Covid Diary: and now, the entire logistic growth curve

One interesting and horrifying feature of the current phase of the Covid-19 epidemic in the US (“the end of the beginning”) is how our ability to track infections has surpassed our ability to control them.   Unlike in March, people are now able to get tested essentially on demand, without first meeting some restrictive set of  criteria meant to conserve testing facilities and materials.  Although testing has once again become scarce in places like Florida, Texas, and Arizona (where long lines at testing sites are certainly preventing many from being tested), outbreak surveillance is just about the only bright-spot in the American response to Covid.

When New York City was the center of the epidemic in the U.S. most people were only tested when they were very sick and in the hospital.  This meant that reported infections were among people far older, sicker, and more likely to die than the entire infected population.  This tendency of under-reporting (an understatement) was worsened by the overcrowding experienced in some NYC hospitals.  Many patients (even among the very ill) were told not to come to the hospital, and were likely never tested.  As a result, infection numbers in NYC did not come close to estimating the true number of infections.

We also missed the beginning of the growth curve.  It turns out that infections in NYC began around early February at the latest, and they were growing quite rapidly by the end of the month.  However, because we were not testing for community transmission at that time (only testing those with a very specific profile of international travel or contact with someone who did), infection counts trailed the beginning of the outbreak by weeks.

This may give us some insight into our current experience.  We are seeing dramatic increases in infections across the country, and especially in Texas and Florida.  Luckily, infections are currently dominated by younger people and death rates have only recently started to increase (and not dramatically).  We probably shouldn’t expect this trend to continue.

It shouldn’t be all that surprising that the most socially active population is infected first.  Given the messaging about risk-factors for Covid, the gap between older and younger people in frequency of social interactions is likely even larger than normal.  However, given enough community spread the coronavirus will come into contact with older community members, and death rates will once again increase.  However, this time the infection is spread across a far larger region of the country and there will be no complete shutdown on the scale of what was seen in March to stop the outbreak.  We should expect to see death counts that surpass what was seen in NYC.