Being an obsessive math nerd, I began tracking the course of COVID19 in my home town (Long Beach) and my state (California) virtually from the beginning of the pandemic. I chose daily new cases in Long Beach and daily deaths in California as my metrics, for no other reason than that the data were pretty easy to find and are pretty reliable indicators of COVID conditions in each population. The California data were extracted from the Worldometer website and the Long Beach data have been published at the Long Beach Health and Human Services website. For both datasets, I am using rolling 7-day averages to smooth out daily reporting anomalies. The figure above shows the data to date plotted side-by-side. Quite coincidentally the vertical scale was comparable for both indicators, so no scaling was required. (There have been 52,659 cases officially reported in Long Beach since the pandemic began as of April 10, 2021, and 60,381 COVID-related deaths in California, remarkably similar numbers.)
The demographics of Long Beach and the state of California are not identical so any conclusions drawn from a direct comparison would not withstand scientific scrutiny. Nevertheless, a superficial comparison of the Long Beach and California data reveals a number of interesting features. When the data are plotted on the same graph, the general shapes of the two curves are remarkably similar with similar peaks and troughs. In addition, the peaks in the CA deaths tend to lag the new case peaks in Long Beach by about 3 weeks and are likely associated with common holiday events.
The first major peak in cases in LB occurred on July 9, 2020, which was likely a result of opening of businesses and public areas after Memorial Day. The time lag between Memorial day and the spike in Long Beach cases was about 40 days, which may reflect the time required for the newly infected population to return home and infect their peers and then allow the onset of reportable symptoms. This timeline is complicated by the fact that the period of increased exposure was protracted, unlike a well defined superspreader event that is easier to correlate with a rise in cases. The surge in cases was followed by a peak in CA deaths on August 3, about 25 days later.
Then in December and January, two peaks in new cases and in deaths occurred. It’s easy to argue that these peaks are correlated with the Thanksgiving and Christmas holiday seasons. The first peak in cases occurred in Long Beach on December 20, 25 days after Thanksgiving. Then the first peak in holiday deaths in CA occurred on January 12, 23 days after the spike in cases. The second peak in LB cases occurred on January 5 (11 days after Christmas) and the second peak in deaths occurred on January 24, just 19 days later. It seems likely that both surges are associated with increased travel and more social interaction during the Thanksgiving and Christmas holidays, including the travel periods leading up to the holidays. While not scientific, this seems like a reasonable timeline to describe the progress of the disease.
The peaks in both datasets are distinct, suggesting that the Thanksgiving and Christmas events were separate and distinct drivers. Even the relative strengths of the Thanksgiving and Christmas events appear to track, with the peak Christmas rate exceeding the Thanksgiving rate by 7 percent for cases and by 13.5 percent for deaths.
So here’s the best news. Both the new case rate in LB and the death rate in CA appear to have plummeted since the January peaks. This is likely due to the absence of any major holidays in the interim combined with the rise in vaccinations. Here’s to hoping that those trends continue.
So please, everybody — get your vaccinations, maintain social distancing, wear your masks, wash your hands, let’s put this horrible plague behind us.