When I became an attending physician at New York–Presbyterian’s Weill Cornell Hospital last summer, after graduating from a residency at Brigham and Women’s Hospital in Boston, I became a hospitalist: a doctor of general internal medicine who takes care of patients for the duration of their inpatient hospital care. Besides my own writing and research, I also teach medical students and residents in a variety of courses and electives. But for me, as for roughly a million other doctors in the United States, that regular routine is changing very suddenly. Part of the challenge we’re facing with Covid-19 is that even knowing this is an event of shocking magnitude, we are not yet able to measure or foresee exactly how fast the disease’s progress through the world’s population will be.
What doctors call the “natural history” of Covid-19 can be envisioned as four stages; most people’s individual cases will stop at the first or, at most, the second stage, while an unlucky minority will experience the third, or all four. First, there is either an asymptomatic or a mildly symptomatic, nonspecifically “flu-like” illness. In Guangzhou province in China, researchers found that the median incubation period is about four days before symptoms, if any, set in. A portion of those with symptomatic, positive disease then experience a second stage: viral pneumonia, often visible on chest X-rays and CT scans by the varying degrees of inflammation of the lung’s interstitium, the connecting and supportive tissues that line the small airways and blood vessels where oxygen and carbon dioxide are exchanged between blood and air with each breath and heartbeat.
The third stage is a process called acute respiratory distress syndrome (ARDS), a rapid and overwhelming inflammatory response that the body mounts against a perceived foreign invader. ARDS is rare but not specific to Covid-19, and can happen in cases as disparate as near-drownings, blunt trauma, overdoses, and infections of any kind, including sometimes the most innocuous-seeming pathogens. (I once saw a patient die of ARDS after catching a fairly standard case of “mono” in college.) Based on data from China and Italy, about 15–20 percent of Covid-19-positive patients who are sick enough to need the hospital will need intensive care unit-level care for severe interstitial pneumonia or full-blown ARDS.
The fourth, rarest, and most frightening stage of Covid-19 is rapid heart failure and cardiac arrest, often after the pneumonia or ARDS has nearly resolved. While we need more organized research to avoid cognitive biases (which can occur by our fixating on particularly memorable but atypical cases), early reports from front-line providers in disease hotspots suggest that these last cases happen to two broad categories of people: relatively healthy adults of about thirty to fifty years of age, and older patients with cardiovascular disease.
There thus appear to be three main sets of circumstances in which people die from Covid-19. Some people at the first stage of illness could die if their access to health care and other resources is already so precarious that mild infections, plus poor connections to lifelines of caregivers and essential goods, are enough to put their lives in peril: the elderly, people in jails and prisons, and those with food and housing insecurity come first to mind. (Thus far, mortality has been much higher among those over sixty, and particularly in those over eighty years old.) Others may die from symptoms that would be severe but curable under ordinary circumstances—requiring extensive oxygen treatment, or even days or weeks on a ventilator—simply because there aren’t enough supplies and carers to treat them.
These two categories are loosely analogous to the ever-present condition that results in millions of deaths globally because of hypothetically “curable” diseases ranging from pediatric rotavirus diarrhea to hepatitis C and HIV. The Covid-19 version of this syndrome perhaps feels more abhorrent because of the speed of the onslaught, and the shock of its occurrence against the unusual backdrop of societies marked by an exceptional degree of wealth and privilege.
The third kind of death is the end-stage respiratory or cardiovascular failure that sometimes even the best and fullest medical care does not succeed in overcoming. Though smaller in number than the other two categories, these deaths in the intensive care unit, which will continue to include young people and health-care workers, are painful beyond measure to witness. On March 13, three physicians from Milan published a commentary in JAMA describing the overwhelming strain imposed by Covid-19 on intensive care units and health-care workers in northern Italy. Such strain creates an atmosphere of heightened distress, haste, and anxiety that in turn leads to even more deaths from Covid-19.
Italy’s first local case of Covid-19 was a thirty-eight-year-old man admitted on February 20 who rapidly required ICU care for ARDS, though he survived and was discharged three weeks later. Within twenty-four hours, thirty-six other cases were identified in the same region; some of the patients were occupying Lombardy’s seven hundred and twenty ICU beds. Existing ICUs in Lombardy were asked to create zones for isolating Covid-19 patients only, as well as distinct zones where patients with suspected Covid-19 who were awaiting test results could also receive critical care, apart from both the known positive and the known negative cases. With these aims in mind, total ICU-level beds increased by over 50 percent in eighteen days.
However, the latest models of the situation in Lombardy predict two very different scenarios for the coming weeks: one assumes that the incidence of new ICU-level cases is linear; the other that the increase will be exponential, at a growth rate of about 24 percent. If linear, there would be thirty-six new ICU admissions per day. If exponential, the steepest curve estimated about 14,000 new ICU admissions by March 20. Thankfully, the actual curve to date falls short of that latter, more dire estimate, but it still far outstrips the medical capacity of the Italian health-care system.
