COVID19: Perspective From A (Sort Of...) Expert-Adjacent Person

OK folks. I have been posting less here even before this blew up (fatherhood and increasing work responsibilities; basically all I’ve been doing lately is posting in my training log) and work has been extra nuts for reasons I’ll explain shortly. But, I decided that I should probably post something here about what’s going on from my perspective.

My bona fides: I have a bachelor’s degree in statistics, a Master’s degree in statistics, and a doctoral degree in epidemiology. I have been on faculty in the School of Medicine for almost seven years, have co-authored over 100 publications in medical journals, and presently work as a statistician on seven randomized controlled trials (across several different medical specialties). I do not work directly on outbreak investigations or infectious diseases, but do work with/on a lot of intensive-care and critical-care studies (where many COVID-19 patients are going to end up once they develop acute respiratory distress syndrome) and have graduate-school friends that work for the government on infectious disease epidemiology.

So, basically, I’m not directly an expert on COVID19 virology or epidemiology, but I’m about as informed of a layperson as can be in this situation. In this thread I will try to discuss what we know about the disease so far and its effects on our health system(s) to the extent that I can. I will not discuss economics, government, or social issues unless they have a direct impact on a relevant tangent. I can’t stop people from doing that, of course, but simply commenting that it is very far outside my expertise and probably best reserved for the other thread if we want to keep this a cleaner sheet of “what do we know about this disease, its spread, treatment, and prognosis of patients as well as health systems” which is probably the best use of this thread.

To be continued…


Very much looking forward to following along.

I realize you probably have a significant amount of introductory/overview material you’d like to present before answering questions or getting out into the weeds, but…I’m very curious what the models indicate about infection dynamics as a function of 1) population density, and 2) socioeconomic status. My hunch is that lower-density areas (ie, smaller towns) as well as zones with low SES (ie, poorer towns) can expect a flatter curve. My question is, does the evidence from this (or previous?) pandemics bear this out?

tl;dr Am I engaging in false hope to think the hospitals in my relatively small (60K) and relatively poor town may be able to ride this out without going through a period during which battlefield triage decisions are required?

Thanks in advance for what is sure to be a very enlightening thread.


Can I weigh in despite holding none of the qualifications as those of @ActivitiesGuy?

If so, my questions would be: do you have confirmed cases?

If yes, how it plays out will be determined by the average number of people someone infected is exposed to each day (social distancing) and the probability of each exposure becoming an infection (hygiene).

If no, and it’s because you don’t have any cases rather than a lack of testing, then if you can limit interactions with people from outside of the town (probably impossible unless you are self-sufficient) then you could ride it out quite safely.

I know of a few islands that have been spared. I’m guessing supplies they cannot manufacture for themselves can be airdropped and disinfected.

1 Like

I would love to hear your thoughts on the current CFR estimates from a statistical perspective. I was tweeting with AviBittMD about it, but you know how Twitter is… More frustrating than educational.

My main question/thought is whether or not current CFR estimates are overstated. He seems to think they’re just as likely understated (people dying from clovid-19 before being diagnosed with it). I think it’s probably the opposite, that a lot of people have it that are either asymptomatic or hav such mild symptoms that they don’t get tested. Meaning CFR is overstated. He said this is taken into account, but I don’t see how.

All that said, I don’t know what the fuck I’m talking about so…

Anyway, glad to hear your perspective on this whole mess.

*If the above falls outside what you want to discuss no worries.

1 Like

To my knowledge, we have no confirmed cases ATM (although I am given to understand several locals are currently under quarantine awaiting results). But surely it’s inevitable that we will, so there’s no point in hoping this event passes us by. And of course you’re right that, once it’s here, subsequent disease dynamics are causally influenced by how stringently the population adheres to mitigation maneuvers.

As for my specific questions…Surely you agree there is face validity to my assumption that, all else being equal, the denser a given population center is, the more likely transmission is to occur. I’m hoping AG can confirm that this ‘density hunch’ is borne out by the data.

