Three of the deaths happened overnight, which is alarming for a disease whose death rate *may* be as low as the flu’s. Combined with three previous fatalities in the state, there are now six people dead among a total of 18 who have tested positive in King County, Washington.

Six out of 18? The death rate is … 33 percent?

No, it isn’t. You already know why if you read this post yesterday. Clearly there are many cases of coronavirus in Washington state right now that remain momentarily undetected, a spread that likely began with a man who returned from China in January, got sick, and then made a full recovery. Genetic testing of his strain of the disease and the strain that killed a different patient within the past few days strongly suggests that that strain has been circulating among the population for the past six weeks. One model of the total number of infections given what we know about how contagious COVID-19 is estimated it at anywhere between 150 and 1,500. If China’s calculations about the disease’s death rate — 2% — are accurate, six deaths would lead us to expect 300 infections locally, right in the range scientists expect. Which would still be … not great, since a 2% death rate is ominous for a highly contagious illness.

If instead the disease’s death rate is more like the common flu’s — 0.2% — then we’d expect 3,000 infections locally. That would be a surprisingly large amount, well beyond the current estimate. Which means either many more people are sick (perhaps mildly sick) than we expect or that coronavirus really is more lethal than the regular ol’ flu, with a death rate higher than 0.2%.

But wait. There are other factors to consider, starting with this one from former FDA commissioner Scott Gottlieb:

It may be that the death rate is flu-like across a sample of the total population of infected people but obviously it’ll be much higher if the sample consists mainly of patients who are far gone. If you try to calculate the death rate of the flu based only on the cases you find in the local ICU, the number you get will be much higher than 0.2%, obviously.

There’s another important complicating factor in the latest Washington cases:

A male in his 50s, hospitalized at Highline Hospital. No known exposures. He is in stable but critical condition. He had no underlying health conditions.

A male in his 70s, a resident of LifeCare, hospitalized at EvergreenHealth in Kirkland. The man had underlying health conditions, and died 3/1/20

A female in her 70s, a resident of LifeCare, hospitalized at EvergreenHealth in Kirkland. The woman had underlying health conditions, and died 3/1/20

A female in her 80s, a resident of LifeCare, was hospitalized at EvergreenHealth. She is in critical condition.

In addition, a woman in her 80s, who was already reported as in critical condition at Evergreen, died. She died on 3/1/20

No patients younger than 50, all three of the deceased 70 or older, two with underlying health problems. That’s exactly what we’d expect based on the information from China, where the death rate for sick and elderly patients is something on the order of 15 percent per some estimates. I actually flagged the outbreak at the LifeCare Center nursing home in yesterday’s post as an example of a worst-case scenario for a localized outbreak precisely because senior citizens are the people most at risk of being killed by coronavirus. According to yesterday’s number, no fewer than 27 residents (and 25 staff) at the home were showing symptoms. I can only assume there’ll be more deaths from that group this week.

But needless to say, an outbreak within a small population of the most vulnerable age group, many of whom had health complications to begin with, isn’t going to tell us a lot about the death rate across the general population. When you factor in Gottlieb’s point that the patients closest to death are apt to be tested first, the numbers nationwide this week may not tell us much of anything at all. Except, of course, that the disease is already here in much greater numbers than we thought, which is already a very safe assumption.

Relatedly, New York City diagnosed its first case of COVID-19 last night but there are doubtless plenty of other hidden ones that have gone undetected so far due to the CDC’s colossal screw-up in testing. That’s about to change:

Is Alex Azar’s employment status about to change too? Politico claims that fingers are being pointed at him within the administration for not being proactive enough in resolving the CDC’s testing snafu more expeditiously:

“This was a management failure,” said one administration official, charging that Azar didn’t adequately plan for a worst-case coronavirus scenario that’s grown more likely by the day — even though Azar touted his bona fides as a veteran of the George W. Bush administration, where he helped fight crises like SARS and an anthrax scare. “CDC and FDA should have been working hand-in-hand to get Plan B, Plan C and Plan D ready to go,” the official said.

“The administration’s response has been reactive, not proactive,” added a former HHS official. “A lot of what has happened has been driven by outside pressure,” like public health labs sounding the alarm that they were unable to perform the CDC’s tests.

I don’t think Trump will fire him, if only because it might add to public anxiety to have him decapitate HHS in the middle of a crisis. Besides, he’s already demoted Azar on the coronavirus response by placing Mike Pence at the head of the task force formed to contain the disease.

Exit question: What do we think of this advice from the surgeon general?

“You can increase your risk of getting it by wearing a mask if you are not a health care provider,” he told Fox & Friends on Monday morning, reasoning that having to adjust a mask gives you more reason to touch your face. The masks apparently are useful not so much in protecting the wearer from contracting the disease but in preventing them from giving it to others (e.g., by projecting droplets containing the disease outward via coughing), which is why authorities want to make sure front-line health-care workers who are screening people have a ready supply. For most of the public I think they’re a psychological security blanket, something that doesn’t do much beyond provide a sense of comfort just in case someone comes up to you and, uh, sneezes in your face.