The Pandemic-Era Emergency Dep’t: Weirder, Wilder & Emptier Than Ever

Matt Bivens, MD
34 min readApr 27, 2020
Coronavirus, image created by JunGSA from the Noun Project

The radio toned at 3 a.m., a single long tone. No big deal, just a heads-up. If the patient had been “priority one” sick, or if paramedics had needed medical orders by radio, the tones would have been frantic and oscillating, and everyone would have stopped to listen. This was not that, and the paramedic’s voice was low-drama: They were arriving, they’d need a room.

Their patient was an elderly gentleman recuperating from a stroke at a Long-Term Acute Care hospital (an LTAC, or “acute rehab”). He’d developed fever and cough, and had just tested positive for the SARS-CoV-2 coronavirus that causes COVID-19. He now had a requirement for oxygen, his heart rate was irregular and a little rapid, and he was headed our way.

Well. He didn’t sound that sick.

Did he really need to be bundled into an ambulance after midnight — disrupting his sleep, exposing a crew of paramedics to the coronavirus — and hustled from one perfectly good hospital to another?

Soon he’d arrive, and I’d get to evaluate the situation. But like a good ER doctor, I’d barely heard about the patient and I was already pondering how to get rid of him: Admit him to my hospital? He’d go to an isolation room, but that still potentially exposes staff upstairs, and we were filling up. Turn him around back to the LTAC and his still-COVID-warm bed? Maybe. Sometimes the LTAC doctors were wrong, and COVID-19 had everyone jumpy.

Then again, sicker-spectrum COVID-19 patients can abruptly go south. Maybe he’d arrive, decompensate, and end up on a ventilator in our ICU. That was clearly the resource nurse’s view. Based on vitals and a two-sentence description over a crackly radio patch, we’d each formed our impression:

Mine: “LTAC doctors are freaking out because someone said, ‘COVID-19’! Patient actually fine and will have a ‘meh’ workup. I will roll my eyes at their sheep-like panic, and send patient back.”

Resource’s: “Sick old man with COVID-19, new oxygen requirement. Once doctor gives up on wishful fantasy that patient is somehow fine even though we have shut down the entire nation over the coronavirus, doctor will put on the ventilator and admit to ICU.”

What to Wear, What to Wear …

“Do you want PAPRs?” asked the resource nurse.

These are Powered Air-Purifying Respirators: In our case, a white plastic hood with a clear face shield, attached by hose to a motorized fan/filter worn on a belt around the waist. We don these spaceman hoods now for high-risk situations like intubations, the prologue to putting a patient on a vent.

If a patient is failing despite oxygen, then he might need sedated, intubated (i.e., have a plastic breathing tube slipped into his trachea), and put on a ventilator. We do this routinely in emergency medicine. But it involves getting up close with a coughing, struggling airway — perhaps between periods of vigorous bag-mask ventilation — and it turns out this is all high-risk for aerosolizing a coronavirus, so that it floats in the air all around us.

This happens in a negative-pressure room — resource was already tracking the patient in the computer to our main resuscitation bay, which has a sliding glass door and a fan that continuously sucks air in from the hall — so no viral particles can wander the ER. The fans draw the air through filters and outside of the building — hopefully someplace up high and remote, where any few scattered viral or bacterial particles that make it so far will be killed off by sunlight. None of this, however, protects those of us inside the room, hence the question: Should we dress like astronauts to meet the new COVID-19 patient? Or go with standard gear?

Standard included an N 95 mask, which each of us had been wearing all shift, for weeks now. They feel like hard cardboard, with moldable edges. When sealed to the face, supposedly they keep out “95%” of whatever’s floating in the air — as long as that whatever is bigger than 0.3 micrometers (300 nanometers). (This is regulated by the National Institute for Occupational Safety and Health; the N stands for “not resistant to oil,” which means it’s fine for healthcare work but not for some industrial processes.)

If you’re wondering: “Is filtering out 95% enough?” — join the club. Sucking in 5% of the coronavirus that comes my way sounds like a bad deal.

Worse, the coronavirus itself is only 0.125 micrometers (125 nanometers). So … small enough to make it thru the mask?

Nevertheless, we have some clinical evidence that N 95s prevent viral or bacterial infections. And we hypothesize that if say a coronavirus is floating in the air, it’s doing so in a large water droplet. Suddenly, the exact size in micrometers of said droplet is of interest, so there’s a brisk trade in math-heavy papers like this one from the Journal of Fluid Mechanics — with its 1,000-frames-per-second images of sneeze- and cough-expelled saliva sprays. This and other literature suggests virus-filled saliva droplets range from 5 to 15 micrometers (5,000 to 15,000 nanometers)— far too big to make it past the N 95.

Maybe so, but the N 95s are miserable things.

Before COVID-19 they were considered “single-use,” worn to see a patient and then discarded upon leaving the room.

Now, in the setting of an international shortage, at every hospital I work at or know of, they are being used in a completely new way: Worn constantly, sometimes with a surgical mask over top to “keep the N 95 clean,” and then turned in for some sort of deep cleaning. The CDC has offered only the most grudging of guidance blessing this sort of reuse, but what can we do? At least we are past the early days, when we doctors were literally studying the specs on vacuum cleaner bags and air conditioner filters, wondering if we could cut them up and sew them into face masks.

