Severe COVID-19

Definition and classification

In a small percentage of patients, COVID-19 will take a severe course. As seen in the chapter Clinical Presentation (page 297), there was no broadly accepted clinical definition for severe COVID-19 at the beginning of the pandemic. According to a brief and practical definition by IDSA (Infectious Diseases Society of America), severe COVID is defined by “SpO2 ≤94% on room air, including patients on supplemental oxygen” and critical COVID is “mechanical ventilation and ECMO” (Bhimraj ). In this chapter, all symptomatic cases not mild or moderate (i.e. severe and critical) are termed as severe.

Open the references of this chapter in a separate window.

Features, course and outcome

The course of the disease as well as the outcome have changed during the pandemic. Most studies refer to the early pandemic phase and cover severely affected regions and countries. The results vary greatly from country to country and depend on the timing as well. The first data from China revealed shocking numbers at a moment when local epidemics were taking off in Europe (Wu 2020). The spectrum of disease was classified as mild in 81% of cases. In total, 14% were classified as severe, and 5% were critical cases. The case-fatality rate was 14.8% in patients aged ≥ 80 years and 8.0% in patients aged 70-79 years. In a large single-center case study on 344 severe and critically ill patients admitted to Tongji hospital in China from January 25 through February 25, 2020, 133 (38.7%) patients died at a median of 15 days (Wang 2020). Besides older age, hypertension and COPD were more common in non-survivors but not diabetes. No difference was seen between patients with or without ACE inhibitors.

An observational study on 10,021 adult patients with a confirmed COVID-19 diagnosis, who were admitted to 920 hospitals in Germany between 26 February and 19 April 2020 revealed a huge death toll. The median age was 72 years and 1,727 patients (17%) needed mechanical ventilation. Patients on mechanical ventilation had more co-morbidities than patients without mechanical ventilation. Morbidity and mortality were particularly high in older patients, with a considerably lower mortality among patients younger than 60 years (Karagiannidis 2020). Mortality was 52% (906/1,727) in patients being mechanically ventilated, with lower rates reaching 63% of patients aged 70–79 years and 72% of patients aged 80 years and older (Table 1). In-hospital mortality in ventilated patients who were also treated with dialysis was particularly high at 73% (342 of 469), and 71% (84 of 119) of patients on extracorporeal membrane oxygenation (ECMO) died.


Table 1. In house mortality in patients with/without ventilation, percentage of absolute numbers
Without ventilation With ventilation (all types)
18-59 years 0.7% (n = 2474) 27.7% (n = 422)
60-69 years 5.4% (n = 1239) 45.5% (n = 382)
70-79 years 14.6% (n = 1623) 62.6% (n = 535)
≥ 80 years 33.8% (n = 2958) 72.2% (n = 388)


High mortality rates were also seen in other countries. During the early phase of the pandemic, chances of surviving an ICU stay in Lombardia, Italy, were only 50% (Grasselli 2020). In a large cohort study of 3988 critically ill patients, most required invasive mechanical ventilation, and mortality rate was high. In the subgroup of the first 1715 patients, 915 patients died in the hospital for an overall hospital mortality of 53.4%.

The mortality in patients requiring mechanical ventilation was equally large in the New York City Area at the beginning of the pandemic (Richardson 2020). A case series from New York included 5700 COVID-19 patients admitted to 12 hospitals between March 1 and April 4, 2020. Median age was 63 years (IQR 52-75), the most common co-morbidities were hypertension (57%), obesity (42%), and diabetes (34%). At triage, 31% of patients were febrile, 17% had a respiratory rate greater than 24 breaths/minute, and 28% received supplemental oxygen. Of 2634 patients with an available outcome, 14% (median age 68 years, IQR 56-78, 33% female) were treated in ICU, 12% received invasive mechanical ventilation and 21% died. Mortality for those requiring mechanical ventilation was 88.1%.

COVID Reference resources:

Home | Daily Top 10 | 7-Day Overview | Download | Epidemiology | Transmission | Prevention | Virology | Vaccines | Diagnosis | Clinical Manifestations | Treatment | Severe COVID-19 | Comorbidities / Special Populations | Pediatrics| Timeline | Preface

In another study in New York City among 1,150 adults who were admitted to two NYC hospitals with COVID-19 in March, 257 (22%) were critically ill (Cummings 2020). The median age of patients was 62 years (IQR 51-72), 67% were men and 82% patients had at least one chronic illness. As of the end of April, 101 (39%) patients had died and 94 (37%) remained hospitalized. 203 (79%) patients received invasive mechanical ventilation for a median of 18 days, 66% received vasopressors and 31% received renal replacement therapy. In a multivariate Cox model, older age, chronic cardiac disease (adjusted HR 1.76) and chronic pulmonary disease (2.94) were independently associated with in-hospital mortality. This was also seen for higher concentrations of interleukin-6 and D-dimer, highlighting the role of systemic inflammation and endothelial-vascular damage in the development of organ dysfunction.

