It could be due to random perturbations in the small data size. Like if 10 people died last year, and 14 people died this year, deaths are up 40%, but the 4 ext

It could be due to random perturbations in the small data size. Like if 10 people died last year, and 14 people died this year, deaths are up 40%, but the 4 extra deaths could be random. You'd have to look at a long trend to infer meaning from such small data. But I'm not sure how big the data is in this case. Something similar happened in my town when I saw a huge-looking spike in a graph of covid deaths. But when I looked closer, deaths had spiked from 0 to 3 on a particular day, and the average is 0.3/day. So that crazy looking spike is probably meaningless over a longer time window.