Scrape tweets using API in python

I am trying to scrape searching results using twitter api. After each request, the program will sleep 60 seconds and continue. But the problem is that the results of two requests are nearly the same. How to avoid this?