A tantalizing data point for the central mystery of the epidemic: How many people have had the disease and how many of those never had symptoms?

In mid-March doctors at New York–Presbyterian Allen Hospital diagnosed one of their obstetric patients with COVID-19. So they changed their admission protocol: Beginning on March 22, they would ask all women admitted for childbirth to submit to a coronavirus screening. They tested more than 200 patients over a two-week span.

The best estimates right now of how many Americans are infected are in the range of one to five percent. The numbers at New York-Presbyterian were conspicuously higher than that.

Four women (1.9%) had fever or other symptoms of Covid-19 on admission, and all 4 women tested positive for SARS-CoV-2 (Figure 1). Of the 211 women without symptoms, all were afebrile on admission. Nasopharyngeal swabs were obtained from 210 of the 211 women (99.5%) who did not have symptoms of Covid-19; of these women, 29 (13.7%) were positive for SARS-CoV-2. Thus, 29 of the 33 patients who were positive for SARS-CoV-2 at admission (87.9%) had no symptoms of Covid-19 at presentation.

Of the 29 women who had been asymptomatic but who were positive for SARS-CoV-2 on admission, fever developed in 3 (10%) before postpartum discharge (median length of stay, 2 days). Two of these patients received antibiotics for presumed endomyometritis (although 1 patient did not have localizing symptoms), and 1 patient was presumed to be febrile due to Covid-19 and received supportive care. One patient with a swab that was negative for SARS-CoV-2 on admission became symptomatic postpartum; repeat SARS-CoV-2 testing 3 days after the initial test was positive.

Scott Gottlieb draws the conclusion that although five percent or less of the overall U.S. population might be infected, it stands to reason that the population of a major hot spot like NYC would have a much higher rate:

Fifteen percent also happens to be the number of residents in the German hot spot of Gangelt who were found recently to have coronavirus antibodies in their blood. Fifteen percent was also the best guess of how many people in Spain might have had the virus as of late last month according to a model published a few weeks ago. This may be a realistic infection rate for populations that have recently had the virus sweep through their communities.

Another bit of circumstantial evidence is how many members of the NYPD have been out sick lately:

On Monday, 6,974 uniformed members of the NYPD were out sick, accounting for 19.3% of the Department’s uniformed workforce, according to the report. That number has jumped from 12% on March 28.

Currently, 1,935 uniformed members and 293 civilian members tested positive for the coronavirus, the NYPD said.

Those were the numbers as of a week ago, at a moment when cases in New York were (apparently) beginning to peak. Needless to say, not every cop who called in sick did so because of COVID-19; on the other hand, given that some segment of the population is walking around with the disease without symptoms, not every cop who showed up to work was free of the virus. It stands to reason that police might have higher rates of infection than the average citizen because of how many people they’re in contact with every day, often in close quarters to make arrests. So we wouldn’t expect the prevalence among New Yorkers generally to match that within the NYPD.

But if, say, 20-25 percent of cops have been infected, might 15 percent of New York City overall have been infected? Hmmmm.

Back to the obstetric study. It’s possible that pregnant women have higher rates of infection than the local general population due to factors that are specific to pregnancy. The Daily Wire points to this recent article from UCSF suggesting that pregnant women are more likely to contract the disease because they’re immunosuppressed. But the official guidance right now from Britain’s Royal College of Obstetricians and Gynecologists is that “Based on the evidence we have so far, pregnant women are still no more likely to contract coronavirus than the general population.” If that’s true, then nothing biological is happening to make them more susceptible.

But could their environments make them more susceptible? Just as cops are probably more at risk due to the physical requirements of their job, it may be that pregnant women are more at risk because they’re visiting doctor’s offices more often than the average person is in the months and weeks leading up to giving birth. Maybe they picked up COVID-19 in the waiting room. Or maybe their doctor picked it up from a colleague and unwittingly passed it along in the exam room. Another great mystery of the epidemic is what percentage of New York doctors and nurses, specifically, might be infected right now after six weeks of being face to face for hours every day with coronavirus patients.

Another question. How many of the 89 percent of asymptomatic infected women admitted to give birth at New York-Presbyterian *remained* asymptomatic for days or weeks after being discharged? The doctors note that three became symptomatic before being discharged from the hospital but we’re left to wonder what happened to the other 26. If they ended up never showing symptoms, that would suggest that many (most?) people who test positive without symptoms never go on to develop any. If they did end up showing symptoms later and were merely “pre-symptomatic” while in the delivery room, that would imply that we have less to fear from asymptomatic carriers spreading the disease for many days on end while they’re out and about among the population. The doctors should follow up.

Exit question: If the 15 percent figure is borne out by antibody testing of a random sample of New Yorkers, is that good news or bad news? If my math is right then that would amount to a true fatality rate of around 0.5 percent, as there have been 6,182 official COVID-19 deaths in NYC thus far and the estimated total infected population would be around 1.2 million people. (Major caveat, though: There’s good reason to believe the official death count is missing many people who died of the disease because they were never tested, which means the fatality rate is higher than we’ve calculated.) On the other hand, even after a month of hell, NYC still wouldn’t be close to herd immunity of 50-66 percent. We’d be a long way off from the rosy projections of the Oxford model that this epidemic will be over soon-ish.