To be clear: At none of the hospitals where I work did we ever run out of protective gear. But at all of them we had reason to worry about it, and if we haven’t run out, it’s in large part because of the ingenuity of the physicians and nurses in suggesting workarounds.

Steam-cleaned and ready for another shift

The N 95 I was wearing this shift had my name and ID badge number written inside it, and had been steam-sterilized and returned to the emergency department. I had found it at the beginning of my shift hung on a wall in a plastic bag, next to multiple other cleaned masks. I would reuse it one more time before starting with a fresh one. I and other doctors had scoured the medical literature about this practice and it seems scientifically defensible. (But science or not, everyone had rebelled at suggestions the N 95s could just go into a common pool, so that we might inherit someone else’s used mask.)

The masks are held in place by elastic straps; the edge and the straps press into the face, making all of us look puffy, wrinkled and ill. Everyone has headaches, and since we’re medical people we debate whether it’s from the muscle tension or the hypercarbia. One of the nurses and I used the end-tidal CO2 detector on ourselves: With the detector’s nasal prongs under my mask, my expired CO2 was high-range normal at 42 mm Hg; it went to 36 mm Hg with the mask off and a couple of refreshing breaths, and I felt better blowing off all of that carbon dioxide waste. My nursing friend’s CO2 readings inside the N 95 were even higher. Everyone in medicine knows that the more CO2 patients retain, the sleepier and loopier they get. “I wonder how many bad medical decisions I’ve made in a hypercarbic fog,” asks one of my physician colleagues.

The N 95 is not only suffocating; it’s giving everyone problems with the bridge of the nose, either skin breakdown, acne or both. Some put bandaids over the nose, but that was officially condemned as an affront to the sacred seal of the mask. Barrier creams like those for diaper rash were trialed; they grease up and discolor the mask, and those in charge of the steam sterilizers objected. (For the same reason, none of my colleagues wear makeup anymore.) The marketing department kicked in a bunch of clear lip balms with the name of the hospital system on them, and rubbing that on the nose bridge seemed to help, some, for awhile. Lately I’ve just been putting paper tape across my nose, the seal be damned. (I figure, “So with tape it’ll be an “N 93.5”) Healthcare workers from the COVID-19 era may well end up with a recognizable permanent scar on the nose bridge, like a tattoo from a deployment.

We treat the N 95s almost superstitiously, like a yin-yang: the outside surface is poisonous evil, the inside pristine goodness. At some point in an 8- or 12-hour shift, the mask has to come off, if only to hydrate and snack. But the donning and doffing for some is akin to religious ceremony. People model ritualistic techniques for removing the mask from the face directly into a Tupperware container, for example.

So, back to resource’s question: Keep on the awful N 95s?

Or take them off — oh so respectfully! — and don a PAPR hood?

It is excellent my hospital even has PAPRs. We are better prepared than many. (Some of us advocated for them after the Ebola encounters of 2014). But crash-donning PAPRs for emergent intubations is a Wild West idea, since they arguably aren’t meant to be donned by people who’ve already been working half of the shift in an infectious hot zone. (And in a recurring theme, our PAPR hoods are also technically single-use. The manufacturer, 3M, has a reluctant document describing how to reuse them after a coronavirus exposure. My ER colleague who found it was transported with nerd-ecstasy, describing the deep-web search he’d had to conduct before he could pry this knowledge free and confirm that “all we have to do” — to quote him paraphrasing 3M — “is just spray the s%$t out of the PAPRs with isopropyl alcohol 70%!”).

Some of our emergency doctors still object to putting a clean PAPR hood on a dirty head, and cite the YouTube videos of the “intubating teams” in Wuhan, China, which have gear and protocols that put American medicine to shame. Others counter by citing Canada’s terrible encounter with SARS — which in 2003 may have infected nearly every 10th doctor in Toronto who performed an intubation. When the Toronto doctors adopted PAPR hoods, the infection rate associated with intubations dropped to zero. (There are lots of confounders: they got more careful overall, once they saw a deadly coronavirus claiming colleagues left and right; but it’s the best data we have.)

How Dangerous is this Virus Anyway?

SARS was the original coronavirus terror: It presented like “a typical viral illness”: Fever, chills, bodyaches, non-productive cough, and about half of the time with copious watery diarrhea. But unlike a “typical viral illness” — does that once-soothing phrase even mean anything now? — half of all people who contracted SARS needed supplemental oxygen; every 5th person ended up on a ventilator, and nearly every 10th died.

The SARS outbreak 18 years ago in central China spread to 29 countries, with more than 8,000 infections causing 774 deaths (9.6% case fatality rate). MERS, the Middle East Respiratory Syndrome, last surfaced in Saudi Arabia six years ago. It had a similar clinical presentation (fever, cough, +/- diarrhea) but was an even nastier coronavirus: only 2,400-plus cases in 27 countries, but 858 deaths for a 34% case fatality rate.

Compared to the other severe coronaviruses, COVID-19 is proving to be more contagious but less virulent. The calculated case fatality rate is all over the place and is hard to establish mid-pandemic, but as Dr. Anthony Fauci and colleagues recently wrote in the New England Journal of Medicine:

… the case fatality rate may be considerably less than 1%. This suggests that the overall clinical consequences of Covid-19 may ultimately be more akin to those of a severe seasonal influenza (which has a case fatality rate of approximately 0.1%) or a pandemic influenza (similar to those in 1957 and 1968) rather than a disease similar to SARS or MERS.

