What COVID-19 models say about Ariz. now
As Arizona continues to reopen after a monthand-a-half under a stay-at-home order, it’s still not clear if COVID-19 cases and deaths have peaked in the state.
A number of mathematical models suggest different trajectories for Arizona cases and deaths.
Some models seem to show that the worst is over, while others project rises in cases and deaths.
The models examined are from the University of Washington’s Institute for Health Metrics and Evaluation, Covid Act Now, Massachusetts Institute of Technology, Northeastern University and the University of Texas at Austin.
Factors such as a swift increase in testing availability through a recent “testing blitz” and expanded testing criteria also affect how many cases get identified, making it difficult to determine if in
creases in case numbers solely are because of more testing or because of increased spread.
As of Friday, Arizona reported 13,169 total confirmed cases and 651 deaths. Last Friday, Arizona reported 9,707 total confirmed cases and 426 deaths. That means cases increased by 25% and deaths increased by 26% over this past week. Over the past two weeks, cases increased by 65% and deaths increased by 97%.
State officials have been consulting a model created by the Federal Emergency Management Agency, the U.S. Centers for Disease Control and Prevention and the assistant Health and Human Services secretary for preparedness and response. That model is based on projection formulas from the Johns Hopkins University and considers current data and mitigation strategies happening in Arizona.
Gov. Doug Ducey and state health department director Cara Christ have said that the model has been the most accurate and realistic. But state officials have yet to release it to the public, saying they do not yet have permission from the federal government to share it.
A FEMA public affairs officer told The Arizona Republic on Wednesday that the agency would not be releasing the model. A public-records request from The Republic to the state health department for a copy of the model has not yet been fulfilled.
“The model we shared with the Arizona Department of Health Services is not for public distribution and was only provided to support their planning needs,” Brandi Richard Thompson of FEMA wrote. She said that there is a “wide variety” of models and that officials make decisions based on the models and tools they think best support their local needs.
“FEMA and federal partners do not release COVID-19 models and planning support tools publicly because they are preliminary results and not forecasts or predictions,” she wrote.
Has Arizona reached its peak?
Neither Ducey nor Christ could say at a Tuesday news conference whether Arizona had reached a peak in cases.
“Has it peaked or not? Well, I don’t know,” Ducey said. “That’s what we’re doing is preparing for the worst-case scenario.”
Several models suggest that Arizona may be past its peak, defined by either daily death rates or infection spread.
The University of Texas model says statewide there’s an 67% chance the peak has already passed, a 70% chance the peak will pass within one week and an 85% chance the peak will pass within two weeks.
For the Phoenix-Mesa-Chandler area, it projects a 54% chance the peak has already passed, a 61% chance the peak will pass within one week and a 75% chance the peak will pass within two weeks.
For Tucson, there’s a 91% chance the peak has already passed, a 92% chance the peak will pass within one week and a 96% chance the peak will pass within two weeks, according to the model.
The University of Texas model defines peak as the day the daily death rate is projected to stop increasing and begin decreasing.
The IHME model also shows that Arizona is past the peak of infections, which are now on a decline.
But still, if mobility increases too quickly without measures like face masks and social distancing, deaths will balloon, Ali Mokdad, IHME professor of health metrics Sciences, said.
A May 8 forecast by the University of Arizona professor Joe Gerald said the number of newly reported cases continues to rise.
“Because of the lag between infection and detection, this increase is attributable to conditions before easing of social distancing restrictions,” he wrote.
Gerald said it doesn’t appear that Arizona is past its peak. Cases, deaths and hospitalizations are all still increasing, he said.
It may matter symbolically whether Arizona is past the peak or not, which might signal to some that the worst is behind us, but regardless of whether we’ve passed the peak, new cases are being identified every day and there is still considerable transmission in the community, he said.
“Our risk of becoming infected remains high, should we be out and about and not be taking precautions,” Gerald said.
Cases in Arizona may not reach a clear peak and instead stay at a plateau, said Joshua LaBaer, executive director at Arizona State University’s Biodesign Institute. Other places, such as New York City, saw clear peaks.
