UAMS Forecast: July 30

[et_pb_section fb_built=”1″ _builder_version=”4.5.2″ _module_preset=”default”][et_pb_row _builder_version=”4.5.2″ _module_preset=”default”][et_pb_column type=”4_4″ _builder_version=”4.5.2″ _module_preset=”default”][et_pb_text _builder_version=”4.5.2″ _module_preset=”default”][/et_pb_text][et_pb_text _builder_version=”4.5.2″ _module_preset=”default”]

Read full report here

The UAMS College of Public Health released its updated forecast on Tuesday, extending out to Aug. 15. All projections are based on Arkansas Department of Health data as of July 30.

In a nutshell:

  • There are no long-term projections in this update as UAMS is adjusting its methods in light of new research on the virus
  • Short term projections predict 55,000 cumulative cases by Aug. 10 and 543 deaths by Aug. 15

The projections

There are two forecasts

  • Through Aug. 10 (includes correctional data)
  • Through Aug. 15 (no correctional data)

Forecast with Correctional Data

Prediction: 55,000 cumulative cases by Aug. 10

The good news is we are falling below projections from this model. To hit the 55,000 mark, we would need 8,707 new cases between today and Aug. 10, or an average of 1,741 new cases a day.

In addition, this model predicted 5,000 new cases between July 27 and Aug. 2. The actual increase was 2,842 between those two dates. We hit the 5,000 case increase on Aug. 5, with the increase taking 9 days instead of the projected 6 days. Either way, that increase is more rapid than anyone should be comfortable with.

This correlates well with the slow-down in new cases I noted last week. That doesn’t mean, however, that the slowing trend will necessarily continue this week. The UAMS report also noted the leveling off of cases in late July.

From the report: “Specifically, over the past week, we have seen roughly the same number of new cases reported each day. This would suggest the curve may be leveling off, albeit at a very high rate of daily new infections. Yet, as this virus has taught us, recent data are not particularly good indicators of long, or even short-term trends.”

Forecast without Correctional Data

Projection through Aug. 15:

  • 52,400 cumulative cases
  • 3,686 cumulative hospitalizations
  • 1,285 cumulative ICU admissions
  • 543 cumulative deaths

How does that square with where we are now?

52,400 Cumulative Cases

I don’t track total cases without including correctional data, so I can’t do a detailed comparison for cases. It does seem high to me, however. We’ve had around 5,000 cases in correctional facilities. If we take those out, it reduces our cumulative cases to around 41,000, which would require more than 11,000 new non-prison cases to reach the projected total in the next 10 days. Again, it seems like our case growth in tracking below these numbers. That’s a good thing.

3,686 Cumulative Hospitalizations

This can look scary since we’re used to seeing hospitalizations as a “point in time” number, and not a cumulative number. We are currently at 3,118 Arkansans ever hospitalized for Covid. To hit the projection, we’d need 568 new hospitalizations over the next 10 days (that includes today). That’s an average of 56.8 new hospital admits per day. Our 7-day rolling average of new admissions is 54, so this seems spot on. 

1,285 Cumulative ICU

This again is a data point I don’t have. I have “point in time” ICU numbers but no cumulative numbers for comparison.

543 Cumulative Deaths

We have currently lost 508 Arkansans to Covid. If we take out the 32 incarcerated population deaths, we are at 476 deaths. I feel a bit uncomfortable not counting any Arkansas life, incarcerated or non-incarcerated, so I’m doing it here for comparison purposes only.

To reach the projection of 543, we’d need 67 new non-correctional deaths in the next 10 days. That’s an average of 7 deaths a day. Our 7-day rolling average is at 10 new deaths per day (including correctional deaths). Given these numbers, we seem likely to hit that target as well.

From this data, we can see that the projections seem to be on target with hospitalizations and deaths, but a bit high on new cases. This could indicate that our deaths and hospitalizations are rising at the same rate they were a few weeks ago, while our new case increase is slowing. 

Since we know that hospitalizations and deaths often lag behind new cases, it could be we’re seeing residuals from the rapid increase in cases a few weeks ago. If this is the case, we should see hospitalizations slow (at least temporarily) soon. Deaths are harder to pinpoint due to delayed reporting and a varied length of illness.

Another possible reason for the disconnect between the trajectory of new cases vs. hospitalizations could be that our testing is not keeping up and we’re subsequently lacking in confirmation of new cases. In this scenario, our hospitalizations and deaths are giving us a truer picture of our current growth.

What about the mask mandate?

I was hoping this report would factor in impacts from the statewide mask mandate, but it looks like that’s coming in the next report. 

“We should expect the number of new cases to continue to increase for the next two weeks. If the executive order for mask wearing in Arkansas is effective as a mitigation measure, it will most likely be reflected in the next two-week report.”

But, I wonder, with the leveling off noted above, if the mandate (along with individual city mandates) isn’t already having an impact. 

Long-term Projections?

There were no long-term projections this week as UAMS adjusts its modeling in light of new research about Covid. New long-term projects should be available in the next update (Aug. 18).

From the report: “There have been a number of changes in our understanding of Covid-19 due to recent reports in the peer reviewed press, and we believe it is time to reassess the assumptions under which the model has been operating. For example, at least one study is reporting there may be as many as 10 to 20 times the number of asymptomatic cases as previously thought. Another study has found the concentration of virus in the upper-respiratory tracts of children is much higher than in adults. This suggests children may be more infectious than previously thought.” 

[/et_pb_text][/et_pb_column][/et_pb_row][/et_pb_section]
Rob Wells

Rob Wells, Ph.D., is an Assistant Professor of Journalism at the University of Arkansas School of Journalism and Strategic Media.

ABOUT ARKANSAS COVID


The University of Arkansas School of Journalism and Strategic Media operates this site as an independent source of news and as a community service for Arkansas residents. Students produce the content here under the supervision of Rob Wells, Ph.D., Assistant Professor of Journalism. The data presented here is collected at roughly the same time each day from the Arkansas Department of Health website.

Questions? Email arkansascovid@gmail.com