If I'm not mistaken, the rate in the paper is bigger, 0.25% against 22k/11M = 0.20% in Belgium.

But if the newspaper I read are right, there is a catch: when tests were not available, Belgium counted as a covid death every case where it could be suspected. Even nowadays, in case of comorbidity, covid is always written down as the cause of death. So 22k is probably a maximum, and, taking into accound the 25% uncertainty on the denominator, the real rate could very different.