COVID-19 characteristics may vary considerably by location. In a United States cohort of 2215 adults who were admitted to ICUs at 65 sites, 784 (35.4%) died within 28 days (Gupta 2020). However, mortality showed an extremely wide variation among hospitals, ranging from 6.6% to 80.8%. Factors associated with death included older age, male sex, obesity, coronary artery disease, cancer, acute organ dysfunction, and, importantly, admission to a hospital with fewer intensive care unit beds. Of note, patients admitted to hospitals with fewer than 50 ICU beds versus at least 100 ICU beds had a higher risk of death (OR 3.28; 95% CI, 2.16-4.99).

Another large prospective observational study in the United Kingdom presented clinical data from 20,133 patients, admitted to (or diagnosed in) 208 acute care hospitals in the UK until April 19 (Docherty 2020). Median age was 73 years (interquartile range 58-82) and 60% were men. Co-morbidities were common, namely chronic cardiac disease (31%), diabetes (21%) and non-asthmatic chronic pulmonary disease (18%). Overall, 41% of patients were discharged alive, 26% died, and 34% continued to receive care. 17% required admission to high dependency or intensive care units; of these, 28% were discharged alive, 32% died, and 41% continued to receive care. Of those receiving mechanical ventilation, 17% were discharged alive, 37% died, and 46% remained in hospital. Increasing age, male sex, and co-morbidities including chronic cardiac disease, non-asthmatic chronic pulmonary disease, chronic kidney disease, liver disease and obesity were associated with higher mortality in hospital.

Free Download: 5th Edition

Spotlight: The situation in a German COVID-19 hospital

The Klinik Mühldorf am Inn Hospital was designated as a COVID-19 clinic on March 16, 2020, in order to keep other facilities free for emergencies and elective care. From that day, a total of 276 SARS-CoV-2 positive and 730 suspected cases were treated there. The largest number of symptomatic patients was admitted at the end of March, and the highest number of simultaneously treated SARS-CoV-2 positive patients was 100 patients on April 6, 2020. In total, 18.5% of these in-patients received intensive care during their hospital stay. The peak of intensive care patients was highest on April 10, 2020 with 17 patients. Due to timely preparation, no triage decisions about withholding ventilation treatments had to be made. All COVID-19 patients who had to be treated in the hospital until July 15th, 2020, and who were in need of mechanical ventilation received it. A total of 51 COVID-19 patients required intensive care treatment (18.5% of all COVID-19 in-patients) and 37 patients (13.4%) were ventilated during their intensive care stay. Seven patients were directly intubated and invasively ventilated without a non-invasive ventilation (NIV) attempt after administration of oxygen through a nasal cannula or mask alone. In total, 9/37 patients did not wish to be intubated. In 16 patients, a prone positioning was carried out, including one patient under NIV.

Management and mechanical ventilation

The cardinal COVID-19 symptom leading to intensive care admission is hypoxemic respiratory failure with tachypnea (> 30/min). Initially, in order to protect staff from aerosols as much as possible, intubation and invasive mechanical ventilation was preferred over non-invasive ventilation (NIV) and nasal high-flow (HFNC).

Likewise, due to lack of knowledge and experience, recommendations on how to deal with these patients were not homogeneous, and ARDS ventilation was the preferred technique (Griffiths 2019). According to the ARDS recommendations, patients should be ventilated with a tidal volume (VT) of < 6ml/kg standardized body weight, a peak pressure of < 30 cmH2O and a PEEP based on the ARDS network table.

In one study, these ventilator settings were used except for the lower PEEP/higher FiO2 table. The driving pressure should not exceed 15 mbar. In addition, prone positioning was recommended in case of a PaO2/FiO2 < 150 for more than 16 hours (Ziehr 2020).

Quickly it became obvious that acute respiratory distress syndrome (ARDS) in COVID-19 is not the same as ARDS. COVID-19 in patients with ARDS – CARDs – appears to include an important vascular insult that potentially mandates a different treatment approach than customarily used for ARDS. It may be helpful to categorize patients as having either type L or H phenotype and accept that different ventilatory approaches are needed, depending on the underlying physiology (Marini 2020). In type L (low lung elastance, high compliance, low response to PEEP), infiltrates are often limited in extent and initially characterized by a ground-glass pattern on CT that signifies interstitial rather than alveolar edema. Many patients do not appear overtly dyspneic and may stabilize at this stage without deterioration. Others may transit to a clinical picture more characteristic of typical ARDS: Type H shows extensive CT consolidations, high elastance (low compliance) and high PEEP response. Clearly, types L and H are the conceptual extremes of a spectrum that includes intermediate stages.

Factors and characteristics to develop one type over the other have been identified: severity of the initial infection, the patient’s immune response, the patient’s physical fitness and comorbidities, the response of the hypoxemia to the ventilation, and the time between first symptoms and hospital admission (Gattinoni 2020). L type patients remain stable before improvement or deterioration. In the latter case the patients develop H type pneumonia (Pfeifer 2020). According to this theory, a ventilation strategy starting with respiratory support with high flow oxygen has been recommended (Gattinoni 2020).