So what’s worse: A highly lethal disease like SARS that’s harder to catch, or a less vicious disease that everyone catches? I guess it depends on whether the question is philosophical or mathematical. At this writing COVID-19 has killed more than 200,000 people worldwide — a death toll 250 times the size of the SARS outbreak — including 55,000 in the United States. (That breaks down further with about 22,000 in New York, and about 2,800 and counting in my state of Massachusetts.)

For the United States, so far this is comparable to the death toll from our regularly scheduled influenza — another viral illness that for the vast majority of the millions who contract it is a minor event, but for some thousands is deadly. CDC estimates that influenza killed 34,000 Americans in the 2018–2019 season, and about 60,000 the year before that, including a few hundred children; COVID-19 has generally not hit the pediatric population.

It’s sobering that in just a couple of months — an era marked by unprecedented societal shutdown as a disease-control measure — COVID-19 has still already surpassed last year’s entire flu season death toll. The latest predictions are for 60,000 to 100,000 COVID-19 deaths by late May. If so, that would be “a success” scenario of sorts — but still a death toll comparable to, for example, the opioid epidemic’s, which was the nation’s other public heaalth crisis just several months ago (and which has been killing 60,000–70,000 a year of late).

I have no idea what happens if we “open up the country” tomorrow. Maybe COVID-19 explodes with all of its contagious furor and we see not 100,000 deaths but 1 million, or more. Time will tell.

When a Single Neb Treatment Killed 15 People

Like COVID-19, SARS also first appeared in China. A 64-year-old physician then traveled from there to Hong Kong, checked into a hotel, infected 16 other people, and 10 days later died. Many of the hotel guests he infected got onto international flights. One flew to Toronto. Another went to a local Hong Kong hospital for shortness of breath, and was treated with nebulized salbutamol (a bronchodilator to open up tight, wheezy airways, like the albuterol people across America use for asthma).

A nebulizer bubbles air through a medical fluid for the patient to inhale; unfortunately the patient also exhales, and this carried the SARS virus in a mist of droplets throughout a tightly packed Hong Kong emergency department. Over the next 2 weeks, some 138 people who’d been there that day ended up hospitalized with SARS, many of them healthcare workers.

Fifteen of those people died, and among survivors studied, even two years later, most had decreased lung function, exercise tolerance and overall health — all from just having been in the same emergency department as a SARS patient getting a neb. It’s no wonder that in the COVID-19 era, hospitals are afraid of nebulizers. All this spring, we’ve had asthmatics show up expecting nebulized medications — I hand them an albuterol inhaler instead.

Even without riding a nebulizer’s mist, back in 2003 SARS — supposedly only spread by close contact — got airborne pretty spectacularly. In a brilliant feat of detective work, Chinese researchers reconstructed step by step how SARS spread to infect more than 300 people at a high-end private apartment complex in Hong Kong, the Amoy Gardens:

Graphic from New England Journal article on the Amoy Gardens investigation

The index patient infected with the SARS virus visited unit 7 on the middle floor of building E on March 14 and again on March 19 and used the toilet; the patient had diarrhea. Subsequent cases of SARS … were located in clusters in four buildings and at certain floor levels …

WHO researchers found problems in the toilet drains; Hong Kong University engineers built “a mock-up of the drainage system” and found that when the building’s flawed toilets were flushed, “huge numbers of aerosols” were vented back into the bathrooms, ready to be dispersed outdoors by the exhaust fans. Computer programs modeled these humid, coronavirus-seeded plumes, wafting through the open windows in the middle floors of neighboring buildings, where residents in turn became infected.

Computer model of airborne SARS (blue plume) from a single dysfunctional bathroom. From the NEJM article.

Others had argued SARS couldn’t possibly have gotten airborne! No, it must have been spread person-to-person through the complex. Or maybe by rats on the rooftops. But with the entertaining contempt of a Sherlock Holmes, the researchers dismissed all that:

Members of the management and security staff of Amoy Gardens, who worked on the ground floor in each building 24 hours a day and would probably have had frequent person-to-person contact with the residents, were not affected by the virus … [Meanwhile,] the theory that roof rats were both amplifiers and distributors of the SARS-associated coronavirus … [is] not supported by the epidemiologic distribution of cases: the middle-level floors were affected more than the upper floors, and certain units were affected more than others in the same building. Roof rats are by nature territorial, and they therefore could not be responsible for the rapid and efficient spread of the infection from a single building, building E, to other buildings … [The roof rat theory also] does not explain the steep decline in the epidemic curve after the peak, because there was no sudden disappearance of roof rats or massive deaths among them.

That was SARS. Can it’s younger brother COVID-19 also be transmitted from fecal matter? Could it also get airborne, if the conditions were right? Officially, we don’t know. But logically, it probably can. As with SARS and MERS, COVID-19 patients present with diarrhea, and the coronavirus has been found in stool samples of more than half (!) of patients tested.