I’m less confident about my hunch concerning a positive correlation between SES and case-fatality rates. On the potential plus side, consider:
—People in my area don’t have a lot of discretionary income, therefore they don’t travel nearly as much or as widely as individuals from wealthier locales; also,
—we are not exactly a sought-out tourist destination.
To my way of thinking, the above seem to reduce the number of ‘Patient Zeros’ we might have, which in turn would reduce the steepness of the curve, which in turn would blunt the acuteness of the demand for ICU services. Also:
—We don’t gather in crowds numbering in multiple thousands. This would seem to reduce the potential impact of each PZ, thereby also flattening the curve.

But on the distaff side of the ledger, I can envision factors correlating with low SES that might offset any advantages we gain otherwise:
—Being low SES means we have higher rates of co-morbidities (read: Us poor folk are fat smokers who all have ‘the sugah and the high blood’), and are thus more vulnerable
—We may have fewer ICU beds per capita
—There may exist an education-level-related unwillingness to adhere to social distancing.

As I have no idea how all these factors interact to determine the relationship (if any) between SES and COVID dynamics,I’m hoping AG can shed data-based light on it.


Currently 12 confirmed cases in our county, 16 total in surrounding counties. 133 statewide, 1 fatality so far.

The majority of confirmed cases in our state (Pennsylvania) are in eastern counties.

1 Like

I’ll start with a few “FAQ” type things (I will edit this post a few times this morning as I duck in and out due to meetings, so I’m going to “post” now at 9:30AM EST and then come back and edit more…if you have any specific questions that you’ve heard, maybe type them below and I’ll address them in this post if I can):

What’s the big deal? I heard this is just like the flu.

It’s natural to reach for the “flu” comparison here because it’s the nearest thing most people can liken it to. There’s some merit to doing so, but the downside (IMO) is that a lot of people already downplay or misunderstand “the flu” (e.g. they conflate “I had a bad cold and stayed home for a few days” with having diagnosed influenza) and so probably think seasonal flu is just something we all get every year and it’s fine.

From what I’ve read, seen, and heard so far, there are a couple of things to know about “the flu” versus the novel coronavirus. The first is to understand that since this is a novel disease, it’s so early that we don’t know much about how it behaves yet, and that is at least part of the reason for excess caution at this time (along with Italy, which we’ll get to shortly). With seasonal flu, we basically know the ceiling for just how bad it gets; with COVID19 we’re still figuring out what we are dealing with here, but there are reasons to believe it could be much worse than just a bad flu season. Let’s unpack some of the reasons why that is.

Case Fatality Rate

The first and seemingly simplest question one might ask about a new disease are “what are the chances someone who gets it will die?”

For reasons you’ve all read and thought about, it’s not a trivial task to pin down the CFR during an outbreak of a novel pathogen. We aren’t testing for it early on; even when we are testing, it’s possible that lots of people get it and don’t realize it (e.g. ride it out at home with a “bad cold”) and we’ve invariably testing the worst cases, so getting a precise estimate of the CFR is really tough in the early days. An imperfect but still (IMO) useful piece of information is the fatality rate among those who actually felt ill enough to seek hospital care - though it will overestimate the true/overall CFR, if that is much higher than for other known pathogens like seasonal flu, it at least tips us off that we’re seeing something that’s bad conditional on the patients feeling bad enough to seek care. We’ll discuss the health systems issue below, but even if the overall CFR is really overestimated right now because of the possibility that lots of people are walking around with no symptoms or mild symptoms and we’re only seeing the worst cases at the hospital, the fact that ICU’s are getting totally overrun with cases in some regions means that we still have a problem even if the CFR is low once you account for all of that. It’s little comfort to hospitals to be told by armchair commentators that the CFR ISN’T ACKSHUALLY THAT HIGH when they’re seeing patients die in front of them because they don’t have enough ventilators.

The best guess right now is about 1% in a system where patients can be adequately cared for (with acknowledgement that this could be much lower if there are tons of us walking around that already “had this” and just experienced it as a bad cold).


Lots of conversations about “R0” numbers these days - just how many new cases spring from one index case? This again is a useful summary measure but also simplistic (highly context-dependent) and hard to calculate in an outbreak of a novel agent without widespread testing.

What we do know is that seems to spread relatively easily, enough so that a few cases going to a conference can generate a whole bunch more cases, and that’s concerning enough that it’s a prudent idea to minimize big gatherings in confined spaces at this time. It’s probably not “walk past someone on the street and you’ll get this” but it certainly seems to be something you can get from shaking hands or touching common surfaces.

