No changes to the requirements from EMA
Many companies are questioning the necessity of “Local Literature Screening” which refers to the part of pharmacovigilance literature screening that is not covered by global databases. We researched the topic and published our results in the first part of our blog on the topic: “Is Local Literature Screening Really Necessary?”. As a part of this research, we asked EMA if they have any plans to support the marketing authorization holders with this expensive legal obligation that typically produces little outcome. We have now received a response from the authority.
The marketing authorization holders (MAH) are required to screen non-indexed biomedical literature in those European countries where their product is authorised irrespective of the type or the marketing status of the product.
Our research into the number of individual case safety reports (ICSR) in EudraVigilance (EV) revealed that less than 1 % of all ICSRs annually reported to EV originate from local literature screening (LLS). Large portion of these ICSRs is comprised of conference reports, duplicate reports and reports that have errors or missing parts in their reference information.
For companies present in all EEA markets or “only” with an approved registration there, this is an activity that substantially racks up costs, for seemingly little purpose. A change in the requirement or further guidance from the authority that made this request of MAHs is needed. Unfortunately, we still need to wait for it…
Response from the EMA
On March 23, 2021, we wrote to EMA[i], asking them if they have any plans related to the requirement to screen non-indexed journals. Our message to them included the findings we have shared in our previous blog on the topic and a general enquiry about future plans to reduce the financial burden on MAHs. We also referred to the Medical Literature Monitoring (MLM) service[ii] that EMA is performing on behalf of the MAHs to reduce the duplicate effort and reporting on widely used substances.
On May 11, 2021 we received the following response:
“At present the Agency has no plans to change how the Medical Literature Monitoring service functions and to extend it to local literature. This is because MAHs are best placed to know both which products they have in which markets and also what are the relevant local journals published in those countries, and they would still have to monitor this literature for the purposes of PSURs.
In addition, one of the reasons for operating the MLM Service in the way we do is to reduce the burden of duplicate ICSRs on NCAs, on MAHs and on the Agency. As you point out, the cost of performing local literature monitoring is not insignificant; but the proportion of literature ICSRs which originate from local literature is rather small. This means that local literature monitoring would have a disproportionately large cost per duplicate ICSR saved.
The Agency keeps the scope of the Medical Literature Monitoring service under periodic review, and we will factor the information you have provided into our future considerations.”
Our response to the EMA
The core message from EMA is “No.”
This was the response we were expecting, but we are not satisfied with the justification.
How would the MAHs have the best knowledge what are the local journals published in the countries where their products are authorized? There is no one party who would have the best knowledge of the local literature published, but in our opinion the most logical candidate to hold this information should be the national competent authorities (NCA) of the Member States[iii]. It would also not really be unfair to expect EU and EMA to have an understanding of these types of sources as these are the authorities who requested their monitoring in the first place.
Those who follow the daily trackers published by the MLM service provider in EV, have probably noticed that it would not be large additional task to expand the screening results to include the benefit-risk relevant hits (non-ICSR) to reduce the need for duplicate effort of global literature monitoring (GLM) for the purpose of PSURs. We hope that EMA is now considering this in their ongoing tender for a new MLM service provider[iv].
Even if there is a disproportionately large cost per ICSR to screen local literature for the EMA, it would still be a cost-saving activity for the community. Now all the individual MAHs end up paying thousands of Euros annually for the LLS, often to find no ICSRs or to report the same articles for the 300th time to EV. MAHs are paying EMA for literature monitoring as part of their annual fees and it is time that LLS and other safety relevant information is covered by the MLM.
We have sent a slightly reformatted version of this blog as a response to the EMA on 27th of May, 2021. While we are not really expecting any action from their end, we want them to hear this.
LLS is not cost-effective
The cost of LLS for one product can often be several times higher than GLM which is performed in global literature databases. The cost of LLS is higher because it must be performed weekly or monthly for the time period the product is authorized. The screening requires involvement of several different pharmacovigilance specialists with different language skills. The sources differ in their format and in most countries, they cannot be screened in one place.
Due to the small amount of ICSRs reported to EV it might seem at first glance like LLS is a minor task for the MAHs. However, completely opposite is true. While LLS produces ICSRs to EV, it is arguable whether the additional information produced for the total benefit-risk profile of the product is worth the effort considering the cost of LLS.
In some European countries the local companies have realised this business opportunity created by the legislation and set up services around LLS, aiming for streamlining and higher level of automation, which we can only commend as we try to spearhead innovation in the industry.
Unfortunately, the value added by these services is limited. The companies subscribe to the local journals, then the journals are either scanned or downloaded by the company’s employees with optical character recognition (OCR) and uploaded to their database. Then a search can be set up for the MAH’s products in the database.
There are currently three problems with this type of service:
- The service is usually limited to one or two countries
- It is probably breaking copyright laws – subscribing to a journal and using the contents for business is probably not completely acceptable (though many organisations and authors may be happy if the information is spread)
- MAH still needs a local PV specialist to interpret the hits (unless they want to rely on Google Translate).
- Because of the manual scanning and uploading, only with a large number of products, the process may be cost-effective
We need a better solution
Tepsivo’s suggestion is that the most sensible course of action would be to have a deep look at the requirement and need for revision (and not just dismiss the idea), with the goal to decrease the time and cost burden by MAHs.
Should the requirement be abandoned altogether, and literature monitoring limited to global databases only, it would not have a significant impact on the signal detection or the continuous risk-benefit monitoring activities. The journal coverage would be reduced by less than 5 % if the product is authorized in all European countries, and Embase is used for GLM. But the legislation is very slow to update, as we have described in our blog Background to the current EU pharmacovigilance legislation.
In an ideal situation, literature should not be a source of ICSRs at all. The articles are usually written by healthcare professionals who should be aware of their possibility (post-marketing) and responsibility (trials) to report adverse events to the NCAs or to the trial sponsor. When the article has gone through the rounds of review and revision and is finally published in a respectable medical journal, the ICSR(s) should have been in EV for months.
Current situation is that the legal requirement stands, no support from EMA is to be expected and no service provider can provide truly full pharmacovigilance literature screening under the same roof.
To fulfil the legal requirement in a more efficient, more automated and less expensive way, we may simply need to rely on new ideas, and anyone who already knows Tepsivo probably realizes our Platform may be the place to start.