Clinical Presentation

After an average incubation time of around 5 days (range: 2-14 days), a typical COVID-19 infection begins with dry cough and low-grade fever (38.1–39°C or 100.5–102.1°F), often accompanied by diminishment of smell and taste. In most patients, COVID-19 remains mild or moderate and symptoms resolve within a week and patients typically recover at home. Around 10% of patients remain symptomatic through the second week. The longer the symptoms persist, the higher the risk of developing more severe COVID-19, requiring hospitalization, intensive care and invasive ventilation. The outcome of COVID-19 is often unpredictable, especially in older patients with comorbidities. The clinical picture ranges from completely asymptomatic to rapidly devastating courses.

In this chapter we discuss the clinical presentation, including

  • The incubation period
  • Asymptomatic patients
  • Frequent and rare symptoms
  • Laboratory findings
  • Outcome: Risk factors for severe disease
  • Reactivations and reinfections
  • Long-term sequelae

The radiological findings are described in the diagnostic chapter, page 251.

Open the references of this chapter in a separate window.

Incubation period

A pooled analysis of 181 confirmed COVID-19 cases with identifiable exposure and symptom onset windows estimated the median incubation period to be 5.1 days with a 95% CI of 4.5 to 5.8 days (Lauer 2020). The authors estimated that 97.5% of those who develop symptoms will do so within 11.5 days (8.2 to 15.6 days) of infection. Fewer than 2.5% of infected persons will show symptoms within 2.2 days, whereas symptom onset will occur within 11.5 days in 97.5%. However, these estimates imply that, under conservative assumptions, 101 out of every 10,000 cases will develop symptoms after 14 days of active monitoring or quarantine. Another analysis of 158 confirmed cases outside Wuhan estimated a similar median incubation period of 5.0 days (95 % CI, 4.4 to 5.6 days), with a range of 2 to 14 days (Linton 2020). In a detailed analysis of 36 cases linked to the first three clusters of circumscribed local transmission in Singapore, the median incubation period was 4 days with a range of 1-11 days (Pung 2020). Taken together, the incubation period of around 4-6 days is in line with that of other coronaviruses causing SARS or MERS (Virlogeux 2016). Of note, the time from exposure to onset of infectiousness (latent period) may be shorter. There is little doubt that transmission of SARS-CoV-2 during the late incubation period is possible (Li 2020). In a longitudinal study, the viral load was high 2-3 days before the onset of symptoms, and the peak was even reached 0.7 days before the onset of symptoms. The authors of this Nature Medicine paper estimated that approximately 44% (95% CI 25-69%) of all secondary infections are caused by such presymptomatic patients (He 2020).

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

Asymptomatic cases

Understanding the frequency of asymptomatic patients and the temporal course of asymptomatic transmission will be crucial for assessing disease dynamics. It is important to distinguish those patients who will remain asymptomatic during the whole time of infection and those in which infection is still too early to cause symptoms (presymptomatic). While physicians need to be aware of asymptomatic cases, the true percentage is difficult to assess. To evaluate symptoms systematically is not trivial and the ascertainment process could lead to misclassification. If you do not ask precisely enough, you will get false negative answers. If questions are too specific, the interviewees may give false positive answers (confirmation bias). For example, in a large study, only two thirds of patients reporting olfactory symptoms had abnormal results in objective olfactory testing (see below). What is a symptom? And, is it possible to interview the demented residents of a nursing home? Sweet grandma will say she was fine over the last few weeks.

In a living systematic review (through June 10, 2020, analyzing 79 studies in a range of different settings), 20% (95% CI 17%–25%) remained asymptomatic during follow-up, but biases in study designs limit the certainty of this estimate (Buitrago-Garcia 2020). In seven studies of defined populations screened for SARS-CoV-2 and then followed, 31% (95% CI 26%–37%) remained asymptomatic. Another review found that asymptomatic persons seem to account for approximately 40-45% of infections, and that they can transmit the virus to others for an extended period, perhaps longer than 14 days. The absence of COVID-19 symptoms might not necessarily imply an absence of harm as subclinical lung abnormalities are frequent (Oran 2020).

The probable best data come from 3,600 people on board the cruise ship Diamond Princess (Mizumoto 2020) who became involuntary actors in a “well-controlled experiment” where passengers and crew comprised an environmentally homogeneous cohort. Due to insufficient hygienic conditions, > 700 people became infected while the ship was quarantined in the port of Yokohama, Japan. After systematic testing, 328 (51.7%) of the first 634 confirmed cases were found to be asymptomatic. Considering incubation periods between 5.5 and 9.5 days, the authors calculated the true asymptomatic proportion at 17.9% (Mizumoto 2020). The outbreak at the aircraft carrier USS Theodore Roosevelt revealed that 146/736 infected sailors (19.8%) remained asymptomatic for the duration of the study period.


Table 1. Larger studies with defined populations; proportion of asymptomatic patients (LTF = long-term facilities)

Population, n Asymptomatic
Alvarado 2020 Young sailors,
US Aircraft Carrier (n=736)
Borras-Bermejo 2020 Nursing Homes Spain,                      residents (n=768) and staff (n=403) 68% of residents, 56% of staff (including pre-symptomatic)
Feaster 2020 LTFs California,                                residents and staff (n=631) 19-86% of residents, 17-31% of staff
Gudbjartsson 2020 Icelandic Population (n=1,221) 43% (including pre-symptomatic)
Hoxha 2020 LTFs Belgium,
residents (n=4,059) and staff (n=2,185)
75% of residents, 74% of staff (including pre-symptomatic)
Lavezzo 2020 (Small town) Vo, Italy,
all residents (n=2,812)
Marossy 2020 LTFs London (n=2,455) 51% of residents, 69% of staff


There is no doubt that asymptomatic patients may transmit the virus (Bai 2020, Rothe 2020). In several studies from Northern Italy or Korea, viral loads in nasal swabs did not differ significantly between asymptomatic and symptomatic subjects, suggesting the same potential for transmitting the virus (Lee 2020). Of 63 asymptomatic patients in Chongquing, 9 (14%) transmitted the virus to others (Wang Y 2020).

Taken together, these preliminary studies indicate that a significant proportion (20-60%) of all COVID-19 infected subjects may remain asymptomatic during their infection. The studies show a broad range, depending on the populations and probably on methodological issues. It will be very difficult (if not impossible) to clarify the exact proportion.

Free Download: 5th Edition


A plethora of symptoms have been described in the past months, clearly indicating that COVID-19 is a complex disease, which in no way consists only of a respiratory infection. Many symptoms are unspecific so that the differential diagnosis encompasses a wide range of infections, respiratory and other diseases. However, different clusters can be distinguished in COVID-19. The most common symptom cluster encompasses the respiratory system: cough, sputum, shortness of breath, and fever. Other clusters encompass musculoskeletal symptoms (myalgia, joint pain, headache, and fatigue), enteric symptoms (abdominal pain, vomiting, and diarrhea); and less commonly, a mucocutaneous cluster. An excellent review on these extrapulmonary organ-specific pathophysiology, presentations and management considerations for patients with COVID-19 was recently published (Gupta 2020).

Fever, cough, shortness of breath

Symptoms occur in the majority of cases (for asymptomatic patients, see below). In early studies from China (Guan 2020, Zhou 2020), fever was the most common symptom, with a median maximum of 38.3 C; only a few had a temperature of > 39 C. The absence of fever seems to be somewhat more frequent than in SARS or MERS; fever alone may therefore not be sufficient to detect cases in public surveillance. The second most common symptom was cough, occurring in about two thirds of all patients. Among survivors of severe COVID-19 (Zhou 2020), median duration of fever was 12.0 days (8-13 days) and cough persisted for 19 days (IQR 12-23 days). According to a systemic review, including 148 articles comprising 24,410 adults with confirmed COVID-19 from 9 countries (Grant 2020), the most prevalent symptoms were fever (78%), cough (57%) and fatigue (31%).

Fever and cough do not distinguish between mild and severe cases nor do they predict the course of COVID-19 (Richardson 2020, Petrilli 2020). In contrast, shortness of breath has been identified as a strong predictor of severe disease in larger studies. In a cohort of 1,590 patients, dyspnea was associated with an almost two-fold risk for critical disease (Liang 2020) and mortality (Chen 2020). Others found higher rates of shortness of breath, and temperature of > 39.0 in older patients compared with younger patients (Lian 2020). In the Wuhan study on patients with severe COVID-19, a multivariate analysis revealed that a respiratory rate of > 24 breaths per minute at admission was higher in non-survivors (63% versus 16%).

Over the last weeks, much cohort data from countries outside China have been published. However, almost all data applies to patients who were admitted to hospitals, indicating selection bias towards more severe and symptomatic patients.

  • Among 20,133 patients in the UK who were admitted to 208 acute care hospitals in the UK between 6 February and 19 April 2020, the most common symptoms were cough (69%), fever (72%), and shortness of breath (71%), showing a high degree of overlap (Docherty 2020).
  • Among 5,700 patients who were admitted to any of 12 acute care hospitals in New York between March 1, 2020, and April 4, 2020, only 30.7% had fever of > 38C. A respiratory rate of > 24 breaths per minute at admission was found in 17.3% (Richardson 2020).
  • Among the first 1,000 patients presenting at the NewYork Presbyterian/Columbia University (Argenziano 2020), the most common presenting symptoms were cough (73%), fever (73%), and dyspnea (63%).

