voice over focus delay

Has anyone come across issues with voice over where after you programatically shift focus from one element to another there is a long delay (5-10 seconds) before voice 'realises' a new element has gained focus and reads it out? It's specifically happening when the element that previously had focus is removed from the view and almost immediately a new element is added to the view which is given focus programatically.

I've tested the same functionality with other screen readers (NVDA and windows narrator) and couldn't reproduce the issue so it seems to be voiceover specific.

Appreciate any help or ideas.