Yes. If these are ‘background’ deaths then you would expect a flat line (haha). Or at least a line with some noise on it.
However as tando mentioned the data are not being collected either objectively or with any reliability. So the shape of the curve may (for example) reflect inconsistent followups or incorrect date-of-death being recorded.
The fact that doctors are being given leeway to not collect data if they think it’s irrelevant is … well, highly inappropriate. The information that is collected is almost certainly useless if there are any more “features” like this in the methodology.