Musculoskeletal symptoms

The cluster of musculoskeletal symptoms encompasses myalgia, joint pain, headache, and fatigue. These are frequent symptoms, occurring each in 15-40% of patients (Argenziano 2020, Docherty 2020, Guan 2020). Although subjectively very disturbing and sometimes foremost in the perception of the patient, these symptoms tell us nothing about the severity of the clinical picture. However, they are frequently overlooked in clinical practice, and headache merits special attention.

According to a recent review (Bolay 2020), headache is observed in 11-34% of hospitalized COVID-19 patients, occurring in 6-10% as the presenting symptom. Significant features are moderate-severe, bilateral headache with pulsating or pressing quality in the temporo-parietal, forehead or periorbital region. The most striking features are sudden to gradual onset and poor response to common analgesics. Possible pathophysiological mechanisms include activation of peripheral trigeminal nerve endings by the SARS-CoV-2 directly or through the vasculopathy and/or increased circulating pro-inflammatory cytokines and hypoxia.

Gastrointestinal symptoms

Cell experiments have shown that SARS-CoV and SARS-CoV-2 are able to infect enterocytes (Lamers 2020). Active replication has been shown in both bats and human intestinal organoids (Zhou 2020). Fecal calprotectin as a reliable fecal biomarker allowing detection of intestinal inflammation in inflammatory bowel diseases and infectious colitis, was found in some patients, providing evidence that SARS-CoV-2 infection instigates an inflammatory response in the gut (Effenberger 2020). These findings explain why gastrointestinal symptoms are observed in a subset of patients and why viral RNA can be found in rectal swabs, even after nasopharyngeal testing has turned negative. In patients with diarrhea, viral RNA was detected at higher frequency in stool (Cheung 2020).

In the early Chinese studies, however, gastrointestinal symptoms were rarely seen. In a meta-analysis of 60 early studies comprising 4,243 patients, the pooled prevalence of gastrointestinal symptoms was 18% (95% CI, 12%-25%); prevalence was lower in studies in China than other countries. As with otolaryngeal symptoms, it remains unclear whether this difference reflects geographic variation or differential reporting. Among the first 393 consecutive patients who were admitted to two hospitals in New York City, diarrhea (24%), and nausea and vomiting (19%) were relatively frequent (Goyal 2020). Among 18,605 patients admitted to UK Hospitals, 29% of all patients complained of enteric symptoms on admission, mostly in association with respiratory symptoms; however, 4% of all patients described enteric symptoms alone (Docherty 2020).

It’s not all criticial illness. Another study compared 92 critically ill patients with COVID-19–induced ARDS with 92 comparably ill patients with non–COVID-19 ARDS, using propensity score analysis. Patients with COVID-19 were more likely to develop gastrointestinal complications (74% vs 37%; p < 0.001). Specifically, patients with COVID-19 developed more transaminitis (55% vs 27%), severe ileus (48% vs 22%), and bowel ischemia (4% vs 0%). High expression of ACE 2 receptors along the epithelial lining of the gut that act as host-cell receptors for SARS-CoV-2 could explain this (El Moheb 2020).

Otolaryngeal symptoms (including anosmia)

Although upper respiratory tract symptoms such as rhinorrhea, nasal congestion, sneezing and sore throat are relatively unusual, it has become clear within a few weeks that anosmia and hyposmia are important signs of the disease (Luers 2020). Interestingly, these otolaryngological symptoms appear to be much more common in Europe than in Asia. However, it is still unclear whether this is a real difference or whether these complaints were not recorded well enough in the initial phase in China. There is now very good data from Europe: the largest study to date found that 1,754/2,013 patients (87%) reported loss of smell, whereas 1,136 (56%) reported taste dysfunction. Most patients had loss of smell after other general and otolaryngologic symptoms (Lechien 2020). Mean duration of olfactory dysfunction was 8.4 days. Females seem to be more affected than males. The prevalence of self-reported smell and taste dysfunction was higher than previously reported and may be characterized by different clinical forms. Anosmia may not be related to nasal obstruction or inflammation. Of note, only two thirds of patients reporting olfactory symptoms and who had objective olfactory testing had abnormal results.

“Flu plus ‘loss of smell’ means COVID-19”. Among 263 patients presenting in March (at a single center in San Diego) with flu-like symptoms, loss of smell was found in 68% of COVID-19 patients (n=59), compared to only 16% in negative patients (n=203). Smell and taste impairment were independently and strongly associated with SARS-CoV-2 positivity (anosmia: adjusted odds ratio 11, 95% CI: 5‐24). Conversely, sore throat was independently associated with negativity (Yan 2020).

Among a total of 18,401 participants from the US and UK who reported potential symptoms on a smartphone app and who had undergone a SARS-CoV-2 test, the proportion of participants who reported loss of smell and taste was higher in those with a positive test result (65 vs 22%). A combination of symptoms, including anosmia, fatigue, persistent cough and loss of appetite was appropriate to identify individuals with COVID-19 (Menni 2020).

Post-mortem histological analysis of the olfactory epithelium in two COVID-19 patients showed prominent leukocytic infiltrates in the lamina propria and focal atrophy of the mucosa. However, it is unclear whether the observed inflammatory neuropathy is a result of direct viral damage or is mediated by damage to supporting non-neural cells (Kirschenbaum 2020). Among 49 confirmed COVID-19 patients with anosmia, there were no significant pathological changes in the paranasal sinuses on CT scans. Olfactory cleft and ethmoid sinuses appeared normal while in other sinuses, partial opacification was detected only in some cases (Naeini 2020).

Cardiovascular symptoms and issues

There is growing evidence of direct and indirect effects of SARS-CoV-2 on the heart, especially in patients with pre-existing heart diseases (Bonow 2020). SARS-CoV-2 has the potential to infect cardiomyocytes, pericytes and fibroblasts via the ACE2 pathway leading to direct myocardial injury, but the pathophysiological sequence remains unproven (Hendren 2020). Post-mortem examination by in situ hybridization suggested that the most likely localization of SARS-CoV-2 is not in the cardiomyocytes but in interstitial cells or macrophages invading the myocardial tissue (Lindner 2020). A second hypothesis to explain COVID-19-related myocardial injury centers on cytokine excess and/or antibody-mediated mechanisms. It has also been shown that the ACE2 receptor is widely expressed on endothelial cells and that direct SARS-CoV-2 infection of the endothelial cell is possible, leading to diffuse endothelial inflammation (Varga 2020). Post-mortem examination cases indicate a strong virus-induced vascular dysfunction (Menter 2020).

Clinically, COVID-19 can manifest with an acute cardiovascular syndrome (termed “ACovCS”, for acute COVID-19 cardiovascular syndrome). Numerous cases with ACovCS have been described, not only with typical thoracic complaints, but also with very diverse cardiovascular manifestations. Troponin is an important parameter (see below). In a case series of 18 COVID-19 patients who had ST segment elevation, there was variability in presentation, a high prevalence of non-obstructive disease, and a poor prognosis. 6/9 patients undergoing coronary angiography had obstructive disease. Of note, all 18 patients had elevated D-dimer levels (Bangalore 2020). Among 2,736 COVID-19 patients admitted to one of five hospitals in New York City who had troponin-I measured within 24 hours of admission, 985 (36%) patients had elevated troponin concentrations. After adjusting for disease severity and relevant clinical factors, even small amounts of myocardial injury (0.03-0.09 ng/mL) were significantly associated with death (Lala 2020).

In patients with a seemingly typical coronary heart syndrome, COVID-19 should also be considered in the differential diagnosis, even in the absence of fever or cough (Fried 2020, Inciardi 2020). For more information, see the chapter Comorbidities, page 379.

Beside ACovCS, a wide array of cardiovascular manifestations is possible, including heart failure, cardiogenic shock, arrhythmia, and myocarditis. Among 100 consecutive patients diagnosed with COVID-19 infection undergoing complete echocardiographic evaluation within 24 hours of admission, only 32% had a normal echocardiogram at baseline. The most common cardiac pathology was right ventricular (RV) dilatation and dysfunction (observed in 39% of patients), followed by left ventricular (LV) diastolic dysfunction (16%) and LV systolic dysfunction (10%). In another case series of 54 patients with mild-to-moderate COVID-19 in Japan, relative bradycardia was also a common finding (Ikeuchi 2020).

Thrombosis, embolism

Coagulation abnormalities occur frequently in association with COVID-19, complicating clinical management. Numerous studies have reported on an incredibly high number of venous thromboembolism (VTE), especially in those with severe COVID-19. The initial coagulopathy of COVID-19 presents with prominent elevation of D-dimer and fibrin/fibrinogen degradation products, while abnormalities in prothrombin time, partial thromboplastin time, and platelet counts are relatively uncommon (excellent review: Connors 2020). Coagulation test screening, including the measurement of D-dimer and fibrinogen levels, is suggested.

