3000 people in one month, from a disease that exhibits exponential spread through the population.
Usually if you look at death by "date of death" you see Gompertz-Functions all around. If you look at death by "reporting date" you often see a very uneven lag forming some exponentials...