In this study, they did not directly do that. Instead they extrapolated from the population/state level delta prevalence estimates (so whatever the state was doing to do that.. probably sequencing) over different time intervals.