But what are the mechanisms? Some studies have found pulmonary embolism with or without deep venous thrombosis, as well as presence of recent thrombi in prostatic venous plexus, in patients with no history of VTE, suggesting de novo coagulopathy in these patients with COVID-19. Others have highlighted changes consistent with thrombosis occurring within the pulmonary arterial circulation, in the absence of apparent embolism (nice review: Deshpande 2020). Some studies have indicated severe hypercoagulability rather than consumptive coagulopathy (Spiezia 2020) or an imbalance between coagulation and inflammation, resulting in a hypercoagulable state (review: Colling 2020).

According to a systematic review of 23 studies, among 7,178 COVID-19 patients admitted to general wards and intensive care units (ICU), the pooled in-hospital incidence of pulmonary embolism (PE) or lung thrombosis was 14.7% and 23.4%, respectively (Roncon 2020).

Some of the key studies are listed here:

  • In a single-center study from Amsterdam on 198 hospitalized cases, the cumulative incidences of VTE at 7 and 21 days were 16% and 42%. In 74 ICU patients, cumulative incidence was 59% at 21 days, despite thrombosis prophylaxis. The authors recommend performing screening compression ultrasound in the ICU every 5 days (Middeldorp 2020).
  • Among 3334 consecutive patients admitted to 4 hospitals in New York City, a thrombotic event occurred in 16% (Bilaloglu 2020). Of these, 207 (6.2%) were venous (3.2% PE and 3.9% DVT) and 365 (11.1%) were arterial (1.6% ischemic stroke, 8.9% MI, and 1.0% systemic thromboembolism). All-cause mortality was 24.5% and was higher in those with thrombotic events (43% vs 21%). D-dimer level at presentation was independently associated with thrombotic events.
  • In a retrospective multicentre study, 103/1240 (8.3%) consecutive patients hospitalized for COVID-19 (patients directly admitted to an ICU were excluded) had evidence for PE. In a multivariate analysis, male gender, anticoagulation, elevated CRP, and time from symptom onset to hospitalization were associated with PE risk (Fauvel 2020).
  • Autopsy findings from 12 patients, showing that 7/12 had deep vein thrombosis. Pulmonary embolism was the direct cause of death in four cases (Wichmann 2020).
  • Acute pulmonary embolism (APE) can occur in mild-to moderate and is not limited to severe or critical COVID-19 (Gervaise 2020).
  • Careful examination of the lungs from deceased COVID-19 patients with lungs from 7 patients who died from ARDS secondary to influenza A showed distinctive vascular features. COVID-19 lungs displayed severe endothelial injury associated with the presence of intracellular virus and disrupted cell membranes. Histologic analysis of pulmonary vessels showed widespread thrombosis with microangiopathy. Alveolar capillary microthrombi and the amount of vessel growth were 9 and almost 3 times as prevalent as in influenza, respectively (Ackermann 2020)
  • Five cases of large-vessel stroke occurring in younger patients (age 33-49, 2 without any risk factors) (Oxley 2020).
  • Five cases with profound hemodynamic instability due to the development of acute cor pulmonale, among them 4 younger than 65 years of age (Creel-Bulos 2020).

Empiric therapeutic anti-coagulation (AC) is now being employed in clinical practice in many centers and will be evaluated in randomized clinical trials. To adjust for bias due to non-random allocation of potential covariates among COVID-19 patients, one study applied propensity score matching methods. Among > 3000 patients, propensity matching yielded 139 patients who received AC and 417 patients who did not receive treatment with balanced variables between the groups. Results suggest that AC alone is unlikely to be protective for COVID-19-related morbidity and mortality (Tremblay 2020).

There is also a quite controversial debate about a possible correlation between the use of ibuprofen and the increased risk of VTE development. According to a recent review (Arjomandi 2020), the causation between the effects of ibuprofen and VTE remains speculative. The role of ibuprofen on a vascular level remains unclear as well as whether ibuprofen is able to interact with SARS-CoV-2 mechanistically. However, the authors recommend careful considerations on avoiding high dosage of ibuprofen in subjects at particular risk of thromboembolic events.

Neurologic symptoms

Neuroinvasive propensity has been demonstrated as a common feature of human coronaviruses. Viral neuroinvasion may be achieved by several routes, including trans-synaptic transfer across infected neurons, entry via the olfactory nerve, infection of vascular endothelium, or leukocyte migration across the blood-brain barrier (reviews: Zubair 2020, Ellul 2020). With regard to SARS‐CoV‐2, early occurrences such as olfactory symptoms (see above) should be further evaluated for CNS involvement. Potential late neurological complications in cured COVID-19 patients are possible (Baig 2020). In a study of 4491 hospitalized COVID-19 patients in New York City, 606 (13.5%) developed a new neurologic disorder (Frontera 2020). The most common diagnoses were: toxic/metabolic encephalopathy (6.8%, temporary/reversible changes in mental status in the absence of focal neurologic deficits or primary structural brain disease, excluding patients in whom sedative or other drug effects or hypotension explained this), seizure (1.6%), stroke (1.9%), and hypoxic/ischemic injury (1.4%). Whether these more non-specific symptoms are manifestations of the disease itself remains to be seen. There are several observational series of specific neurological features such as Guillain–Barré syndrome (Toscano 2020), myasthenia gravis (Restivo 2020) or Miller Fisher Syndrome and polyneuritis cranialis (Gutierrez-Ortiz 2020).

Especially in patients with severe COVID-19, neurological symptoms are common. In an observational series of 58 patients, ARDS due to SARS-CoV-2 infection was associated with encephalopathy, prominent agitation and confusion, and corticospinal tract signs. Patients with COVID-19 might experience delirium, confusion, agitation, and altered consciousness, as well as symptoms of depression, anxiety, and insomnia (review: Rogers 2020). It remains unclear which of these features are due to critical illness–related encephalopathy, cytokines, or the effect or withdrawal of medication, and which features are specific to SARS-CoV-2 infection (Helms 2020). However, in a large retrospective cohort study comparing 1,916 COVID-19 patients and 1,486 influenza patients (with emergency department visits or hospitalizations), there were 31 acute ischemic strokes with COVID-19, compared to 3 with influenza (Merkler 2020). After adjustment for age, sex, and race, the likelihood of stroke was almost 8-fold higher with COVID-19 (odds ratio, 7.6).

Of note, there is no clear evidence for CNS damage directly caused by SARS-CoV-2. In a study on 21 cerebrospinal fluid (CSF) samples from patients with confirmed COVID-19, all were negative. These data suggest that, although SARS-CoV-2 is able to replicate in neuronal cells in vitro, SARS-CoV-2 testing in CSF is not relevant in the general population (Destras 2020). In a large post-mortem examination, SARS-CoV-2 could be detected in the brains of 21 (53%) of 40 examined patients but was not associated with the severity of neuropathological changes (Matschke 2020) which seemed to be mild, with pronounced neuroinflammatory changes in the brainstem being the most common finding. In another study, brain specimens obtained from 18 patients who died 0 to 32 days after the onset of symptoms showed only hypoxic changes and did not show encephalitis or other specific brain changes referable to the virus (Solomon 2020).

Dermatological symptoms

Numerous studies have reported on cutaneous manifestations seen in the context of COVID-19. The most prominent phenomenon, the so-called “COVID toes”, are chilblain-like lesions which mainly occur at acral areas. [Chilblain: Frostbeule (de), engelure (fr), sabañón (es), gelone (it), frieira (pt), 冻疮 (cn)] These lesions can be painful (sometimes itchy, sometimes asymptomatic) and may represent the only symptom or late manifestations of SARS-CoV-2 infection. Of note, in most patients with “COVID toes”, the disease is only mild to moderate. It is speculated that the lesions are caused by inflammation in the walls of blood vessels, or by small micro-clots in the blood. However, whether “COVID toes” represent a coagulation disorder or a hypersensitivity reaction is not yet known. Key studies:

  • Two different patterns of acute acro-ischemic lesions can overlap (Fernandez-Nieto 2020). The chilblain-like pattern was present in 95 patients (72.0%). It is characterized by red to violet macules, plaques and nodules, usually at the distal aspects of toes and fingers. The erythema multiform-like pattern was present in 37 patients (28.0%).
  • Five clinical cutaneous lesions are described (Galvan 2020): acral areas of erythema with vesicles or pustules (pseudo-chilblain) (19%), other vesicular eruptions (9%), urticarial lesions (19%), maculopapular eruptions (47%) and livedo or necrosis (6%). Vesicular eruptions appear early in the course of the disease (15% before other symptoms). The pseudo-chilblain pattern frequently appears late in the evolution of COVID-19 disease (59% after other symptoms).
  • In a case series on 22 adult patients with varicella-like lesions (Marzano 2020), typical features were constant trunk involvement, usually scattered distribution and mild or absent pruritus, the latter being in line with most viral exanthems but not like true varicella. Lesions generally appeared 3 days after systemic symptoms and disappeared by day 8.
  • Three cases of COVID-19 associated ulcers in the oral cavity, with pain, desquamative gingivitis, and blisters (Martin Carreras-Presas 2020).

