I got the impression that
@yodon and
@Ed Panek were relating to Clinical Investigations (clinical trials), whilst
@Raisin picker related to the literature-based Clinical Evaluation...?
I assume that the OP was referring to the latter.
I agree with
@Raisin picker that most companies (especially small ones) outsource it to specialists. It is indeed a specialty, and sadly the level of competence varies a lot among those who offer this service. I highly recommend cross referencing anyone you are about to hire, and get confirmation from multiple past clients to ensure consistent good performance.
I can't comment on cost, but I can say for sure that lead time is a real issue. In my experience, the job is seldom done in the allocated timeframe, so whatever you go ahead with, I recommend accounting for at least double in your project plan. I also recommend allowing for multiple rounds of review (by yourself) and revision by the author. Obviously, the less straightforward the device is, it all has the potential to explode exponentially. It's one of the trickiest, and most important, tasks in medical devices development and regulatory clearance.
Also important to get an excellent understanding (in house) of your specific NB expectations around it, before engaging with anyone, and then making sure the CER author is on the same page (preferably, put it in the contract). At the end of the day, it's the NB clinical reviewer who will give the final blessing on the CER (or not). Which reminds me... include wording in the contract to ensure the CER author is available and committed for fixing any issues after NB review.
Please also note that you will need a clinical specialist in the relevant niche to endorse the evaluation (unless the author is also such a specialist, which is uncommon, they can't do it). CER authors don't necessarily provide that service, and expect you to bring on board your own clinical specialist (typically a Dr, but not always).