When SARS traveled by plane from Hong Kong to Toronto, it devastated local hospitals. Some 438 people became infected and 43 died. More than half of those infected were doctors or other healthcare workers, some of whom also infected their family and even children. Later investigations found that nothing was so dangerous as intubating an infected patient: It increased the risk of getting infected yourself by from 7- to 9-fold. For those of us who intubate patients — paramedics in the field, and emergency, anesthesia and ICU doctors in the hospitals — these are stark numbers.

COVID-19 is far more contagious than SARS. It turns out it ain’t SARS (killing every 10th victim) or MERS (killing every 3rd) when it comes to the likelihood of death or disability if you catch it. So, great! Nevertheless, more than 200 doctors and nurses worldwide have died from COVID-19, and I’d prefer to avoid it.

We had recently been gifted with huge Plexiglass boxes, with circular arm holes; the idea was they could be placed over a supine patient’s head, and the doctor could reach through the arm holes to perform the tracheal intubation. This is again the Wild West, totally untested. A proof-of-concept article in the New England Journal of Medicine, by some smart folks at Boston Medical Center, had included a dramatic simulation of a patient’s cough during the intubation procedure:

Flourescent dye on the intubator’s face shield after “the mannequin” coughs

To approximate a forceful cough and generate a spread of droplets and aerosols, a small latex balloon containing 10 ml of fluorescent dye was placed in the hypopharynx of the mannequin. The balloon was inflated with compressed oxygen … until the balloon burst … in a crude simulation of a cough. … [Without the box] dye was found on the laryngoscopist’s gown, gloves, face mask, eye shield, hair, neck, ears, and shoes. Contamination of the floor occurred within approximately 1 meter from the head of the bed and also on a monitor located more than 2 meters away.

Gross. And sobering.

Our medical teams have been highly focused on how our procedures — nebulizers that bubble breathing mists, or CPR, or tracheal intubation — can seed the air around us with coronavirus. But actually there’s nothing so likely to seed the air far and wide as the patient him or herself, simply coughing. Coughing is literally the body’s way of trying to eject the infectious material (and associated sputum and gunk) as far from itself as possible.

Yet in the odd logic of our times, none of us are PAPR’ed up for coughing patients; there’s no proposal to put every awake and alert coughing COVID-19 patient into a square plastic helmet. (We do make them wear a surgical mask). No, those sorts of heroic gestures are reserved for the 10 minutes or so when we perform our “aerosolizing procedures,” not for the next 10 hours of arguably-as-dangerous coughing. It feels like a cognitive error. But just try suggesting to already-stressed hospitals that now every nurse should be in a PAPR hood.

There’d be another sick COVID-19 patient here soon. I’d intubated two so far this season and seen many more, and I still felt like I was making it up on the fly every time.

No PAPRs, I said. Let’s just meet him. (I still secretly hoped he’d be fine.)

No plastic boxes either.

If we decide we have to intubate, we’ll re-group.

‘I Just Can’t Cough’

Paramedics wheeled the man into our resuscitation bay. He looked fatigued but alert, with a normal respiratory rate, and in no distress. (Note: I have changed non-essential details of this case to protect patient privacy.)

He’d been hypoxic at the LTAC, but the paramedics had him on a high-flow nasal cannula at 10 liters of oxygen a minute, and his oxygen saturation was 100% — too good! It actually needed dialed down a bit. (Too much oxygen is also bad for a patient).

The LTAC had started him on hydroxychloroquine and azithromycin. The data for those therapies is so weak, we might as well prescribe chicken soup and mustard plasters. (At least chicken soup doesn’t make you sicker, the way hydroxychloroquine might.) But when the president of the United States keeps insisting something’ll work, there’s moral pressure to try it.

Some patient details changed to protect privacy.

“How do you feel?” I asked him.

“What?” he half-shouted. He was heard of hearing.

“HOW DO YOU FEEL?” I shouted.

“Better,” he said. “Pretty good. Only I want to cough, and I can’t. Nothing comes up.”

He demonstrated, making a wet noise. All of us in our gowns, face shields and N 95s (and me in my N 93.5 with the tape sticking out) involuntarily swayed back.

He had no fever, and his blood pressure was solid. His heart rate was in the 130s (normal rates are from 60 to 100 beats per minute), but that’s not uncommon when an illness combines with atrial fibrillation, a chronic dysrhythmia for this man. He also hadn’t gotten his morning medications yet, including some to keep that rate slow.

In the pandemic era, hospital policies change minute-by-minute. On this given morning, we had a soft system-wide rule against use of high-flow nasal cannulas. (Neither the LTAC nor the ambulances had this rule). It was a decree laid down over a theoretical concern, not borne out well in the literature, that if a patient opens his or her mouth, pumping oxygen into the nose at high flow rates could spray coronavirus out the mouth.

(C’mon. At some point, we have to calm down a little.)

Yes, there are studies of this too. Researchers have created simulated snot out of tap water thickened with potato starch, smeared it on mannequins, and used nasal cannulas to try to spray it out of the mouth; they have bathed the mouth and nose of mannequins with yeast broth, pumped in the high flow oxygen through nasal prongs, and seen if yeast grows on petri dishes set up vertically and facing the mannequin at 30 centimeter intervals. No water-potato snot and no yeast cultures appeared beyond 2 feet from the patient’s face. Other researchers, in China, set up petri dishes around ICU patients with particularly aggressive pneumonias as they received oxygen either by high-flow nasal cannula or by the supposedly safer oxygen facemask, and didn’t find much difference.

