Will the Current USA Pandemic Incidence Continue to Decline?

 Summary

The current decline in new cases at the rate of 2% per day is not the first decline.  The January –February decline at the rate of 2.5% ended in stagnation in March.  The July-August decline at the rate of 1.5% ended in the great spike which began in mid-September.  Can these past declines help us predict what will happen in the rest of this calendar year?

Prevalence refers to proportion of persons who have a condition at or during a particular time period, whereas incidence refers to the proportion or rate of persons who develop a condition during a particular time period. All rates of new cases per day, both projected and actual, are an average of the last seven days.

The Current USA Pandemic Incidence Decline which began on April 10

On April 10 the national number of new cases per day began to decline. About two weeks later I decided the rate of two percent per day best fitted the decline. So in the following chart the projected estimates are given, along with the actual incidence for that day, and then the difference. Positive numbers indicate that the incidence exceeded the projection.  As can been seen below the actual incidence on May 15 (average of preceding seven days) was 302 cases below the projected number.  Typically when there is a smooth decline the smoothed daily variability has been within about five thousand cases on either side of the projected figure.

 

APRIL 10 (CURRENT) PANDEMIC DECLINE
Pandemic Incidence Decline Projection  at Rate of 2% per Day
Date	Projected	Actual	Difference
April 10	73193	73193	0
May 1	47887	52747	+2835
May 15	36090	35788	-302
June 1	25599
June 15	19293
July 4	13143
July 15	10524
August 1	7465
August 15	5626
September 5	3581
September 15	3007
October 1	2221
October 15	1640
November 1	1164
November 15	877

 

The projection would bring incidence down to 25599 on June 1, then 13143 on July 4,  then 7465 on August 1 and then 3581 on Labor Day, September 5.  Ultimately it predicts below a thousand new cases a day by Thanksgiving.  So unless we are in a hot spot or a large gathering, we should be reasonably safe on July 4th, mostly safe by Labor Day and very safe by Thanksgiving. 

How likely is this to happen?   Well our last major decline which began in early January stagnated in early March. Our previous major decline which began on July 25, 2020 reversed on September 11, 2020 into the largest surge of the pandemic.

The largest and most rapid decline which began on January 9, 2021

On January 9, 2021 the national incident rate per day was 256954 new cases. A decline began which a few weeks later I estimated would be at the rate of 2.5 percent per day. The projections worked well until after March 6. There were in fact 61833 cases on that day, 409 cases lower than the predicted 62222.  By April 1 the incidence was climbing again with a new case rate of 63914, more than thirty thousand cases above the projected 32215 cases for that date.  Cases continued to climb erratically until April 10 when the next firm decline started.

JANUARY FEBRUARY PANDEMIC DECLINE
Pandemic Decline at Rate of 2.5% per Day
Date	Projected	Actual	Difference
January 9	256954	256964	0
February 1	143480	145073	+1592
February 15	100660	85593	-15067
March 1	70619	66681	-3938
March 6	62222	61833	-409
March 15	49543	55030	+5487
April 1	32215	63914	+31698
April 10	25651	73193	+47541
May 1	14696	49765	+35068
May 15	10575	35788	+25213
June 1	6876
June 15	4824
July 4	2982
July 15	2257
August 1	1468
August 15	1030
September 5	605

 

Why did the rate level stagnate with a slow erratic climb?  Most likely this was due to new faster spreading variants.  When the decline did resume it was at a 2.0% rate rather than the previous 2.5 % rate. However other factors (vaccination rate, mask wearing, social distancing, indoor vs. outdoor activity, and size of gatherings) likely played some role. 

What would have happened if the 2.5 percent per day downward trend had continued?  Cases would have been at 25651 instead of 73193 on April 10, then reached 15073 per day on May 1, and 6876 on June 1, and finally 2982 on July 4.   In other words the whole downward trend got moved back by two to three months during the stagnation period.  

The Decline that began on July 25, 2020

Back on July 25, 2020 there were 67567 cases per day. Then a decline started at the rate of 1.5 percent per day that continued until September 11^th when the predicted number of new cases was to be at 32709. It was actually 31799 cases, almost a thousand lower.  What happened after that day is that everything reversed and new cases began to climb rapidly upward.

AUGUST 2020 PANDEMIC DECLINE
Pandemic Decline at Rate of 1.5% per Day
Date	Projected	Actual	Difference
July 25	67567	67567	0
August 15	49192	52901	+3709
Sept 1	38046	41231	+3184
September 11	32709	31799	-911
October 1	24177	41500	+17323
October 15	19566	53066	+33500
November 2	14906	87168	+72262

 

What would have happened had the downward trend at the rate of 1.5 percent continued?   It would have reach 24177 by October 1, and then 14906 by election eve. Trump would have been saying all during October that the virus was going away even without the vaccine.  I think he would have been re-elected.  Shortly afterward all the announcements of the vaccines would have been made.  He would have said, “Look not only is the virus going away, but we have the vaccines to keep it away.”  If the Republicans had won the House and Senate Trump would be talking about “president for life.” 

CONCLUSIONS

It is interesting that a downward trend of 1.5 percent occurred without a vaccine. The downward trend accelerated to 2.5 percent with the vaccine, but came back to 2.0 percent with the change in variants. Behavior, vaccines, and virus variants are each important in controlling the virus.

We should wary of attributing the 2.5 percent decline to the vaccine alone; up to 1.5 percent of that may have been due to behavior.  Similarly the plateau between the 2.5 percent decline and the 2.0 percent decline may have been because of behavioral changes as much as new variants.

Our predictions for what is next should include both the extreme cases of a continued 2.0 percent decline due mostly to continued vaccine rollout, and another flat plateau due to relaxation of behavioral control.  Mostly likely it will be something in between. I have calculated three additional scenarios at 1.5 percent, 1.0 percent, and 0.5 percent depending on the relative mix of vaccine rollout, variants, and behavioral control.  

PROJECTIONS

In the chart below there are five projections: The April 10 base at two percent, the May 15 base at 1.5 percent, the May 15 base at one percent, the May 15 base at 0.5 percent and the May 15 base at zero percent (i.e. a plateau).

FUTURE PANDEMIC DECLINE PROJECTIONS
Pandemic Incidence Decline Projection at Rate of
Date	April 10 2%	May 15 1.5%	May 15 1%	May 15 0.5%	May 15 0%
May 15	36090	35788	35788	35788	35788
June 1	25599	27679	30167	32865	35788
June 15	19293	22401	26208	30637	35788
July 4	13143	16809	21652	27854	35788
July 15	10524	14235	19386	26360	35788
August 1	7465	11009	16341	24207	35788
August 15	5626	8910	14196	22566	35788
September 5	3607	6891	11380	20723	35788
September 15	3007	5577	10396	19319	35788
October 1	2221	4379	8852	17830	35788
October 15	1640	3544	7690	16621	35788
November 1	1164	2741	6482	15264	35788
November 15	877	2218	5631	14229	35788

 

PREDICTIONS

The extreme projections are unlikely to happen.  We will not likely be able to maintain the present two percent per day decline as restrictions vanish, and it is unlikely that the virus will be able to maintain its rate of infection as vaccinations increase even if the rate of vaccination is slow. Since it is unlikely that behavioral restrictions will be increased, the rate of decline will likely depend upon the rate of vaccinations.  If we make a strong effort to vaccinate people as soon as possible we will be closer to the two percent column, but if vaccinations stagnate we will be closer to the zero percent column.  We should try to be under ten thousand new cases a day before September.  We certainly don’t want the rate to be above ten thousand new cases when we get into the Thanksgiving and Christmas holidays.