COVID-19 BY THE NUMBERS – What does the data really say about Arizona?

Posted at 9:44 PM, Apr 07, 2020
and last updated 2020-04-08 00:46:21-04

By the gold standard COVID-19 model, Arizona is looking pretty good.

The model created by the University of Washington’s Institute for Health Metrics and Evaluation now estimates that the COVID-19 pandemic will probably not cause either a hospital bed or ICU shortage in the state. It also projects the total death rate by August 4th to be 570. While that is still one too many, the same model showed an ICU bed shortage, as well as over 1,500 deaths only 10 days ago.


So what has changed? And what does the model really say about how Arizona is doing in combating the COVID-19 epidemic?

What has changed?

The biggest change is in the model itself. The earliest projections made by the IHME were based on COVID-19 cases in Wuhan City. Shortly after publication, it began to consider data from Italy and Spain as well, however model builders did not have a lot of information on how social distancing would impact the estimates. With the passage of time and more data, the model is now able to account for the impact of social distancing policies in places around the world where viral infections are now thought to have peaked or are beginning to peak.

More recently, and perhaps more important to Arizona, is the model is now including case data that is reported daily from the state health agencies across the country. What this means is that every day at 9 a.m. the Arizona Department of Health reports new cases and deaths due attributable to COVID-19, and that data, as well as data from the other states, are fed into the model to change the direction of the model’s curve.

The model requires four government-mandated social distancing policies for its estimates to be accurate, according to the model builders. A stay at home order, educational facilities being closed, non-essential services being closed, and travel being severely limited. In previous versions of the model, builders had given each of these mandates equal weight. After reviewing the data however, model builders decided to weigh the mandates differently, giving school closures more emphasis in some communities than a stay at home order for example.


What does the model say about how Arizona is combating the virus?

First the good news, according to the model, Arizonans are very likely “flattening the curve.” Even though the state is only rated to have completed two of the four government mandates, the decision by Governor Ducey and Superintendent Kathy Hoffman to implement a statewide closure of schools early seems to have had a strong impact on Arizona’s trajectory.

462020.JPG<div class="Figure-credit" itemprop="author">Garrett Archer

Even if social distancing policies enacted in the state are proving successful, there are some largely unknown variables to consider that also have an impact on the model. The first being how reflective is Arizona’s Department of Health’s COVID-19 reporting of deaths and hospital utilization of COVID-19 patients to the actual numbers? If there is a large, or growing number of people being admitted to hospitals for respiratory illnesses, but do not fall into the state’s testing criteria for COVID-19, the model’s projections would suffer.

There are reasons for Arizonans to be optimistic at this point about their fight against the COVID-19 pandemic. But just as the earliest models of COVID-19’s catastrophic infection rates were met with a healthy degree of skepticism, so should these much rosier projections.