Spontaneous failure of nRF52832

We've got kind of a strange problem. A rather large percentage of the chips experience high current. About 30mA. We can do all the testing, and they are fine. Then a few weeks later, the coin battery fails, and we find it in a high current state. The chip seems to be functional however.

With the semiconductor shortage, we think its possible we could be getting defect or used chips. Hard to say.

Another possibility is software. I read something about the chip getting stuck in high power debug mode, but don't know much about it.

Anyone experienced this before?