 This paper presents a new method called COVID near-term which uses an autoregressive model and a parametric bootstrap approach to accurately predict hospitalizations within two to four weeks of the initial prediction. The authors compared their results with those of the California COVID Assessment Tool, CalCAT, and found that the predictions made by COVID near-term were more accurate than those of CalCAT. This suggests that this method could be used to provide more accurate estimates of future hospitalizations due to COVID-19. This article was authored by Adam B. Olsson, Riyadh Nagarsia, Christopher I. Kaphan, and others.