Other case reports include digitate papulosquamous eruption (Sanchez 2020), petechial skin rash (Diaz-Guimaraens 2020, Quintana-Castanedo 2020). However, it should be kept in mind that not all rashes or cutaneous manifestations seen in patients with COVID-19 can be attributed to the virus. Co-infections or medical complications have to be considered. Newer studies reporting in negative PCR and serology have questioned a direct association between acral skin disease and COVID-19:

  • Of 31 patients (mostly teenagers) who had recently developed chilblains, histopathologic analysis of skin biopsy specimens (22 patients) confirmed the diagnosis of chilblains and showed occasional lymphocytic or microthrombotic phenomena. In all patients, PCR and serology remained negative (Herman 2020).
  • Among 40 young patients with chilblain lesions and with suspected SARS-CoV-2 infection, serology was positive in 12 (30%). All had negative PCR results at the time of presentation, suggesting that in young patients SARS-CoV-2 is completely suppressed before a humoral immune response is induced (Hubiche 2020).
  • In a cohort series from Valencia following 20 patients aged 1 to 18 years with new-onset acral inflammatory lesions, all lacked systemic manifestations of COVID-19. Surprisingly, both PCR and serologic test results were negative for SARS-CoV-2 (Roca-Ginés 2020).

Comprehensive mucocutaneous examinations, analysis of other systemic clinical features or host characteristics, and histopathologic correlation, will be vital to understanding the pathophysiologic mechanisms of what we are seeing on the skin (Review: Madigan 2020).


SARS-CoV-2 has an organotropism beyond the respiratory tract, including the kidneys and the liver. Researchers have quantified the SARS-CoV-2 viral load in precisely defined kidney compartments obtained with the use of tissue micro-dissection from 6 patients who underwent autopsy (Puelles 2020). Three of these 6 patients had a detectable SARS-CoV-2 viral load in all kidney compartments examined, with preferential targeting of glomerular cells. Renal tropism is a potential explanation of commonly reported clinical signs of kidney injury in patients with COVID-19, even in patients with SARS-CoV-2 infection who are not critically ill (Zhou 2020). Recent data indicate that renal involvement is more frequent than described in early studies (Gabarre 2020). Of the first 1,000 patients presenting at the NewYork-Presbyterian/Columbia University, 236 were admitted or transferred to intensive care units (Argenziano 2020). Of these, 78.0% (184/236) developed acute kidney injury and 35.2% (83/236) needed dialysis. Concomitantly, 13.8% of all patients and 35.2% of patients in intensive care units required in-patient dialysis, leading to a shortage of equipment needed for dialysis and continuous renal replacement therapy.

In recent months, some case reports of collapsing glomerulopathy akin to those seen during the HIV epidemic have been published. All of these cases were in patients of African ethnicity (Velez 2020).


One of the largest studies, evaluating liver injury in 2273 SARS-CoV-2 positive patients, found that 45% had mild, 21% moderate, and 6.4% severe liver injury. In a multivariate analysis, severe acute liver injury was significantly associated with elevated inflammatory markers including ferritin and IL‐6. Peak ALT was significantly associated with death or discharge to hospice (OR 1.14, p = 0.044), controlling for age, body mass index, diabetes, hypertension, intubation, and renal replacement therapy (Phipps 2020). In another meta-analysis of 9 studies with a total of 2115 patients, patients with COVID-19 with liver injury were at an increased risk of severity (OR 2.57) and mortality (1.66).

Ocular and atypical manifestations

Ocular manifestations are also common. In a case series from China, 12/38 patients (32%, more common in severe cases) had ocular manifestations consistent with conjunctivitis, including conjunctival hyperemia, chemosis, epiphora, or increased secretions. Two patients had positive PCR results from conjunctival swabs (Wu 2020). The retina can also be affected, as has been shown using optical coherence tomography (OCT), a non-invasive imaging technique that is useful for demonstrating subclinical retinal changes. Twelve adult patients showed hyper-reflective lesions at the level of the ganglion cell and inner plexiform layers more prominently at the papillomacular bundle in both eyes. Since their initial report, the authors have extended their findings to more than 150 patients, demonstrating an absence of blood flow within the retinal lesions of “many” patients (Marinho 2020).

Other new and sometimes puzzling clinical presentations have emerged (and will emerge) in the current pandemic. There are case reports of non-specific symptoms, especially in the elderly population, underlining the need for extensive testing in the current pandemic (Nickel 2020).

Laboratory findings

The most evident laboratory findings in the first large cohort study from China (Guan 2020) are shown in Table 2. On admission, lymphocytopenia was present in 83.2% of the patients, thrombocytopenia in 36.2%, and leukopenia in 33.7%. In most patients, C-reactive protein was elevated to moderate levels; less common were elevated levels of alanine aminotransferase and D-dimer. Most patients have normal procalcitonin on admission.


Table 2. Percentage of symptoms in first large cohort study from China (Guan 2020). Disease severity was classified according to American Thoracic Society (Metlay 2019) guidelines
Clinical symptoms All Severe Disease Non-
Fever, % 88.7 91.9 88.1
Cough, % 67.8 70.5 67.3
Fatigue, % 38.1 39.9 37.8
Sputum production, % 33.7 35.3 33.4
Shortness of breath, % 18.7 37.6 15.1
Myalgia or arthralgia, % 14.9 17.3 14.5
Sore throat, % 13.9 13.3 14.0
Headache, % 13.6 15.0 13.4
Chills, % 11.5 15.0 10.8
Nausea or vomiting, % 5.0 6.9 4.6
Nasal congestion, % 4.8 3.5 5.1
Diarrhea, % 3.8 5.8 3.5
Radiological findings
Abnormalities on X-ray, % 59.1 76.7 54.2
Abnormalities on CT, % 86.2 94.6 84.4
Laboratory findings
WBC < 4,000 per mm3, % 33.7 61.1 28.1
Lymphocytes < 1,500 per mm3, % 83.2 96.1 80.4
Platelets < 150,000 per mm3, % 36.2 57.7 31.6
C-reactive protein ≥ 10 mg/L, % 60.7 81.5 56.4
LDH ≥ 250 U/L, % 41.0 58.1 37.1
AST > 40 U/L, % 22.2 39.4 18.2
D-dimer ≥ 0.5 mg/L, % 46.6 59.6 43.2



Parameters indicating inflammation such as elevated CRP and procalcitonin are very frequent findings. They have been proposed to be important risk factors for disease severity and mortality (Chen 2020). For example, in a multivariate analysis of a retrospective cohort of 1590 hospitalized subjects with COVID-19 throughout China, a procalcitonin > 0.5 ng/ml at admission had a HR for mortality of 8.7 (95% CI: 3.4-22.3). In 359 patients, CRP performed better than other parameters (age, neutrophil count, platelet count) in predicting adverse outcome. Admission serum CRP level was identified as a moderate discriminator of disease severity (Lu 2020). Of 5279 cases confirmed in a large medical center in New York, 52% of them admitted to hospital, a CRP > 200 was more strongly associated (odds ratio 5.1) with critical illness than age or comorbidities (Petrilli 2019).

Some studies have suggested that the dynamic change of interleukin-6 (IL-6) levels and other cytokines can be used as a marker in disease monitoring in patients with severe COVID-19 (Chi 2020, Zhang 2020). In a large study of 1484 patients, several cytokines were measured upon admission to the Mount Sinai Health System in New York (Del Valle 2020). Even when adjusting for disease severity, common laboratory inflammation markers, hypoxia and other vitals, demographics, and a range of comorbidities, IL-6 and TNF-α serum levels remained independent and significant predictors of disease severity and death. These findings were validated in a second cohort of 231 patients. The authors propose that serum IL-6 and TNF-α levels should be considered in the management and treatment of patients with COVID-19 to stratify prospective clinical trials, guide resource allocation and inform therapeutic options.

There is also one study suggesting that serum cortisol concentration seems to be a better independent predictor than other laboratory markers associated with COVID-19, such as CRP, D-dimer, and neutrophil to leukocyte ratio (Tan 2020).

Hematological: Lymphocytes, platelets, RDW

Lymphocytopenia and transient but severe T cell depletion is a well-known feature of SARS (He 2005). In COVID-19, lymphopenia is also among the most prominent hematological features. Lymphopenia may be predictive for progression (Ji 2020) and patients with severe COVID-19 present with lymphocytopenia of less than 1500/µl in almost 100% of cases (Guan 2020). It’s not only the total lymphocyte count. There is growing evidence for a transient depletion of T cells. Especially the reduced CD4+ and CD8+ T cell counts upon admission were predictive of disease progression in a larger study (Zhang 2020). In another large study on COVID-19 patients, CD3+, CD4+ and CD8+ T cells as well as NK cells were significantly decreased in COVID-19 patients and related to the severity of the disease. According to the authors, CD8+ T cells and CD4+ T cell counts can be used as diagnostic markers of COVID-19 and predictors of disease severity (Jiang 2020). Beside T cells, B cells may also play a role. In 104 patients, a decrease in B cells was independently associated with prolonged viral RNA shedding (Hao 2020).

Another common hematological finding is low platelet counts that may have different causes (Review: Xu 2020). A meta-analysis of 24 studies revealed a weighted incidence of thrombocytopenia in COVID-19 patients of 12.4% (95% CI 7.9%–17.7%). The meta-analysis of binary outcomes (with and without thrombocytopenia) indicated an association between thrombocytopenia and a 3-fold enhanced risk of a composite outcome of ICU admission, progression to acute respiratory distress syndrome, and mortality (Zong 2020). Cases of hemorrhagic manifestation and severe thrombocytopenia responding to immunoglobulins fairly quickly with a sustained response over weeks have been reported (Ahmed 2020).