In any case: This gentleman was doing well on high-flow, so for the moment I left it. At 3 am no one was around to see.

I gave him an intravenous dose of medicine for his irregular fast heart rate, and it slowed nicely. Nursing titrated his oxygen down from high-flow rates to just a few liters per minute by regular old nasal cannula, and his sats were a beautiful 95%. Within minutes of arrival, he suddenly looked quite well.

Back at my desk, I further reviewed his history, typed up my preliminary findings, and watched as his lab results started arriving in the computer: A reassuring white cell count, good kidney function, no liver or heart injury, normal electrolytes … I was starting to mentally pat myself on the back: The rehab doctors had overreacted. Sure, he had a flu-like illness and would have fevers and coughing and feel cruddy; but he could ride this out with supplemental oxygen and supportive care back at his rehab.

Then his portable chest Xray resulted. It was startling. I was looking for a mild viral pneumonia picture. Instead he had near-total whiteout of his left lung.

It looked like a new, large pleural effusion: Fluid inside the chest but outside the lungs, compressing them. The patient had a weakly pumping heart at baseline; was this congestive heart failure? Did he have a cancer in his lungs, which can sometimes spread fluid like this? Or was this just COVID-19 pneumonia? It has been known to present to one lung side, and with associated pleural effusions.

Even as I frowned over this X-ray, nursing informed me they were starting to dial back up his oxygen.

The Dread Ventilator

A lot of attention has focused recently on “rationing limited ventilators.” We weren’t there this particular early morning in Massachusetts. But a ventilator is a brutal therapy. Even young, healthy patients find themselves profoundly deconditioned after a few days lying in a bed sedated and perhaps paralyzed, with the machine doing the breathing. The lungs aren’t used to being inflated actively; they are used to being tenderly tugged open by the expanding chest and dropping diaphragm, which gently draws air in; with a ventilator, the air is forced in, the lungs are forced open from the inside, so they expand and smush up against the inside of the torso, lifting the dead weight of the chest wall out, and forcing the diaphragm down. It’s debatable whether the modern ventilator is always an improvement over the Iron Lung, the big tank used for patients with polio paralysis in the 1930s — in those the patient’s head would stick out the top, like a volunteer at a magic show about to be sawed in half, and the Iron Lung would create negative pressure gradients inside, to suck outward the torso and draw in the air.

An Iron Lung (from the Wikimedia commons).

Even with exquisitely selected and titrated volume and presssure settings, ventilators can cause lung damage; drawn-out sedation leads to delirium and sometimes cognitive decline; nutrition is challenging; muscle wasting and weakness can set in; bacteria can migrate down the plastic endotracheal tube to cause a ventilator-associated pneumonia; and the old and frail, if not quickly weaned off, can simply waste away.

My older patient was definitely frail. He had left-sided muscle weakness from prior strokes; he’d been on the vent before, some years ago, and had such a hard time weaning, he’d had a tracheostomy for a time — basically a hole cut into the neck to allow the patient to be alert and sitting up and interacting but still hooked up to a machine. He’d eventually been weaned off of the vent, and the tracheostomy site closed, but I could still see the scar.

Like every other hospital, ours has a pandemic-era “no visitors” policy. But family are crucial for both doctors and patients — there’s a lot that needs negotiated, and a sick person needs their loved ones at hand. Calls to his son and healthcare proxy went to voicemail. But per the documented discussions at the LTAC, the patient and his family had made clear that if he worsened with this illness, they would want us to try a ventilator to bridge him through; and if his heart stopped beating, they would want us to perform CPR and other heroics before “giving up” and declaring him dead. He was, in medical jargon, a Full Code.

How frank had those discussions at the LTAC really been, back in the days when he was more alert and less sick? With his profound baseline weakness and comorbidities, if he went on a ventilator, he might survive … for awhile … but he’d further decondition, likely have further cognitive decline, likely run into complications, and likely die without ever escaping the hospital.

COVID-19 is a viral pneumonia — and medicine calls pneumonia “the Old Man’s Friend” for a reason. It can be a quiet, relatively easy way to slip away — maybe after some hours or a day or two of low-key final interactions with loved ones. The alternative could well involve ending all human contact now — the last person you interact with being me, a doctor in a space suit hood, right before the medications put you under. This is so highly likely in the COVID-19 era — when literature suggests few people put on ventilators actually make it off of them — that we recently put a computer tablet into the emergency department, specifically to allow patients to interact with loved ones by video a final time before going under onto the vent.

I didn’t want my patient to die today, and I always want to respect family and patient wishes. But I also didn’t think a ventilator was in his interests — especially if it could be avoided. I thought about that left lung whiteout on chest Xray. If it was mostly pneumonia, that’d be terrible; but if enough of it was just pleural effusion, I could do a semi-emergent thoracentesis — stick a needle into the back of his chest, drain the fluid from around his lungs. It’d be challenging; he’d need to sit up and preferably follow commands, and since he also happened to be on blood-thinning medications, the procedure would come with increased bleeding risks. But if it worked, and a large amount of fluid could be drained off, it might stabilize his slipping oxygenation; he might be able to avoid the ventilator.

In gown, gloves, N 95, and face shield, I wheeled a cumbersome ultrasound machine into the resus room. He was resting with his eyes closed, wearing a nasal cannula with a surgical mask on top of it. He was breathing faster and looked worse.

