Following the publication of the latest EMA guidance, my previous blog examined the five new XEVMPD fields that are required by the EMA. This time I’ll cover how to improve the quality of your data and the impact of the tidying up the controlled vocabularies. Local language SmPCs and timelines will be in the final blog of the series.
Wanted: Quality as well as quantity
On the one hand, I was surprised at how poor quality the data within the EVMPD appears to be, if the EMA’s two pilot assessments are anything to go by. On the other hand, given the mad rush to get the data submitted by the July 2012 deadline, and the lack of clear guidance about what’s required, it’s probably not surprising. What to do about it though?
Well, this time there is more time, and also the guidance is clearer. We’re still waiting for a new FAQs document and the publication of the data quality guidelines, but we do have eleven months to comply instead of four, and chapter 3.II is now a useful document. Plus, we’re not starting from scratch this time; much of the data is already gathered, so it’s more a case of checking and adjusting.
The good news is that there’s going to be a way of extracting all your existing EVMPD data out of the EMA’s database. This can then be easily imported into a Regulatory Information Management (RIM) or XEVMPD tool for checking, correction and resubmission. Better still, if you have a properly structured RIM tool many of the errors are either not possible, due to the rules enforced within the database, or ‘helper’ tools are provided to help you check and correct the data.
If you’re stuck with using EVWeb then your best bet is to use its query tools to extract and double-check the data in sections via spreadsheets then enter the corrected data.
Don’t forget to download and study the data quality guidance as soon as it’s published though; this is expected in March.
We’re all aware of the mess that some of the Controlled Vocabularies have got into. The good news is that it’s the EMA’s responsibility to clean them up.
They’ve done the pharmaceutical forms and the routes of administration and published spreadsheets giving the mapping between the incorrect and the correct ones. Here’s the document explaining this:http://www.ema.europa.eu/docs/en_GB/document_library/Other/2013/11/WC500153999.pdf. For our customers we’ll be providing a report to identify if/where the obsolete EVCodes are used and offering a script to correct them.
The substance list is still being worked on, and once complete this will be a similar story. The difference is that the EMA cannot just delete the old substances as so many of them are in use; instead they are planning an amendment to the acknowledgement message schema so that they can ‘report back’ if there are any incorrect substances codes in use. The correct code to use will also be provided, so that users and software vendors can correct their data in an orderly fashion.