Red blood cell distribution width (RDW) is another component of complete blood counts that quantifies the variation of individual red blood cell (RBC) volumes and has been shown to be associated with elevated risk for morbidity and mortality in a wide range of diseases. In a large cohort study including 1641 adults diagnosed with SARS-CoV-2 infection and admitted to 4 hospitals in Boston (Foy 2020), RDW was associated with mortality risk in Cox models (hazard ratio of 1.09 per 0.5% RDW increase and 2.01 for an RDW > 14.5% vs ≤ 14.5%).

However, there are also cohorts in which hematological parameters such as thrombocytes, neutrophil-to-lymphocyte ratio or D-dimers do not allow prediction of patient outcome (Pereyra 2020). These routine parameters, despite giving guidance on the overall health of the patient, might not always accurately indicate COVID-19-related complications.

Cardiac: Troponin

Given the cardiac involvement especially in severe cases (see above), it is not surprising that cardiac parameters are frequently elevated. A meta-analysis of 341 patients found that cardiac troponin I levels are significantly increased only in patients with severe COVID-19 (Lippi 2020). In 179 COVID-19 patients, cardiac troponin ≥ 0.05 ng/mL was predictive of mortality (Du 2020). Among 2736 COVID-19 patients admitted to one of five hospitals in New York City who had troponin-I measured within 24 hours of admission, 985 (36%) patients had elevated troponin concentrations. After adjusting for disease severity and relevant clinical factors, even small amounts of myocardial injury (0.03-0.09 ng/mL) were significantly associated with death (adjusted HR: 1.75, 95% CI 1.37-2.24) while greater amounts (> 0.09 ng/dL) were significantly associated with higher risk (adjusted HR 3.03, 95% CI 2.42-3.80). However, it remains to be seen whether troponin levels can be used as a prognostic factor. A comprehensive review on the interpretation of elevated troponin levels in COVID-19 has been recently published (Chapman 2020).

Coagulation: D-dimer, aPTT

Several studies have evaluated the coagulation parameter D-dimer in the progression of COVID-19. Among 3334 consecutive patients admitted to 4 hospitals at New York City, a thrombotic event occurred in 16.0%. D-dimer level at presentation was independently associated with thrombotic events, consistent with early coagulopathy (Bilaloglu 2020). In the Wuhan study, all patients surviving had low D-dimer during hospitalization, whereas levels in non-survivors tended to increase sharply at day 10. In a multivariate analysis, D-dimer of > 1 µg/mL remained the only lab finding which was significantly associated with in-hospital death, with an odds ratio of 18.4 (2.6-129, p = 0.003). However, D-dimer has a reported association with mortality in patients with sepsis and many patients died from sepsis (Zhou 2020).

In a considerable proportion of patients, a prolonged aPTT can be found. Of 216 patients with SARS-CoV-2, this was the case in 44 (20%). Of these, 31/34 (91%) had positive lupus anticoagulant assays. As this is not associated with a bleeding tendency, it is recommended that prolonged aPTT should not be a barrier to the use of anti-coagulation therapies in the prevention and treatment of venous thrombosis (Bowles 2020). Another case series of 22 patients with acute respiratory failure present a severe hypercoagulability rather than consumptive coagulopathy. Fibrin formation and polymerization may predispose to thrombosis and correlate with a worse outcome (Spiezia 2020).

Lab findings as risk factor

It is not very surprising that patients with severe disease had more prominent laboratory abnormalities than those with non-severe disease. It remains unclear how a single parameter can be of clinical value as almost all studies were retrospective and uncontrolled. Moreover, the numbers of patients were low in many studies. However, there are some patterns which may be helpful in clinical practice. Lab risk factors are:

  • Elevated CRP, procalcitonin, interleukin-6 and ferritin
  • Lymphocytopenia, CD4 T cell and CD8 T cell depletion, leukocytosis
  • Elevated D-dimer and troponin
  • Elevated LDH

Clinical classification

There is no broadly accepted or valid clinical classification for COVID-19. The first larger clinical study distinguished between severe and non-severe cases (Guan 2020), according to the Diagnosis and Treatment Guidelines for Adults with Community-acquired Pneumonia, published by the American Thoracic Society and Infectious Diseases Society of America (Metlay 2019). In these validated definitions, severe cases include either one major criterion or three or more minor criteria. Minor criteria are a respiratory rate > 30 breaths/min, PaO2/FIO2 ratio < 250, multilobar infiltrates, confusion/disorientation, uremia, leukopenia, low platelet count, hypothermia, hypotension requiring aggressive fluid resuscitation. Major criteria comprise septic shock with need for vasopressors or respiratory failure requiring mechanical ventilation.

Some authors (Wang 2020) have used the following classification including four categories:

  1. Mild cases: clinical symptoms were mild without pneumonia manifestation through image results
  2. Ordinary cases: having fever and other respiratory symptoms with pneumonia manifestation through image results
  3. Severe cases: meeting any one of the following: respiratory distress, hypoxia (SpO2 ≤ 93%), abnormal blood gas analysis: (PaO2 < 60mmHg, PaCO2 > 50mmHg)
  4. Critical cases: meeting any one of the following: Respiratory failure which requires mechanical ventilation, shock, accompanied by other organ failure that needs ICU monitoring and treatment.

In the report of the Chinese CDC, estimation of disease severity used almost the same categories (Wu 2020) although numbers 1 and 2 were combined. According to the report, there were 81% mild and moderate cases, 14% severe cases and 5% critical cases. There are preliminary reports from the Italian National Institute of Health, reporting on 24.9% severe and 5.0% critical cases (Livingston 2020). However, these numbers are believed to strongly overestimate the disease burden, given the very low number of diagnosed cases in Italy at the time. Among 7,483 US heath care workers with COVID-19, a total of 184 (2.1–4.9%) had to be admitted to ICUs. Rate was markedly higher in HCWs > 65 years of age, reaching 6.9–16.0% (CDC 2020).


We are facing rapidly increasing numbers of severe and fatal cases in the current pandemic. The two most difficult but most frequently asked clinical questions are 1. How many patients end up with severe or even fatal courses of COVID-19? 2. What is the true proportion of asymptomatic infections? We will learn more about this shortly through serological testing studies. However, it will be important that these studies are carefully designed and carried out, especially to avoid bias and confounding.

Case fatality rates (CFR)

The country-specific crude case fatality rates (CFR), the percentage of COVID-19-associated deaths among confirmed SARS-CoV-2 infections, have been the subject of much speculation. There are still striking differences between countries. According to assessed on October 12, 2020, the crude CFR between the 100 most affected countries (in terms of absolute numbers) ranged from 0.05 (Singapore) to 10.2 (Mexico). Within the 10 most affected countries in Europe, the CFR range was between 0.8% (Czechia) and 10.2% (Italy).

Although it is well known that the CFR of a disease can be biased by detection, selection or reporting (Niforatos 2020), and although it became quickly clear that older age is a major risk factor for mortality (see below), many other factors contributing to regional differences throughout the world have been discussed in recent months. These factors include not only differences in the overall age structure of the general population of a country and co-residence patterns, but also co-morbidity burden, obesity prevalence and smoking habits as well as societal and social psychological factors. Others include heterogeneity in testing and reporting approaches, variations in health care system capacities and health care and even political regime. Different virus strains or even environmental factors such as air pollution have also been discussed, as well as potential differences in genetic variability or even “trained immunity” induced by certain live vaccines such as bacillus Calmette-Guérin (for references see Hoffmann C 2020).

We can probably exclude most of these speculations. SARS-CoV-2 is not deadlier in Italy (CFR 10.2%), United Kingdom (7.1%), or Sweden (6.0%), compared to Slovakia (0.3%), Israel (0.8%), India (1.5%) or USA (2.7%). Instead, there are three major factors that have to be taken into account:

  • The age of the pandemic, especially of the population which is first affected. Data from the 20 most affected European countries and the USA and Canada show that the variance of crude CFR of COVID-19 is predominantly (80-96%) determined by the proportion of older individuals who are diagnosed with SARS-CoV-2 (Hoffmann C 2020). Of note, the age distribution of SARS-CoV-2 infections is still far from homogeneous. The proportion of individuals older than 70 years among confirmed SARS-CoV-2 cases still differs markedly between the countries, ranging from 5% to 40% (Figure 1).

Countries’ testing policies (and capacities). The fewer people you test (all people, only symptomatic patients, only those with severe symptoms), the higher the mortality. In Germany, for example, testing systems and high lab capacities were established rapidly, within weeks in January (Stafford 2020).

  • Stage of the epidemic. Some countries have experienced their first (or second) waves early while others lagged a few days or weeks behind. Death rates only reflect the infection rate of the previous 2 to 4 weeks.

There is no doubt that the marked variation of CFR across countries will diminish over time, for example, if less affected countries such as Korea or Singapore fail to protect their older age groups; or if countries with high rates at the beginning (such as Italy, Belgium or Sweden) start to implement broad testing in younger age groups. This process has already begun. In Belgium, for example, CFR peaked on May 11 with an appalling rate of 16.0%; it has now dropped to 6.3%. The CFR in the USA peaked on May 16 (6.1%) and is now less than half that. Germany started with a strikingly low CFR of 0.2% by the end of March (prompting much attention even in scientific papers), peaked on June 18 (4.7%) and is now (October 10) at 3.0%.