Health Systems

This is the biggest issue and one I’m going to return and update later, because it’s probably the most important thing to understand to grasp why this could be a public health disaster

We don’t even have that many cases where I live yet, so why the massive lockdown?

See above re: health systems and early unknowns. The major concern isn’t that Allegheny County (where I live) has 12 cases today; it’s that we might have 100 cases a week from now, 250 the week after that, and 500 the week after that; at this point we’ve probably exhausted the supply of available ICU beds and ventilators, which means hospitals and ICUs have to start making decisions like “COVID19 patients over 60 years old don’t get ventilator support, we have too many patients and need to spend the resources on people with the best chance of survival” which is how the CFR that’s 1% under normal conditions with sufficient care available creeps up towards 7-8-9% as is happening in Italy.

One other note that occurs to me - the clinical course of this disease seems to require that patients who do advance to needing ventilator support tend to need ventilator support longer than other comparable diseases/patients, which is another challenge here. The “number of cases” isn’t the only thing that matters here. The same number of cases will stress the system differently if all of the cases need to be on a ventilator for 14 days versus 5 days.

But just the elderly are at risk, right? Why do young people have to stop what they’re doing too?

Two things here. For one, it’s true that the elderly are at much higher risk than the young and healthy, but the risk to non-elderly is still greater than zero; my Twitter timeline is full of stories of deceased 40- and 50-somethings from COVID19 (admittedly there have been very few deaths to date in 20-somethings).

The second is the fucking shocking degree of callousness this seems to show towards people older than you. Whenever I see some dipshit teenager on spring break getting interviewed and saying “So what, if I get the corona I get the corona” - don’t you have parents? Grandparents? Friends that are older? Do you not understand that if you get the virus, come home and give this to them, it could be curtains for them? We can’t prevent all deaths from all causes, I get that, and as alluded (though I don’t really wanna go there in THIS THREAD) there are going to be some really complex social consequences with respect to people losing income (not just the rich, but the wait staff, day care babysitter, gig musician, wedding caterer, etc) and fallout from that. But anyways, I’m not a fan of the “so just the elderly get it, what’s the big deal?” line of thinking I’ve been seeing from young folks.


I agree there is face validity but I cannot tell you that I’ve read anything supporting that, nor have I read anything disproving that either.

With regards to SES, that’s tough to weigh-in on, as it’s not just income that matters but also social culture. Where I live, it seems as if the higher socio-economic status you enjoy the more socially isolated you already are. Meanwhile, being poor here isn’t conflated with bad health in the same sense (sugar, high blood pressure), although it’s not as if it doesn’t have a health impact.

Anyway, I shouldn’t weigh in on it anyway.

And immunocompromised people.

1 Like

Me either.

Hell of a way to find out that you have an underlying condition that exacerbates the illness or causes fatality.


And it’s coming from a generation that needs safe spaces.

Anyway, thanks to OP for this thread and ED as well for posting on it.

1 Like

Do you know what the CFR is for those under 60 with no pre-existing conditions? From what I’ve read, it seems like the majority of deaths are people over 60 or people over 40 with a pre-existing medical condition.

I gotta be honest, I don’t really blame them nor am I surprised. It’s hard for me to imagine the majority of people acting any differently at 18-22. I mean, I personally know Marines that raw dogged Thai girls knowing the various STD rates in Thailand (we were briefed)…

That also ignores some of the obvious political issues, but out of respect for the thread I’ll leave it at that.

To date this is true, yes, and that is generally consistent with the behavior of other viral conditions leading to respiratory distress. However, there are definitely cases of younger, healthier patients dying. It seems that some of these are healthcare workers, which has created theories that there’s an issue with cumulative exposure to the virus; my personal, non-fact-based guess is that many of the HCW’s who are falling victim and dying from COVID19 are physically and mentally exhausted and have decreased reserve to sustain themselves through the infection because this is happening in regions where the ICU and hosp is overwhelmed.

I’m going to try to discuss this a bit more later.

1 Like

Also, as we have seen in Italy, younger people will be given priority when it comes to intensive care over an elderly patient. We don’t know how many of the elderly who died would have survived if there were enough resources available.

Something else in Italy: in Lombardia, the region most affected, cell phone tracking has shown that 40% of the people are still going outside in spite of the lockdown. This includes elderly people.

