Table 2. The number of COVID-19 deaths reported by each U.S. state and the District of Columbia as of May 5, 2020, according to the COVID Tracking Project's spreadsheet.
By contrast, the total number of deaths predicted by the IHME to happen in the United States by July 30, 2020 was 81,111. This means that, nationally, 62,698/81,111 = about 77.3% of the cumulative number of COVID-19 deaths in the United States predicted to happen by July 30, 2020 had already happened by May 5, 2020.
Of the 50 states and the District of Columbia (n = 51), 8 of them (7 states--Maryland, Massachusetts, Connecticut, Illinois, New Jersey, New York, and Rhode Island--and the District of Columbia) had already surpassed the best-estimate prediction of the cumulative number of deaths. Clearly, given that they had surpassed these predictions by May 5 even though the predicted numbers were not supposed to be reached until July 30, the prediction was significantly too low in these regions. In other words, 8/51 = about 15.7% of the included regions had surpassed the cumulative predicted number of deaths almost three months earlier than expected. Of these 8 regions, 2 of them (25% of the states that had already surpassed the predicted number of deaths, and 2/51 = 3.9% of the total) had surpassed even the IHME's upper bound. In total, of the 51 included regions in the current study, 24 of them (24/51 = 47.1%) had surpassed the IHME's lower bound. These 24 included the aforementioned 8 regions that had surpassed the IHME's best-estimate prediction, as well as 16 that had not. The remaining 27 regions (all states) had not yet surpassed even the IHME's lower bound. Figure 1, below, illustrates this pattern of results in a map.