dream product!

One final thought:

According to the EMA, the amount of MLM screened products will vary as budget is available. The way the products are selected based on budget is very simple: more money means the EMA screens from product 1 to 175; less money, the EMA screens from 1 to 25. Simple and clean… but everyone should be ready for the day EMA has money to screen from 1 to 300, as well as the day EMA can’t even screen product 1. Do not rebudget your post-marketing safety departments, assuming EMA’s MLM will be there as a cushion/replacement for your Literature Screening I will go as far as to say EMA’s MLM will create a whole new chapter in Literature Screening audit findings.

Stay Up to Date on Pharmacovigilance News

Sign up for Elsevier's Pharmacovigilance newsletter and get the latest thought-leadership on pharmacovigilance and information management.