He woke with a touch to his shoulder.

“HOW DO YOU FEEL?” I shouted.

“Weak,” he said. “I want to cough, and I can’t.”


He gave me a thumbs up.

I dropped the railing of his bed. I pumped the foot jack to raise the whole bed higher. I moved aside a little wheeled table. I wrestled the ultrasound up snug against the bed and locked the wheels. I ducked under a jungle of monitor cords between the back wall and the head of the bed to plug in the ultrasound, and reflected once again as I did so on how much of my time at work I spend climbing over stuff, and under stuff, and rearranging furniture.

I laid his bed back 30 degrees. The ultrasound flickered to life and with gel on the probe I went to look at his lower left posterior side, as he lay flat.

“COLD GEL!” I warned him, but he didn’t flinch.

I could see some fluid but views were poor.

I tried having him sit up. He said he was too weak, and he was — even with him grasping a railing with his good right hand to help pull himself upright, I struggled to lift him, he was all deadweight. I wedged my left shoulder and arm behind his back and tried to run the ultrasound probe with my other hand but he kept flopping sideways.

His nurse came in to help. Like me she was in a flimsy yellow disposable gown overtop of scrubs, an N 95 mask, gloves, a hat — hers a bouffant, mine a cap — and a plastic face shield.

“Let me help you!” she said to the patient, and she grabbed each shoulder and pulled him straight upright, facing her.

“I’m too weak,” he complained through his surgical mask. “I can’t cough!”

He made a little cough, straight into her face, but she just smiled encouragingly.

Example of how consolidation (pneumonia) looks like liver on ultrasound. From an article in the Western Journal of Emergency Medicine.

My first good views of the lung I did a double-take: Was I seeing liver? I automatically double-checked I was on the left side of the body where the effusion was, not the right where the liver lay. But no: This was lung consolidation. The lung usually just looks like white noise on an ultrasound screen, because of all of the air pockets, which don’t carry sound waves well; but when it goes from a damp airy sponge to a sopping wet sponge — when the tissue is heavy with a consolidating pneumonia — then on ultrasound it often shows up like a solid organ similar to liver or spleen. There’s even a term for the ultrasound appearance of a densely consolidating pneumonia: “lung hepatization,” or liver-ization. There was fluid as well, between the walls of the lung and the inside of the chest cavity, the so-called pleural space. But it was smaller than I’d expected from the chest Xray.

“That’s enough!” the patient said. He’d been complaining almost from the first, for the entire 30 seconds or so of the exam. I’d been automatically dismissing his protests like a typical doctor — “JUST A BIT MORE SIR!” “HANG IN THERE ALMOST DONE!”

But now he was actively struggling to escape our grasp — “I can’t sit like this! It hurts my back! Let go!” — and we lay him back.

He sighed with relief. But he looked sicker than ever.

Roll him onto his left side, so the fluid would all be dependent down, and still do a thoracentesis? No. Now that I’d seen a clearer picture by ultrasound, there was just too much pneumonia here. Drawing off the fluid layered around it would not be game-changing.

What about proning him? To lie prone is to lie face-down. Physicians treating COVID-19 have found the backs of the lungs often more heavily involved, as were my patient’s by the ultrasound views. They’ve also found that having patients roll onto their bellies for stretches — or, if they are intubated and sedated, having the ICU team position them this way — can improve oxygenation. It’s a hot topic in COVID-19 care, and a term has been coined — “the happy hypoxics” — for young, healthy people with COVID-19 and startlingly low oxygen levels, who nevertheless aren’t particularly distressed, and who improve lying on their bellies. But patients who self-prone have to prop themselves up on their elbows or be carefully positioned with pillows; my patient was so weak, I feared he’d just press his face into the pillow and give up. Self-proning for him was no more of a solution than a thoracentesis.

Time to Suit Up

“Sir, I don’t think you can keep your oxygen levels up without the vent,” I told him. It was hard to believe that I’d known him for less than an hour. Or that I once thought he’d looked well-enough to go back to rehab. COVID-19 cases really can decompensate rapidly. (It made me wonder about all of the cases who seemed perfectly fine and who I’d sent home.)

“What?” he said.


“Oh, vent!” Thumbs up. He wanted that. He was so tired now, he looked relieved at the idea. We had him on an oxygen face mask by now and his O2 saturation was only 78%.

We started preparations to intubate him. I gave a verbal order for the meds, then discarded my gown and gloves, washed my hands, left the room. I doffed my face shield and my N 95 mask. Officially I am supposed to put an N 95 I plan to reuse into “a paper bag” to “keep the dirty outside from contaminating the clean inside.” Where was this mythical paper bag? I’m sure there was an e-mail about it. Instead I nested the mask face-down in an empty cardboard tissue box at my desk. What a relief to have the damn thing off. I threw away the paper scrub cap, went to the staff bathroom and washed my face. I tried to rub away the elastic band and mask creases — and then, as an afterthought, washed my hair in the bathroom sink, and dried up with paper hand towels. Which is not part of the PAPR protocol, but what the heck. We’re all improvising these days — this ain’t the Wuhan intubation team, with its “3rd level protection” protocols.