In the face of these mounting pressures in northern Italy, it became clear that there needed to be protocols for making the difficult decisions that would follow. Accordingly, on March 6, the Italian Society of Anesthesia, Analgesia, Resuscitation, and Intensive Care (SIAARTI) published a document offering recommendations regarding the clinical ethics of intensive care “in exceptional conditions of disequilibrium between needs and available resources.” Contrary to some of the more alarmist rumors on social media about malign and inflexible rationing of care (about which the science journalist Laurie Garrett has written forcefully in The Lancet), these nonbinding guidelines are not particularly prescriptive or specific.
Moreover, they underscore that rationing health-care resources to only some patients is justifiable only after every effort has been made to obtain more resources, or to transfer patients to better-resourced sites. But when those options are exhausted, the guidelines propose “distributive justice” based on prioritizing treatment to patients with “the greatest possibility of therapeutic success” based on their medical history, comorbidities (other health problems), and the likelihood that they will recover. In other words, doctors are advised to use their own judgment in allocating care and resources to those patients most likely to benefit.
By focusing on those who have “higher probability of survival and secondarily to those who can have more years of life saved,” these principles would “maximize benefits for most people.” The SIAARTI guidelines reject the notion of “first come, first served” as untenable, since it excludes patients for no other reason than the vagaries of time and chance.
The guidelines’ specific recommendations include common-sense best practices such as seeking the ethical and practical opinion of other providers or senior staff when making a charged decision, encouraging all patients to make advance directives and clarify their wishes in the event of rapid deterioration—we should all do this; I now have one taped to my refrigerator—and providing absolute transparency and continual communication with patients and families, particularly when resource rationing is involved. (They also emphasize that even a patient deemed ineligible for mechanical ventilation ought to receive the best available care in every other respect.)
While its grim recommendations are particularly timely now, the SIAARTI document is not the first of its kind. Back in 2015, New York State published a lengthier but, in essence, similar set of “Ventilator Allocation Guidelines” in preparation for future influenza pandemics. This document, written by the New York State Task Force on Life & the Law for the state’s Department of Health, and with the aim of “reflect[ing] the values of New Yorkers,” also places a priority on patients “for whom ventilator therapy would most likely be lifesaving.”
When descriptions of the SIAARTI document started appearing in the US media, the tone of the reporting alone could be panic-inducing: a piece in The Atlantic referred to patients not ventilated “gasping for air” and the untreated sick being “left to die.” This language is not quite right: even the most rudimentary or resource-strapped clinical treatment affords palliative care and comfort, and bears witness. We can at least prevent physical suffering. And we are not turning anyone away. In the past weeks, health-care workers who haven’t seen the inside of an ICU, or perhaps even a general ward, since their training are stepping up in droves to prepare to provide the best and most attentive care they can, should they be needed under conditions of staff shortage.
We should also remember that health-care rationing happens continually in the US: organ donation, mental health care, medication-assisted treatment for substance use disorders, shelter beds, and public housing are all forms of health care for which demand far outstrips supply. In a sense, US health care rations even conventional care, such as access to a primary care doctor, based on employment status and insurer (though free clinics and sliding-scale Federally Qualified Health Centers work hard to bring this resource to as many Americans as possible).
In the case of Covid-19, Italian bioethicists commenting on SIAARTI remind us that the guidelines are a template designed to inform doctors’ decision-making, not a mold to stamp identically on all patients. “It is not possible to define a priori criteria such as age, social position, or other arbitrary criteria; the evaluation will have to be done at the patient’s bed, in the singularity of the case,” said Antonio Gioacchino Spagnolo, the director of bioethics at the Catholic University of the Sacred Heart in Rome.
There’s good reason for giving doctors and health-care workers discretion in their practice, because it’s entirely likely that Covid-19 will pose new and different challenges over time. For example, the reason why some patients develop the most grave cases of Covid-19 and some do not is currently unclear. We know that many slightly different strains of the SARS-CoV-2 virus that causes the disease Covid-19 are circulating, and some virologists speculate that the strains vary in virulence and ability to overwhelm the immune system. Other scientists point to differences among host immune systems, or suggest that the virus sometimes reactivates latent viruses encountered earlier in patients’ lives, so that some people’s bodies are forced to fight infections on multiple fronts.
In the meantime, as the situation worsens, my colleagues and I find ourselves engaged in many small forms of resource reallocation. To name one example among hundreds, our hospital’s interview-based research project on homelessness and hospitalization has stalled: it would clearly be irresponsible to put a person suffering homelessness at further risk by asking them to huddle around an audio recorder with health-care workers like us who have potentially been exposed to the virus.
Similar challenges face one colleague’s study on the overlooked work of home health aides, and another colleague’s study on harm reduction centers for patients using IV drugs. Medical students at my hospital have suspended many aspects of their studies but are volunteering after hours as childcare providers to physician families embarking on extra shifts and overtime amid school closure. Many of my coworkers, whose love of clinical practice is matched by their passion for teaching, research, and advocacy, are having to sideline such pursuits. There is just one priority: crisis mitigation.