Figure 2a shows a sharp decline in transmission among children (dotted) over the past 2 weeks with the smallest gains being observed among the elderly (solid). While difficult to draw causal conclusions, recent mask mandates in certain districts, aggressive case identification, and limited classroom closures are likely helping as the steepest declines were among school-age children 5 – 19 years (Figure 2b). However, the reversal among adolescents may be somewhat attenuated once the impacts of university re-openings are fully realized (https://www.nature.com/articles/s41467-021-25169-3).

Comparing school reopening this year (in-person) with last year (virtual), levels of community transmission are much higher this year than last, particularly among those <15 years (Figure 2c). Higher transmission among those 15 – 24 years associated with university re-openings was first observed about this time last year (orange bars, ages 15 – 24). It is hard to believe how good conditions were this time last year – without reinforcing how impactful good better public policy and high better levels of individual cooperation can be.

Over the past 3 weeks, the trajectory of the current outbreak and that of the winter 2020/2021 outbreak have meaningfully diverged (Figure 2d). While there have been temporary reversals before, it appears we are now experiencing a more permanent shift away from last year’s trajectory.

Test positivity among those undergoing traditional nasopharyngeal testing unexpectedly increased this week from to 19% to 21% (Figure 3). This level of positivity indicates that testing is inadequate for optimal public health practice and that many cases are going undiagnosed. Prior troughs in test positivity have preceded increasing transmission rates, so it is possible improvements could slow next week perhaps due to the impact of Labor Day. Alternatively, fewer people (i.e., the most symptomatic) may have been willing to test over the holiday.

As of September 8th, 2061 (24%) of Arizona’s 8743 general ward beds were occupied by COVID-19 patients, a 0% decrease from last week’s 2070 occupied beds (Figure 4 and Figure 5 Panel A). Another 689 (8%) beds remained available for use. The number of available beds is somewhat lower than last week’s 737 beds.

As of September 8^th, 561 (32%) of Arizona’s 1749 ICU beds were occupied by COVID-19 patients, a 2% increase from last week’s 549 occupied beds (Figure 4 and Figure 5 Panel B). An additional 143 (8%) ICU beds remained available for use, about the same as last week’s 139 beds.

As a lagging indicator, hospital occupancy appears to be plateauing as expected following declining case rates.

Arizona hospital occupancy remains above seasonal levels with safety margins, as measured by excess available beds, remaining quite low (Figure 6). Available capacity is essentially as low as it was at the height of the winter 2020/2021 surge. Although increases in occupancy are moderating soon in response to declining community transmission, hospitals should be prepared for >20% ward occupancy and >25% ICU occupancy for several more weeks. Even if you are vaccinated, the tail of the COVID-19 surge will still impact your access to healthcare services should you become ill for quite some time.

Arizona continues to experience >100 COVID-19 deaths per week (Figure 7). The week ending August 15th has recorded 180 deaths so far with the week ending August 22nd close behind with 178 deaths. Even so, the rise in the number of deaths is more muted than expected given the rise in cases. It now appears unlikely that Arizona will reach 300 deaths per week. Instead, the peak should occur at 200 deaths per week. This indicates that the case fatality rate has declined from its historical average of 1.5% to a new base rate nearing 1.0%. This is likely attributable to lower severity among younger cases and breakthrough cases.

For the week ending September 5th, 1792 Pima County residents were diagnosed with COVID-19 (Figure 8). This marks the 2^nd week of declining case counts although the decline is more modest than the state trend.

New cases are being diagnosed at a rate of 174 cases per 100K residents per week and is essentially holding steady. Trends across the various age groups appear in Figure 9.

The Pima County comparison of fall 2020 with fall 2021 show a more dramatic signal of the impact of university reopening on community transmission. Case rates among those 15 – 24 years of age were actually higher this time last year. It is important to acknowledge though that there is considerably less testing on university campuses this year than last. Nevertheless, being on the inside where the sausage is being made, I don’t sense there is as much transmission on campus this year as last. Many of our students have either been vaccinated or previously recovered providing at least some buffer against transmission.