Figure 1. Association between case fatality rate (CFR) and the proportion of persons over 75 years of age among all confirmed SARS-CoV-2 cases (R2=0.8034, p<0.0001). The circle sizes reflect the country-specific numbers of COVID-19 associated deaths per million habitants; the linear fit prediction plot with a 95% confidence interval was estimated by weighted linear regression (weight = total number of COVID-19 associated deaths).

CFR among health care workers, well-defined populations

In well-monitored populations in which under-reporting is unlikely or can be largely determined, the mortality rates may better reflect the “true” CFR of COVID-19. This applies to healthcare workers (HCW) but also to populations of “well-defined” (limited) outbreaks and in populations with available serology data. The low mortality rates in these populations are remarkable.

  • In a large study of 3387 HCW from China infected with SARS-CoV-2, only 23 died, corresponding to a mortality rate of 0.68%. The median age was 55 years (range, 29 to 72), and 11 of the 23 deceased HCW had been reactivated from retirement (Zhan 2020). Current studies in the US have found similar mortality estimates of 0.3-0.6% (CDC 2020). Of the 27 HCW who died from COVID-19 until mid-April, 18 were over 54 years of age. The overall low mortality rates were probably due to the fact that HCWs were younger and healthier, but also that they had been tested earlier and more frequently.
  • On the cruise ship Diamond Princess, as of May 31, the total number of infected reached 712, and 13 patients died from the disease leading to a CFR of 1.8% (Moriarty 2020). Of note, around 75% of the patients on the Diamond Princess were 60 years or older, many of them in their eighties. Projecting the Diamond Princess CFR onto the age structure of the general population, mortality would be in a range of 0.2-0.4%.
  • According to an investigation of the shore-based USS Theodore Roosevelt outbreak, only 6/736 infected sailors were hospitalized, and one (a “senior listed member in his 40s”) died during the study period (CFR 0.1%) (Alvarado 2020).
  • Using population-based seroprevalences in Geneva (Switzerland) and after accounting for demography, the population-wide infection fatality rate (IFR) was 0.64% (0.38–0.98) (Perez-Saez 2020).

CFR compared to influenza

More time and data are needed before the COVID-19 pandemic can be accurately compared with past pandemics. But what makes SARS-CoV-2 different from pandemic influenza virus? It’s not only that SARS-CoV-2 is a new pathogen and influenza is not and that the diseases differ clinically. The picture is more complex. It also depends on which flu season you are talking about – the influenza pandemic excess mortality ranged from extreme (1918) to mild (2009) over the past 100 years. Another key difference between SARS-CoV-2 and pandemic influenza is the age distribution of patients who are severely ill. Mortality due to SARS-CoV-2 and SARS-CoV is strongly skewed towards people older than 70 years, very dissimilar to the 1918 and 2009 influenza pandemics.

Pooled estimates of all-cause mortality for 24 European countries for the period March–April 2020 showed that excess mortality of COVID-19 particularly affected ≥ 65-year-olds (91% of all excess deaths) and to a far lesser extent those 45–64 (8%) and 15–44-year-olds (1%) (Vestergaard 2020). The excess mortality of COVID-19 is markedly higher than for major influenza pandemics in the past. For example, the 2009 pandemic influenza A H1N1 globally led to 201,200 respiratory deaths (range 105,700-395,600) with an additional 83,300 cardiovascular deaths (Dawood 2012). This is by far lower than the deaths caused by COVID-19 to date. According to a recent review, the population risk of admission to the intensive care unit is five to six times higher in patients infected with SARS-CoV-2 than in those with the fairly mild 2009 influenza pandemic (Petersen 2020).

In New York City, a study analyzed standardized mortality ratios (SMR) of comparator pandemics and epidemics relative to the first 2020 wave of COVID-19 (Muscatello 2020). In older people, COVID-19 mortality until June 2020 was more than 10-fold higher than a severe influenza season, and more than 300-fold higher than the 2009–10 influenza pandemic. Compared to the catastrophic 1918–19 winter wave of the influenza pandemic, there are marked differences for different age groups. The 1918-19 influenza had a high mortality, especially in younger persons (5–15 years; ~25% of total deaths), possibly due to antibody-dependent enhancement and ‘cytokine storms’ in younger people but also due to some protective cross-immunity from previous influenza outbreaks among those older. Compared to COVID-19, the overall age-adjusted, all-age mortality rate of the influenza 1918-19 was 6.7 times higher. In younger people (< 45 years), the SMR was 42; that is, 42 times higher for influenza in 1918–19 than for COVID-19. However, in people older than 44 years of age, the SMR was 0.56; that is, 44% lower in 1918–19 than for COVID-19.

Modeling scenarios without appropriate mitigation measures, simulations predict incredibly high peaks in active cases and alarmingly high numbers of deaths far into the future. In Germany, for example, 32 million total infections would result in 730,000 deaths over the course of the epidemic, which would seem to occur only by the end of the summer 2021 under the assumption that no reliable treatment is available before then (Barbarossa 2020).

Older Age

From the beginning of the epidemic, older age has been identified as an important risk factor for disease severity (Huang 2020, Guan 2020). In Wuhan, there was a clear and considerable age dependency in symptomatic infections (susceptibility) and outcome (fatality) risks, by multiple folds in each case (Wu 2020). The summarizing report from the Chinese CDC found a death rate of 2.3%, representing 1023 among 44,672 confirmed cases (Wu 2020). Mortality increased markedly in older people. In cases aged 70 to 79 years, CFR was 8.0% and cases in those aged 80 years older had a 14.8% CFR. There is now growing data from serology-informed estimates that the same is true for the infection fatality risk (IFR). After accounting for demography and age-specific seroprevalence, IFR was 0.0092% (95% CI 0.0042–0.016) for individuals aged 20–49 years, 0.14% (0.096–0.19) for those aged 50–64 years but 5.6% (4.3–7.4) for those aged 65 years and older (Perez-Saez 2020).

In recent months, these data have been confirmed by almost all studies published throughout the world. In almost all countries, age groups of 60 years or older contribute to more than 90% of all death cases.

  • In a large registry analysing the epidemic in the UK in 20,133 patients, the median age of the 5165 patients (26%) who died in hospital from COVID-19 was 80 years (Docherty 2020).
  • Among 1591 patients admitted to ICU in Lombardy, Italy, older patients (> 63 years) had markedly higher mortality than younger patients (36% vs 15%). Of 362 patients older than 70 years of age, mortality was 41% (Grasselli 2020).
  • According to the Italian National Institute of Health, an analysis of the first 2003 death cases, median age was 80.5 years. Only 17 (0.8%) were 49 years or younger, and 88% were older than 70 years (Livingston 2020).
  • Detailed analysis of all-cause mortality at Italian hot spots showed that the deviation in all-cause deaths compared to previous years during epidemic peaks was largely driven by the increase in deaths among older people, especially in men (Piccininni 2020, Michelozzi 2020).
  • In 5700 patients admitted to New York hospitals, there was a dramatic increase of mortality among older age groups, reaching 61% (122/199) in men and 48% (115/242) in women over 80 years of age (Richardson 2020).
  • The median age of 10,021 adult COVID-19 patients admitted to 920 German hospitals was 72 years. Mortality was 53% in patients being mechanically ventilated (n=1727), reaching 63% in patients aged 70–79 years and 72% in patients aged 80 years and older (Karagiannidis 2020).
  • In an outbreak reported from King County, Washington, a total of 167 confirmed cases was observed in 101 residents (median age 83 years) of a long-term care facility, in 50 healthcare workers (HCW, median age 43 years), and 16 visitors. The case fatality rate was 33.7% among residents and 0% among HCW (McMichael 2020).

There is no doubt that older age is by far the most important risk factor for mortality. Countries failing to protect their elderly population for different reasons (such as Italy, Belgium or Sweden) are facing a higher CFR, while those without many older patients infected by SARS-CoV-2 (such as the Republic of Korea, Singapore, Australia) have markedly lower rates.

What are the reasons? Severe endothelial injury as seen in critically ill patients (Ackermann 2020) and endotheliopathy is an essential part of the pathological response to severe COVID-19, leading to respiratory failure, multi-organ dysfunction and thrombosis (Goshua 2020). Circulating endothelial cells are a marker of endothelial injury in severe COVID-19 (Guervilly 2020) and there is a direct and rapid cytotoxic effect of plasma collected from critically ill patients on vascular endothelial cells (Rauch 2020). It is therefore tempting to speculate that endothelial injury will be particularly harmful in older patients with atherosclerosis.

But maybe not all is due to arteriosclerosis. “Inflammaging”, a common denominator of age-associated frailty, may also contribute to the severe COVID-19 course in older people. One hypothesis is that pre-existing inflammatory cells, including senescent populations and adipocytes, create the inflammaging phenotype that amplifies subsequent inflammatory events. Nevertheless, high amounts of inflammation alone do not explain the devastating tissue destruction and it may be that age-associated changes in T cells have a role in the immunopathology (review: Akbar 2020). There is growing evidence that coordination of SARS-CoV-2 antigen-specific responses is disrupted in older individuals. Scarcity of naive T cells was also associated with ageing and poor disease outcomes (Rydyznski 2020).

