A July 3 summary of forecasts for the spread of the novel coronavirus and the disease it causes, COVID-19, has been posted online for folks who would like to take a deep dive into predictive data.
Two takeaways from the modeling: How much is wearing a cloth mask worth? Keeping 63 people alive and 14,000 people a day virus-free this fall. Why are we seeing the upswing in cases? Because we’re not social distancing the way we were when businesses and bars were closed and we were sheltering in place.
The models, which use Arkansas-specific data to predict what is coming, change as the real numbers do, and there is some good news, despite the recent rise in cases. For example: The UAMS/UA COVID-19 research team predicts a rate of transmission at 2.41 for every person infected, down from predictions made mid-June. It also appears that at the peak — now pushed to Oct. 30 — the forecast for hospital beds, including ICU — will be below the state’s capacity.
As of July 2, the short-term models, which are more accurate than long-term models, predicted the state’s cumulative number of infections will reach 35,000 by July 10 (the “best-fit” model) or 27,000 (the “time-series” model). (According to numbers presented today, the total case numbers stand at 24,512. That doesn’t represent today’s actual number of cases, because it is based on results from tests that were taken over the past several days. Commercial labs are running as much as nine days behind because of increased testing nationally.)
The less accurate long-term forecast puts the peak of cumulative cases (150,000) at around Oct. 30. It also predicts a peak of 2,794 hospitalizations, 838 intensive care cases and 586 ventilations at that point.
Simulations show a daily new infection rate of 20,000 at the end of September “if conditions do not change” — that is, if there is still the same amount of resistance to wearing masks in public. It also assumes that colleges and schools reopen. “With almost complete compliance with mask-wearing in public, the simulation is showing significantly few daily infections, around 6,000. A similar pattern emerges with respect to daily deaths.” The simulations show the death rate at 75 per day by Sept. 24 if conditions do not change; about 12 per day if stricter precautions are taken.
The “time-series” models are based on community infection information gleaned at the first 25 days of collection, 75 days and current data. The 75-days out model predicted the state would have 8,000 total cases. That’s way out of whack with today’s totals because it was based on a presumption of stricter safety measures.
“Following the loosening of social distancing measures, we see a large increase in confirmed cases that deviates significantly from trend,” the report says of the total case data. The prediction of 27,000 cases by July 13 is more than three times greater than the earlier prediction.
However, the time-series model for deaths is only somewhat higher than the 75-day data predicted. The model now puts total deaths at around 370 (as I read the model) by July 13. A total of 361 deaths were reported at today’s update by the governor.
“The increase in actual and predicted cases is due to the relaxing of social distancing regulations,” the report states. “Previously, we thought the number of increased cases was associated with increased testing. As [the time-series models show], predicted deaths from covid-19 have increased off-trend as of this reporting period, suggesting a real increase in coronavirus infections and resulting deaths.”