I put on new gown and gloves, put the PAPR air flow unit belt around my waist, turned it on and donned the hood. Instantly I was in a cool and refreshing bubble. The hubbub of the emergency department was distant and muffled. All that was missing was some piped-in soft jazz. I remembered how all of the literature had insisted on “comfort” being one of the PAPR advantages — that had seemed oddly frivolous, until I’d worn an N 95 for a week.

Back at bedside, the patient was tiring rapidly. His O2 sat was 63% on the facemask. I was starting to feel I’d failed, letting things get so far gone; that I should have just intubated him right up front, when he had more reserve and it was safer. I knew if I pushed sedatives and paralytics for the intubation now, his oxygen levels would sink like a stone. This was one of the main ways a patient abruptly dies peri-intubation: If the doctor stops the patient from breathing but then can’t slip the tube in place efficiently and start providing oxygen quickly enough, patients can go into cardiac arrest.

It’s not as easy as it sounds: Every intubator has at least occasionally been convinced they’d put the breathing tube into the trachea, but actually missed and slipped it into the esophagus. In such cases, it might take only a squeeze or two from a bag-valve mask to realize the error — but that can already open up the gastroesophageal sphincter, releasing a geyser of stomach contents. Vomit can come jetting up the tube like a garden hose, even as it bubbles up the esophagus all around the tube; stomach acid and contents pour out of the mouth, all over the patient’s face, soaking the bed sheets around the head; it sprays all over the healthcare workers involved, who are frantically suctioning it with catheters; it gushes down the patient’s open airway, paralyzed and unable to cough protectively, and pools into the lungs. For that matter, sometimes sedatives and paralytics alone loosen the stomach sphincter: You can be doing everything right and smiling at a clear view of vocal cords either side of the trachea, right before a flash flood of vomit washes it away. This is why the anesthesiologist doesn’t want you eating after midnight the day before elective surgery.

These were the routine concerns with every emergent, profoundly hypoxic intubation. But in COVID-19 there were added twists. One, there was often extensive airway edema — another quirky charm of this virus, it was known to cause swelling in the back of the throat, sometimes to the point of making it hard to fit a tube through. Would he have airway edema? Two, among the bewildering proliferation of pandemic-era hospital policies — changing sometimes hour-by-hour — we this morning had another soft hospital-wide rule against bag-mask ventilating a patient; instead, we would immediately hook an endotracheal tube to a ventilator circuit. That sounds fine, except that the vent hums along in its mindless, sedate fashion, while a hypoxic patient who’s just been intubated sometimes wants to be vigorously ventilated, to push back fast against the plummeting oxygen levels. (Then again, maybe that’s just adrenalin surge and panic by doctors and respiratory therapists; maybe the mindless slow-and-steady vent has it right). Three, there had been talk of immediately pronouncing dead any COVID-19 cardiac arrest, given the terrible prognosis — but taken literally, that might mean pushing medications, having a patient flat-line, and then all looking at each other in frightened confusion. Oops? Now what? Are we really not even going to try to fix a cardiac arrest that happens in front of us?

Mallampati scores, and expected corresponding views of vocal cords and trachea (trachea is black triange in Grade 1 view; cords the white rectangles either side).

I had planned to sedate and paralyze him and quickly take over the airway — a so-called rapid-sequence intubation, or RSI. But with O2 sats so low, I wasn’t sure I’d have even 20 seconds to work with. I had him open his mouth wide for me: It was a pretty encouraging view, I could see soft palate and uvula, no obvious edema. If this had been an elective intubation, I would have formally called out, and nursing would have recorded, his Mallampati score — a predictor of how hard it would be, once he was paralyzed, with all of his face and neck and airway muscles loose, to open his mouth wider, press his tongue out of the way with the laryngoscope, and find his vocal cords. I would have scored him a Mallampati 2, which was not bad.

Intubate him without paralysis? With the head of the bed elevated at 15 degrees — a little incline so gravity would help keep stomach contents in the stomach — I could put him under lightly with ketamine, a sedative unique for keeping a patient breathing on his own. I decided I’d try it; it’d buy us many more seconds of the patient working on his own to keep his oxygen levels up.

“Do you want that Plexiglass box?” It was hard to hear each other in the PAPRs — but it was also hard to hear people speaking thru N 95s, and at least now we could see each other’s faces, and lip read.

That was a hard “no.” I couldn’t imagine adding in another variable. What would I do with a box in my way if I needed suction, or help manipulating the larynx to push the vocal cords into view, or if God forbid it all went to hell and I had to perform a crash surgical airway?

Nursing pushed the Ketamine, and I told the patient to think some happy thoughts. He closed his eyes, his face went slack, but he was still breathing, and his oxygen saturation held steady at 63%. I removed his oxygen mask, and slid the plastic laryngoscope blade into his mouth, following his tongue back and looking at a video screen on the handle — it had a partial view of vocal cords. I was used to seeing them standing still in a paralyzed patient, but he was still breathing, and they were opening and closing with a shuddering, 2-step motion. There was no airway edema. (But I did suddenly remember all of my coronavirus reading — the unanimity about always controlling airways with a paralytic on-board, to prevent coughing. Too late.) At this point I called for the paralytic; If I had to say why, I’d say to prevent coughing, to give myself an even better view, to keep him from bucking against the vent air flows, and to still the cords to prevent accidental trauma to them; but even as nursing administered it, and without really waiting for it to take effect, I slid the endotracheal tube home, inflated the balloon to hold it in place, and we hooked him up to the ventilator. The whole process took less than 10 seconds. (We may not have the Wuhan gear, but we’ve got the skills — raise the bed higher, Wuhan, why are you intubating all hunched over and crooked!) His oxygen saturation percentage dipped to the 40s before climbing back up nicely; even more surprising was his end-tidal CO2, which was high — it startled me how high — he’d been tiring out and failing to blow off his ventilatory waste products. He’d been even closer than I thought to respiratory collapse.