To adequately assess oxygenation, the oxygen content (CaO2) in the blood is helpul, as it describes the actual oxygen supply (DO2) better than the oxygen partial pressure (pO2), particularly when combined with the cardiac output (CO):

DO2 = CaO2 x CO    and    CaO2 = Hb x SaO2 x 1.4

With a CaO2 limit of 10 g/100 ml blood, and an appropriate cardiac output, i.e., absence of cardiac failure, a lower O2 saturation (hypoxemia) can be tolerated in the blood before a critical oxygen shortage in the tissue (hypoxia) develops.

Therefore, rather than strictly focusing on pO2 values ​​as represented by the oxygenation index PaO2/FiO2 of < 150, it is more reasonable to consider the overall clinical picture while setting individual target values before intubation. Attempting high-flow oxygen and non-invasive ventilation in patients with type L pneumonia is recommended. Intubation should only be performed if there is significant clinical deterioration (Lyons 2020, Pfeifer 2020).

Special situations in severe COVID-19

Prone positioning

Prone position (PP) has become a therapeutic option, even in awake, non-intubated patients, during spontaneous and assisted breathing (Telias 2020). In one study, among 50 patients, the median SpO2 at triage was 80%. After supplemental oxygen was given to patients on room air it was 84%. After 5 minutes of proning was added, SpO2 improved to 94% (Caputo 2020). Whether PP prevents intubation is not known yet.

In a prospective before-after study in Aix-en-Provence, France among 24 awake, non-intubated, spontaneously breathing patients with COVID-19 and hypoxemic acute respiratory failure requiring oxygen supplementation, the effect of PP was only moderate. 63% were able to tolerate PP for more than 3 hours. Oxygenation increased in only 25% and was not sustained in half of those after resupination. However, prone sessions were short, partly because of limited patient tolerance (Elharrar 2020).

In a small single-center cohort study, use of the prone position for 25 awake, spontaneously breathing patients with COVID-19 was associated with improved oxygenation. In addition, patients with an SpO2 of 95% or greater after 1 hour of the prone position had a lower rate of intubation. Unfortunately, there was no control group and the sample size was very small. Ongoing clinical trials of prone positioning in non–mechanically ventilated patients (NCT04383613, NCT04359797) will hopefully help clarify the role of this simple, low-cost approach for patients with acute hypoxemic respiratory failure (Thompson 2020).

Extracorporal Membrane Oxygenation (ECMO)

Since the beginning of the pandemic, extracorporeal lung replacement procedures such as ECMO have been recommended with caution and only in selected patients with severe and persistant hypoxemia (PaO2/FiO2 < 80), with minor comorbidities and with full usage of all other measures, such as relaxation and recruiting maneuvers (Smereka 2020).

In a single center narrative study regarding ECMO, support for 27 patients with COVID-19 was described (Kon 2020). At the time of the paper submission, survival was 96.3% (one death) in over 350 days of total ECMO support. Thirteen patients (48.1%) remained on ECMO support, while 13 patients (48.1%) were successfully decannulated. Seven patients (25.9%) were discharged from the hospital while six patients (22.2%) remained in the hospital, of which four were on (unmodified) room air. The authors conclude that the judicious use of ECMO support may be clinically beneficial.


During the pandemic, an old problem in a new situation arose: When to perform tracheostomy (and how) in COVID-19 patients? In a review of the current evidence and misconceptions that predispose to uncontrolled variation in tracheostomy among COVID-19 patients, the authors conclude that decisions on tracheostomy must be personalized; that some patients may be awake but cannot yet be extubated (favoring tracheostomy); while others may have immediate, severe hypoxemia when lying supine or with any period of apnea (favoring deferral) (Tay 2020, Schultz 2020). Meanwhile, detailed consensus guidance has been published, including on important issues such as timing of tracheostomy (delayed until at least day 10 of mechanical ventilation and considered only when patients are showing signs of clinical improvement), optimal setting (hierarchic approach to operative location, enhanced PPE), optimal procedure and management after tracheostomy (McGrath 2020).

Lung Transplantation

As in other terminal lung diseases, lung transplantation (LTX) can be a potential therapeutic option. Of course, the indication needs to be considered especially careful. In an editorial published in August 2020, the authors list ten considerations that they believe should be carefully weighed when assessing a patient with COVID-19-associated ARDS regarding potential candidacy for lung transplantation (< 65 years, only single-organ dysfunction, sufficient time for lung recovery, radiological evidence of irreversible lung disease, such as severe bullous destruction or established fibrosis, etc) (Cypel 2020).

Up to now, only case reports have been published. After 52 days of critical COVID-19, ECMO and several complications, a comprehensive interdisciplinary discussion on the direction of treatment resulted in a consensus that the lungs of the otherwise healthy 44-year-old woman from Klagenfurt, Austria had no potential for recovery. On day 58, a suitable donor organ became available, and a sequential bilateral lung transplant was performed. At day 144, the patient remained well. Despite the success of this case, the authors emphasize that lung transplantation is an option for only a small proportion of patients (Lang 2020).

By Markus Unnewehr , Peter Rupp and Matthias Richl