Advice : How to deal with (0 and small population) when calculating incidence rates

[Wad Faris]

Hello! I am working on a project about trends in cancer incidence by age, gender, nationality, and region. I faced a probelm that some age groups and regions and years for one nationality is 0., What are the known and easy solutions for zero or small numbers of population,, and if you have a reference for that, please share it with me ..I heared about grouping age groups or regions and also smoothing techniques .. but I am not sure ! Thank you for your help and support

[Carlo Lazzaro]

Wad:
welcome to this forum.
The usual recipe is to group together the levels of ctegorical variables with few observations.
Whether this sound acceptable or not in your research field, I do not know.

[Wad Faris]

ok thank you, but how to group them ? is there any ref

[Carlo Lazzaro]

Wad:
the smallest ones.