COVID-19 Surveillance Updates in US Metropolitan Areas: Dynamic Panel Data Modeling (Preprint)

Abstract

BACKGROUND COVID-19 cases in U.S. metropolitan areas fell dramatically in January and early February, but flattened out in the last two weeks of February and first week of March. The reopening of states and municipalities coupled with the emergence of SARS-Cov-2 variants raises the specter of a re-ignition of explosive growth. Vigilant surveillance can help identify any re-ignition and validate an early and strong public health policy response. OBJECTIVE This surveillance reports aimed to provide up to date information about a potential re-ignition of the pandemic using the novel metrics of speed, acceleration, jerk, and 7-day persistence. METHODS COVID-19 pandemic dynamics for the 25 largest U.S. metropolitan areas were analyzed through 3/7/2021 using the novel metrics calculated on the basis of observed data on the cumulative number of cases as reported in usafacts.org. Statistical analysis was conducted using dynamic panel data models estimated with Arellano-Bond regression techniques. Results are presented in tabular and graphic forms for visual interpretation. RESULTS On average, speed in the 25 largest U.S. metropolitan areas declined from 68 new cases per day per 100,000 population during the week of 1/4-1/10/21 to 20 during the week of 3/1-3/7/2021. However, the decline stagnated and speed dropped only one case per day over the past two weeks from a value of 21 during the week of 2/15-2/21/2021. This stagnating decline is confirmed by acceleration and jerk data. Houston exhibited a smaller than average overall decline in speed and a bounce off its low speed of 17 during the week of 2/15-2/21/2021 to 27 during the week of 2/22-2/28-2021 and 29 during the week of 3/1/-3/7/2011. CONCLUSIONS The stagnation is evidence of the persistence of the pandemic and the possibility of a surge in new cases and possibly explosive growth as states reopen and if people choose not to follow recommended guidelines including social distancing and face mask wearing. Our evidence that Houston showed signs of a bounce during the last week of February and the first week of March coupled with the presence of all the major SARS-Cov-2 variants in the metropolitan area strongly suggests that reopening will lead to an upsurge in Houston’s COVID-19 cases with the potential for re-igniting exponential growth.

Implementation and Comparison of Several Register Transfer Level Simulation Engines for the HDLRuby Language

“…As soon as a value does not include any undefined bit (i.e., Z or X) and it is small enough to match primitive C types (e.g., long long), it is converted to the numeric representation. Forth, list or array-based data structures like the bit strings, the lists of signals, or the list of behaviors to process are all implemented by dynamic arrays (15) that are compact, fast to access, fast to append to, and require a very little number of allocations compared to both static arrays and linked lists.…”