Sex and ethnicity

A striking finding is the lower mortality in female patients, evident through almost all available data. In Italy, for example, male gender was an independent risk factor associated with mortality at ICU with a hazard ratio of 1.57 (Grasselli 2020). Using a health analytics platform covering 40% of all patients in England, COVID-19-related death was associated with being male, with a hazard ratio of 1.59 (95% CI 1.53–1.65) (Williamson 2020). The hitherto largest registry study with detailed data on demographics and other clinical factors is shown in Table 3. There is some evidence that there are sex-specific differences in clinical characteristics and prognosis and that the presence of comorbidities is of less impact in females (Meng 2020). It has been speculated that the higher vulnerability in men is due to the presence of subclinical systemic inflammation, blunted immune system, down-regulation of ACE2 and accelerated biological aging (Bonafè 2020).


Table 3. Age and co-morbidities in a large registry study (Docherty 2020), providing multivariate analyses and hazard ratios.
  UK, n = 15,194
Hazard Ratio (95% CI) Death
Age 50-59 vs < 50 2.63 (2.06-3.35)
Age 60-69 vs < 50 4.99 (3.99-6.25)
Age 70-79 vs < 50 8.51 (6.85-10.57)
Age > 80 vs < 50 11.09 (8.93-13.77)
Female 0.81 (0.75-0.86)
Chronic cardiac disease 1.16 (1.08-1.24)
Chronic pulmonary disease 1.17 (1.09-1.27)
Chronic kidney disease 1.28 (1.18-1.39)
Diabetes 1.06 (0.99-1.14)
Obesity 1.33 (1.19-1.49)
Chronic neurological disorder 1.18 (1.06-1.29)
Dementia 1.40 (1.28-1.52)
Malignancy 1.13 (1.02-1.24)
Moderate/severe liver disease 1.51 (1.21-1.88)


An in-depth analysis performed on 137 COVID-19 patients found that male patients had higher plasma levels of innate immune cytokines such as IL-8 and IL-18 along with more robust induction of non-classical monocytes. A poor T cell response negatively correlated with patients’ age and was associated with worse disease outcome in male patients, but not in female patients. Conversely, higher innate immune cytokines were associated with worse disease progression in female patients, but not in male patients (Takahashi 2020). Emerging knowledge on the basic biological pathways that underlie gender differences in immune responses needs to be incorporated into research efforts on SARS-CoV-2 pathogenesis and pathology to identify targets for therapeutic interventions aimed at enhancing antiviral immune function and lung airway resilience while reducing pathogenic inflammation in COVID-19 (review: Bunders 2020).

Ethnic minorities may be disproportionately affected by the COVID-19 pandemic. Among the first 1.3 million lab-confirmed COVID-19 cases reported to CDC until May 30, 2020, 33% of persons were Hispanic (accounting for 18% of the US population), 22% (13%) were black, and 1.3% (0.7%) were non-Hispanic American Indian or Alaska Native (Stoke 2020). However, in a large cohort study on 5902 COVID-19 patients treated at a single academic medical center in New York, survival outcomes of non-Hispanic Black and Hispanic patients were at least as good as those of their non-Hispanic White counterparts when controlling for age, sex, and comorbidities (Kabarriti 2020). Several other US studies have also found no differences, after controlling for confounders such as age, gender, obesity, cardiopulmonary comorbidities, hypertension, and diabetes (McCarty 2020, Muñoz-Price 2020, Yehia 2020). There is some evidence indicating a longer wait to access care among black patients in the US, resulting in more severe illness on presentation to health care facilities (Price-Haywood 2020).


Several studies have found obesity to be an important risk factor (Goyal 2020, Petrilli 2019). Among the first 393 consecutive patients who were admitted to two hospitals in New York City, obese patients were more likely to require mechanical ventilation. Obesity was also an important risk factor in France (Caussy 2020), Singapore and the US, especially in younger patients (Ong 2020, Anderson 2020). Of 3222 young adults (age 18 to 34 years) hospitalized for COVID-19 in the US, 684 (21%) required intensive care and 88 patients (2.7%) died. Morbid obesity and hypertension were associated with a greater risk of death or mechanical ventilation. Importantly, young adults aged 18 to 34 years with multiple risk factors (morbid obesity, hypertension, and diabetes) faced risks similar to 8862 middle-aged (age 35-64 years) adults without these conditions (Cunningham 2020). A recent review has described some hypotheses regarding the deleterious impact of obesity on the course of COVID-19 (Lockhart 2020), summarizing current knowledge on the underlying mechanisms. These are:

  1. Increased inflammatory cytokines (potentiate the inflammatory response)
  2. Reduction in adiponectin secretion (abundant in the pulmonary endothelium)
  3. Increases in circulating complement components
  4. Systemic insulin resistance (associated with endothelial dysfunction and with increased plasminogen activator inhibitor-1)
  5. Ectopic lipid deposited in type 2 pneumocytes (predisposing to lung injury).


Besides older age and obesity, many risk factors for severe disease and mortality have been evaluated in the current pandemic.

Early studies from China found comorbidities such as hypertension, cardiovascular disease and diabetes to be associated with severe disease and death (Guan 2020). Among 1,590 hospitalised patients from mainland China, after adjusting for age and smoking status, COPD (hazard ratio, 2.7), diabetes (1.6), hypertension (1.6) and malignancy (3.5) were risk factors for reaching clinical endpoints (Guan 2020). Dozens of further studies have also addressed risk factors (Shi 2020, Zhou 2020). The risk scores that have been mainly proposed by Chinese researchers are so numerous that they cannot be discussed here. They were mainly derived from uncontrolled data and their clinical relevance remains limited. An interactive version of a relatively simple, so called “COVID-19 Inpatient Risk Calculator” (CIRC) evaluated in 787 patients admitted with mild-to-moderate disease between March 4 and April 24 in five US hospitals in Maryland and Washington (Garibaldi 2020), is available at

Smoking as a risk factor is under discussion, as well as COPD, kidney diseases and many others (see chapter Comorbidities, page 379). Among 1150 adults admitted to two NYC hospitals with COVID-19 in March, older age, chronic cardiac disease (adjusted HR 1.76) and chronic pulmonary disease (2.94) were independently associated with in-hospital mortality (Cummings 2020).

The main problem of all studies published to date is that their uncontrolled data is subject to confounding and they do not prove causality. Even more importantly, the larger the numbers, the more imprecise the definition of a given comorbidity. What is a “chronic cardiac disease”, a mild and well-controlled hypertension or a severe cardiomyopathy? The clinical manifestations and the relevance of a certain comorbidity may be very heterogeneous (see chapter Comorbidities, page 379).

There is growing evidence that sociodemographic factors play a role. Many studies did not adjust for these factors. For example, in a large cohort of 3481 patients in Louisiana, US, public insurance (Medicare or Medicaid), residence in a low-income area, and obesity were associated with increased odds of hospital admission (Price-Haywood 2020). A careful investigation of the NYC epidemic revealed that the Bronx, which has the highest proportion of racial and ethnic minorities, the most persons living in poverty, and the lowest levels of educational attainment, had higher rates (almost two-fold) of hospitalization and death related to COVID-19 than the other 4 NYC boroughs Brooklyn, Manhattan, Queens and Staten Island (Wadhera 2020).

Taken together, large registry studies have found slightly elevated hazard ratios of mortality for multiple comorbidities (Table 3). It seems, however, that most patients with preexisting conditions are able to control and eradicate the virus. Co-morbidities play a major role in those who do not resolve and who fail to limit the disease to an upper respiratory tract infection and who develop pneumonia. Facing the devastation that COVID-19 can inflict not only on the lungs but on many organs, including blood vessels, the heart and kidneys (nice review: Wadman 2020), it seems plausible that a decreased cardiovascular and pulmonary capacity ameliorate clinical outcome in these patients.

However, at this time, we can only speculate about the precise role of co-morbidities and their mechanisms to contribute to disease severity.

Is there a higher susceptibility? In a large, population-based study from Italy, patients with COVID-19 had a higher baseline prevalence of cardiovascular conditions and diseases (hypertension, coronary heart disease, heart failure, and chronic kidney disease). The incidence was also increased in patients with previous hospitalizations for cardiovascular or non-cardiovascular diseases (Mancia 2020). A large UK study found some evidence of potential socio-demographic factors associated with a positive test, including deprivation, population density, ethnicity, and chronic kidney disease (Lusignan 2020). However, even these well perfomed studies cannot completely rule out the (probably strong) diagnostic suspicion bias. Patients with co-morbidities could be more likely to present for assessment and be selected for SARS-CoV-2 testing in accordance with guidelines. Given the high number of nosocomial outbreaks, they may also at higher risk for infection, just due to higher hospitalization rates.