You really want to see the numbers go down, though, not just level off, he said.
“I don’t know if this state will see that kind of peak because we did start social distancing in a way that New York didn’t do by the time they were in real rapid growth rate. So, hopefully, we won’t see that kind of peak,” he said.
Are Arizona’s cases increasing?
The MIT model shows Arizona’s total detected cases continuing to increase over the next month, up to more than 19,600 by June 15.
But the model projects that Arizona’s active cases will steadily decline from now on. It shows 4,202 active cases on May 21 followed by a steady decrease in active cases down to 3,343 on June 15.
The Northeastern model shows increased numbers of total infections through June 8, when the projection ends, more rapidly increasing growing case trends of weeks past.
The Covid Act Now model for Arizona shows “elevated risk” given reopening metrics. Covid Act Now describes itself on its website as “a multidisciplinary team of technologists, epidemiologists, health experts, and public policy leaders working to provide disease intelligence and data analysis on COVID in the U.S.”
Arizona has a high rate of infection growth with an increasing number of active cases, according to the Covid Act Now model. The model says that on average, each person with COVID-19 is currently infecting 1.18 other people, meaning COVID-19 is growing, but slowly. Arizona’s infection rate has not yet dropped below 1.0, according to the model. The model says that most experts recommend an infection rate of less than 1.0 for two weeks before reopening.
Gerald pointed to the state’s infection growth rate as well, saying social distancing efforts didn’t push the number below 1.0. “Our social distancing has helped and has helped to slow the transmission of the virus, but it hasn’t been sufficient to suppress or extinguish the virus,” he said.
The IHME model shows that daily reported confirmed cases are still increasing but the total estimated infections have started to decrease after peaking around May 2.
The infection has been spreading less quickly over the past week and a half.
Depending on where you look in the state, the trends in cases differ, LaBaer said. In Maricopa County, cases flattened a bit in April, then started to rise again in May, he said. The past week has flattened a bit for the county again, he said.
Other counties, including Yuma and Yavapai, were going down, but have started increasing again, he said, while
Pima County seems to be flattening.
“Certainly for about a month there, definitely the stay-at-home was doing a lot,” LaBaer said. “Where things started to take off again was even before we opened up, and I think that’s just people, probably, my guess is, getting anxious and wanting to get out more on their own even. That, probably plus a mix of probably more aggressive testing. So we’re probably picking up more cases that way.”
The increase in testing alone doesn’t appear to be the sole cause of cases increasing, LaBaer said. Part of the increase likely is due to actual spreading throughout the state, he said.
“I am moderately concerned that the April trend of really flattening things” may be changing, he said. “I’m a little concerned that we’re taking off a little bit. I’m hopeful that in the last week, that that hasn’t panned out. So maybe we’re flattening again. But this trend here worries me a little bit.”
Are Arizona’s deaths increasing?
While the number of deaths publicly reported in the past week has been high, many of those deaths actually occurred much earlier and are just now being added to the state’s database.
The University of Texas model projects Arizona’s daily reported deaths will continue to increase until May 28, when they will start to decline.
But the MIT model shows that total detected deaths will increase steadily from now until June 15, when the total death count is projected to reach 1,155.
The Northeastern model projects a continued steady increase in daily new deaths through June 8, the model’s end.
The Covid Act Now predicts a significant amount of deaths — 39,000 — by Aug. 11 if all restrictions are lifted. Based on current trends, it projects 2,000 total deaths by Aug. 11.
Mokdad of IHME said deaths show what happened three weeks back. This means mortality is expected to continue to increase until around May 20 when it starts to decrease in terms of daily reported deaths.
The model projects a total of 2,871 deaths by Aug. 4, a high number in large part due to Arizonans’ increase in mobility when active infections were still increasing, Mokdad said. “We’re assuming with mobility that mortality would increase,” he said. “The only thing we don’t know is when people in Arizona go out, how well they will do that. Will they wear a mask? Will they give a safe distance from each other or not?”