The new revision of the EC guidance on Clinical Evaluations (i.e, CERs), was issued in June of this year.  MEDDEV 2.7.1 Rev 3 is now Rev 4.  The industry has provided many ways to learn about the new document via white papers, webinars and conferences (thanks!).  The education has been great, but what’s it like for someone who is used to writing CERs to the old Rev 3 to actually go through the process of writing to Rev 4?  I’m happy to share.

First, the good news – there is little content that is actually new.  The general requirements in both revisions are basically the same.  What is different is the expected rigor in a CER produced to Rev 4.  The Rev 3 document is 46 pages long, with 10 of those being a checklist for Notified Bodies.  The checklist is gone in Rev 4, but the document has still grown to 65 pages.  The earlier revision was almost “easy-breezy” in its wording by comparison.  In the new version, the tone has changed from “This is what you should do for this section” to “You really have to do this, in great detail and with relevant support, and you don’t have a choice, so just do it.  Cross-reference it and make sure someone double checks your work.  And don’t talk back.”

For example, Rev 3 spent all of 72 words to describe who should perform the clinical evaluation.  Rev 4? 275 words.  And it’s all muscle mass and no fat.  The section on appraisal of clinical data grew from one page to nearly 8.  But I’m not complaining; to an engineer, clarity is good and ambiguity is bad.  Having specifics helps us, the client and the Notified Body to all be on the same page.

So what does this all mean for the practical task of writing the CER?  I was recently asked to upgrade an existing CER to Rev 4 as part of a routine update.  The existing document had been well-written to the Rev 3 guidance by a regulatory expert.  When I was done, I had spent over half again as much time as I had planned and produced an updated CER that was twice as long as the original.  Part of that effort was spent on a new data set appraisal scheme, since Rev 4 of the MEDDEV no longer includes the poorly conceived “possible method” of Rev 3 (good riddance).

We learned from the experience and embraced the need for rigor.  We upgraded the clinical evaluation SOPs and forms we use at Medavise to be as robust as the Rev 4 requirements.  As a result, the assessment of each data set is deeper, the demonstration of equivalence is stronger and the conformance to essential requirements is clearer.  Which, of course, seems to be the whole point of the new MEDDEV revision.