This has some condition specific stuff, some general, but it’s long and dry. You have to be really interested to hang in there on this one.

It is pretty informative though.


From what I understand, going outside isn’t the problem. It’s going outside to areas where virus transmission is likely.

If people are going for a walk in the woods, or around the block to get some quick exercise, I doubt that’s much of a problem. But if people are going to stores, meeting up, going to common areas, I see that as a problem.

It’s Italy. People are socializing.

Fair point.

AD: I just read a post from a buddy’s orthopedic surgeon saying that Backcountry skiing is dangerous/selfish from a virus transmission standpoint due to the heavy breathing involved in the way up, and the virus being an aerosol.

In my mind, being out in the mountains with only my SO and dog in my group, where the closest people are hundreds of yards, to 1/4 mile away seems pretty safe from a virus standpoint (not counting the numerous dangers you normally deal with out there).

What are your thoughts on safety of outdoor recreation?

Without touching what people in Italy are doing:

Yes, for those of you wondering about this: even under whatever state your, well, state may be in (for those of us in the US) it is fine to go outside and leave your house. Go for a walk/jog, go to a park, etc. If you normally run or cycle with a group of people, that’s fine (though anyone who feels ill should stay home, and you’d do best to avoid your usual handshakes and high-fives in favor of elbow bumps or dapping, lol).

What we want to do at this time is minimize large gatherings of people where super-spreading events are a higher risk. It’s fine to go to the store if you need something, but try to minimize the number of trips. Work from home to the extent possible. We’re not at “you can’t even open your front door! Don’t go outside” levels of lockdown nor should we ever be at that point. We just want to avoid the kind of gatherings where one person showing up with the virus can turn into 50 people leaving with it (which sadly includes things like a restaurant, since the virus does live on surfaces).

We’ll know more in two weeks. The reason for the “cancel everything now and distance as much as practical” urgency was to buy ourselves some time to figure out just how bad this is going to get. Pennsylvania is up to 185 confirmed cases (still only 16 in Allegheny County) while the US overall is approaching 10,000 cases (and again, we’re still way behind on how many people have been tested, and we’re only like a week into this). The concern isn’t just about 10,000 cases today (for the fucking morons that are still like “but there’s more flu cases than this every year”). It’s that if we’re on our way to 100,000 next week and a million in a month, this could get ugly in a hurry, and this disease has a long enough incubation period before people show symptoms that we have no idea how many time bombs are walking around right now, but we’ll have a better picture of that in 7-10 days.


Up from last night.

Total is hasn’t gone crazy, but it’s rising none the less.

It’s downright spooky at times. Some of my usual day to day stuff that would usually be crowds and traffic is almost like ghost town levels of people.

1 Like

Lol. You’re telling me, man. I’m still coming to Oakland to work. Technically, we have ordered everyone that can work remotely to do so, and canceled all in person meetings (with everything being done by various teleconference software - Zoom, GoToMeeting, etc). Since our office is entirely empty, I actually feel okay about coming into the office - there’s little practical difference between working from home and “driving myself to the office to go sit in my office with the door closed and not see anyone” - the only “exposure” I get that I wouldn’t otherwise is touching a few door handles, and I’m washing my hands basically every time I come in and out of the building.

Like I said, the “16” ain’t the concern, it’s that as the number rises, we know those people almost assuredly passed it to a few others before they knew they had it. Exponential growth happens slowly and then all at once, and the downside of not acting was potentially catastrophic (with full acknowledgement that shutting everything down for 2 weeks or more is gonna have its own very unpleasant fallout eventually, and that’s where the GUBMINT needs to step up IMO). One of the problems with this is that we’ll never know for sure if we overreacted, but we’d sure as hell know if we under-reacted, and by then it would’ve been too late.

One of the things that everyone who’s suddenly an expert will be spouting in the weeks to come is the sorta-surprisingly low number of ICU beds per capita around the nation. Hospitals end up kinda like hotels - if they have lots of unfilled beds on a regular basis, the admin geniuses close or repurpose the beds. Good for profit margins, bad for “Oh shit we have 1,000 people with this new virus, 10% of them need to be hospitalized…wait what do you mean my local hospital only has eight ICU beds!!!”