Several papers have looked at the mortality rate from Covid-19 as a function of the infection rate. The slope, a measure of the relationship between the two rates, varies depending on (among other things) the population being studied, the level of supportive care available and the currently circulating variants.

Suppose you have a dataset from adults 18-55 in 50 different US cities in June 2020. When you build a simple linear regression model to look at the number of covid-19 deaths (Y) as a function of the number of cases (X), you get an estimated linear model Y = 0.12 * X.

Based on this model, if city A had 100 more cases in June 2020 than city B, how many more deaths would you predict in city A than in city B?