Then his blood pressure went to a catastrophic 50s/30s but that was not unexpected — it was dropping for a multitude of reasons, including the sedative medications, the underlying illness, some mild dehydration from insensible losses of an entire day of rapid open-mouth breathing, and the loss of adrenalin tone now that his body could relax and was no longer fighting for every breath. We already had the vasopressor norepinephrine hanging and we titrated it up quickly. His blood pressure rose nicely, his CO2 levels normalized, his oxygen levels were back in the mid-90s, his color improved. He was comfortable and meta-stable: Ready for the ICU.


An hour later, and another routine pandemic-era shift has ended. All PAPR gear has been doffed, cleaned and set aside for the next team. I’ve showered, changed, and am finishing up charting in our ER admin offices. Once those offices were orderly workplaces for people key to our medical practice, like the business manager and the QA coordinator. Now they are empty — the “non-essentials”, that oh-so-insulting term, all work from home. The offices are overflowing with haphazardly-stored supplies — piles of plastic face shields, boxes of isopropyl alcohol bottles, a huge open cardboard box filled with blue and pink Bombas socks (a donation to our group from a well-wisher).

I sat at a computer wedged between socks and isopropyl alcohol boxes to finish up my charting. A Procedure Note for the intubation. A Procedure Note for the central line. A Procedure Note for Critical Care provided. I would be paid separately for each of those, but even after years doing this I honestly don’t know how much. (It varies by insurance, among other things; somewhere between one hundred and two hundred dollars?) My ICU colleagues are already putting in their initial notes, and I take a peek at their thinking; I’ll monitor the patient remotely for the next few days, to see how he does, but I’m not optimistic. Anyway he belongs to ICU now.

It occurs to me that if I’d been mercenary enough, or organized enough, I could have put a formal order in for my lung ultrasound and recorded the images; then I could have billed separately for that as well. Instead, I missed a legitimate opportunity to make more money — off of my patient’s misery? off of his health insurance? — and I’d simply used the ultrasound like a stethoscope. I’m actually relieved I didn’t think about billing for the ultrasound; it’s a reassuring sign that I’m doing the right things for the right reasons. Putting in Procedure Notes that document the care provided but also double as invoices — the billers will use them to demand payment from insurers — is a requirement at every hospital, but has always felt faintly, vaguely wrong. I would prefer a flat salary, frankly, but the healthcare system has insidiously roped me in as a reluctant participant in the “do more things, get more pay” model.

Usually a public health problem is a healthcare system opportunity. Not so this time. Oddly, during a pandemic hospitals turn out to be like restaurants or barber shops — customer-service businesses suddenly without customers. Arguably it’s even worse than that: The restaurants can close their doors to reduce expenses.

Hospitals everywhere have cancelled elective surgeries — a major driver of revenues — to free up beds for COVID-19 surge capacity. With everyone staying at home, no one’s getting in trouble: No car crashes, bar fights, dog bites, work-related back pains. Those few who do need medical care are also often staying away — afraid to come to the hospital, because afraid they’ll catch the coronavirus. So even as we’ve seen dozens of COVID-19 patients every day, the ER’s variety is gone, and it remains eerily half-empty.

More empty than it should be: I’ve seen one person who arrived days after a stroke, and I’ve seen others who’d been suffering terrible misery — agonizing pain of pancreatitis, or of severe peptic ulcers — who tried to gut out 24 or even 48 hours of intractable vomiting, before finally, in reluctant terror, coming to the ER. They were surprised to find it so peaceful, and to be seen in clean rooms with no COVID-19 cases around (we’ve divided our ER into hot and cold zones).

Surgical specialties have seen a collapse of their income. When a surgeon recently declared he’d take a patient emergently to the OR — for a situation that in normal times might be an elective matter — it provoked some raised eyebrows. What’s the emergent indication? “Intractable boat payments,” quipped a colleague.

At least the surgeons get some time off; ER doctors are working as much or more than ever, but with the department half-empty our pay has collapsed as well. The medical practice I work for, like many nationally, has already cut compensation, with more cuts on the horizon. The joke is that’s why we’re stockpiling crates of isopropyl alcohol and blue and pink ankle socks — for the Molotov cocktails, when the front-line providers revolt against the for-profit healthcare system. It’s a half-hearted joke: Everyone knows we’ll just keep coming to work. Meanwhile, the news that ISIS, the Islamic State in Iraq & Syria, is calling on followers to attack U.S. hospitals, because coronavirus, while the U.S. president suggests we try injecting disinfectant to treat it, somehow doesn’t even surprise.



Matt Bivens, MD

Born in DC, studied at UNC-Chapel Hill, now living in Massachusetts. ER physician, EMS medical director, recovering journalist & Russia-watcher.