If it is a postmarket trial, same intended use as a marketed device for which you receive complaints through other channels (e.g., from users), then that would make sense
The whole point of complaints is to monitor the safety and performance of the devices you are marketing. If the device in the trial is not a marketed device, what's the point?
On the other hand, if the device is not marketed, but the trial is submitted to FDA and FDA then clears it, and you start selling it, then those complaints are the first complaints received about the marketed device.
On the other other hand, if it was a premarket trial, then these complaints are not PMS, and, depending on your processes, they could get mixed in with PMS and post-market risk management. To me, the data in the pivotal trial establishes a baseline at the time of market release, and complaints monitor changes to that risk profile.
Also, although you ask about Part 820, if you also do EU CERs, and include both trial data and complaints data, then the device failures and adverse events from the trial will be reported twice.
I think it's very messy, and I personally would keep clinical trial data separate from complaints data. I would also note in my complaints procedure that data from clinical trials are evaluated as clinical data in the clinical trial report and not as compliants.
Others here may have different insights.