COVID-19 shows an extremely variable course, from completely asymptomatic to fulminantly fatal. In some cases it affects young and apparently healthy people, for whom the severity of the disease is neither caused by age nor by any comorbidities – just think of the Chinese doctor Li Wenliang, who died at the age of 34 from COVID-19 (see chapter The First 8 Months, page 429). So far, only assumptions can be made. The remarkable heterogeneity of disease patterns from a clinical, radiological, and histopathological point of view has led to the speculation that the idiosyncratic responses of individual patients may be in part related to underlying genetic variations. Many single nucleotide polymorphisms (SNPs) across a variety of genes (eg, ACE2, TMPRSS2, HLA, CD147, MIF, IFNG, IL6) have been implicated in the pathology and immunology of SARS-CoV-2 and other pathogenic coronaviruses (Ovsyannikova 2020). The ‘COVID-19 Host Genetics Initiative’ brings together the human genetics community to generate, share, and analyze data to learn the genetic determinants of COVID-19 susceptibility, severity, and outcomes (CHGI 2020). It seems that regions on chromosome 3 are significantly associated with severe COVID-19 at the genome-wide level. The risk variant in this region confers an odds ratio for requiring hospitalization of 1.6 (95% confidence interval: 1.42-1.79).

Some further key studies are listed here:

  • A large study identified a 3p21.31 gene cluster as a genetic susceptibility locus in patients with COVID-19 with respiratory failure and confirmed a potential involvement of the ABO blood-group system (Elinghaus 2020). A higher risk in blood group A was found compared to other blood groups (odds ratio, 1.45; 95% CI, 1.20 to 1.75) and a protective effect in blood group O as compared with other blood groups (odds ratio, 0.65; 95% CI, 0.53 to 0.79)
  • In a meta-analysis of 7 studies, comparing 7503 SARS-CoV-2 positive patients with 2,962,160 controls, SARS-CoV-2 positive individuals were more likely to have blood group A (pooled OR 1.23, 95% CI: 1.09–1.40) and less likely to have blood group O (pooled OR 0.77, 95% CI: 0.67–0.88) (Golinelli 2020).
  • Associations between ApoEe4 alleles and COVID-19 severity, using the UK Biobank data (Kuo 2020). ApoEe4e4 homozygotes were more likely to be COVID-19 test positives (odds ratio 2.31, 95% CI: 1.65-3.24) compared to e3e3 homozygotes. The ApoEe4e4 allele increased risks of severe COVID-19 infection, independent of pre-existing dementia, cardiovascular disease, and type 2 diabetes.
  • A report from Iran describes three brothers aged 54 to 66 who all died of COVID-19 after less than two weeks of fulminating progress. All three had previously been healthy, without underlying illnesses (Yousefzadegan 2020).
  • Two families with rare germline variants in an innate immune-sensing gene, toll-like receptor 7 (TLR7), that leads to severe disease even in young males who inherit the mutated gene on a single copy of their X chromosome (van der Made 2020).

In addition to the genetic predisposition, other potential reasons for a severe course need to be considered: the amount of viral exposure (probably high for Li Wenliang?), the route by which the virus enters the body, ultimately also the virulence of the pathogen and a possible (partial) immunity from previous viral diseases. If you inhale large numbers of virus deeply, leading rapidly to a high amount of virus in the pulmonary system, this may be much worse than smearing a small amount of virus on your hand and, later, to your nose. In this latter case, the immune system in the upper respiratory tract may have much more time to limit further spread into the lungs and other organs. After an outbreak at a Swiss Army base, soldiers had to keep a distance of at least 2 m from each other at all times, and in situations where this could not be avoided (e.g., military training), they had to wear a surgical face mask. Of the 354 soldiers infected prior to the implementation of social distancing, 30% fell ill from COVID-19. While no soldier in a group of 154 in which infections appeared after implementation of social distancing developed COVID-19 (Bielecki 2020).

Pre-existing SARS-CoV-2 S-reactive T cells may also play a role, contributing to the divergent manifestations of COVID-19. These cells represent cross-reactive clones, probably acquired during previous infections with endemic human coronaviruses (HCoVs). In healthy SARS-CoV-2-unexposed donors, they were found in 35% (Braun 2020). However, the clinical effect of these T cells and other immunological factors on clinical outcomes remains to be determined. There are hundreds of immunological papers focusing on the unresolved question why some patients develop severe disease, while others do not (review: Gutierrez 2020). It remains also to be seen whether T cells provide long-term protection from reinfection with SARS-CoV-2 and if there is a natural immunity, induced by cross-reactive T cells (Le Bert 2020, Mateus 2020).

Over the coming months, we will get a clearer view of 1) correlates of immunoprotection, such as virus-specific antibodies that limit disease and 2) correlates of immune dysregulation, such as cytokine over-production that may promote disease.

Overburdened health care systems

Mortality may be also higher in situations where hospitals are unable to provide intensive care to all the patients who need it, in particular ventilator support. Mortality would thus also be correlated with health-care burden. Preliminary data show clear disparities in mortality rates between Wuhan (> 3%), different regions of Hubei (about 2.9% on average), and across the other provinces of China (about 0.7% on average). The authors have postulated that this is likely to be related to the rapid escalation in the number of infections around the epicenter of the outbreak, which resulted in an insufficiency of health-care resources, thereby negatively affecting patient outcomes in Hubei, while this was not the case in other parts of China (Ji 2020). Another study estimated the risk of death in Wuhan as high as 12% in the epicentre and around 1% in other more mildly affected areas (Mizumoto 2020)­.

Finally, there may be differences between hospitals. In a US cohort of 2215 adults with COVID-19 who were admitted to ICUs at 65 sites, 784 (35.4%) died within 28 days. However, mortality showed a wide variation between hospitals (range, 6.6%-80.8%). One of the well known factors associated with death was a hospital with fewer intensive care unit beds (Gupta 2020)! 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).

Reactivations, reinfections

Seasonal coronavirus protective immunity is not long-lasting (Edridge 2020). There are several reports of patients infected with SARS-CoV-2 who became positive again after negative PCR tests (Lan 2020, Xiao 2020, Yuan 2020). These reports have gained much attention, because this could indicate reactivations as well as reinfections. After closer inspection of these reports, however, there is no good evidence for reactivations or reinfections, and other reasons are much more likely. Methodological problems of PCR always have to be considered; the results can considerably fluctuate (Li 2020). Insufficient material collection or storage are just two examples of many problems with PCR. Even if everything is done correctly, it can be expected that a PCR could fluctuate between positive and negative at times when the values ​​are low and the viral load drops at the end of an infection (Wölfel 2020). The largest study to date found a total of 25 (14.5%) of 172 discharged COVID-19 patients who had a positive test at home after two negative PCR results at hospital (Yuan 2020). On average, the time between the last negative and the first positive test was 7.3 (standard deviation 3.9) days. There were no differences to patients who remained negative. This and the short period of time suggest that in these patients, no reactivations are to be expected.

However, in recent months several case reports of true (virologically proven: phylogenetically distinct strains) re-infections have been reported (To 2020, Gupta 2020, Van Elslande 2020). In most cases, the second episode was milder than the first. However, there is at least one case where the second infection was more severe, potentially due to immune enhancement, acquisition of a more pathogenic strain, or perhaps a greater in-oculum of infection as the second exposure was from within household contacts (Larson 2020). Up to now, however, these are anecdotal case reports.

Animal studies suggest that re-infection is unlikely (Chandrashekar 2020). Following initial viral clearance and on day 35 following initial viral infection, 9 rhesus macaques were re-challenged with the same doses of virus that were utilized for the primary infection. Very limited viral RNA was observed in BAL on day 1, with no viral RNA detected at subsequent timepoints. These data show that SARS-CoV-2 infection induced protective immunity against re-exposure in nonhuman primates. There is growing evidence for a long-lived and robust T cell immunity that is generated following natural SARS-CoV-2 infection (Neidleman 2020).

Reactivations as well as rapid new infections would be very unusual, especially for coronaviruses. If a lot of testing is done, you will find a number of such patients who become positive again after repeated negative PCR and clinical convalescence. The phenomenon is likely to be overrated. Most patients get well anyway; moreover, it is unclear whether renewed positivity in PCR is synonymous with infectiousness.

Long-term sequelae

The profound physical impairments associated with critical COVID-19 illness are well known. Many patients with severe COVID-19, especially older patients and those with ARDS, will suffer long-term complications from an intensive care unit stay and from the effects of the virus on multiple body systems such as the lung, heart, blood vessels and the CNS. However, there is growing evidence that even in some younger people with non-severe COVID-19 the illness may continue for weeks, even months. The persistent symptoms in these so-called “long haulers” fluctuate and range from severe fatigue, breathlessness, fast heart rate with minimal exertion, chest pain, pericarditis/myocarditis, hoarseness, skin manifestations and hair loss, acquired dyslexia, headaches, memory loss, relapsing fevers, joint pains, and diarrhea. Symptoms may arise through several mechanisms including direct organ damage and involvement of immune function and the autonomic nervous system. The following key papers address post-acute findings in patients with mild-to-moderate COVID-19.

  • In Rome, 143 patients discharged from hospital were assessed after a mean of 60 days after onset of the first COVID-19 symptom. During hospitalization, 73% had evidence of pneumonia but only 15% and 5% received non-invasive or invasive ventilation, respectively. Only 13% were completely free of any COVID-19-related symptom, while 32% had 1-2 symptoms and 55% had 3 or more. Many patients reported fatigue (53%), dyspnea (43%), joint pain (27%) and chest pain (28%). A worsened quality of life (QoL) was observed in 44% of patient (Carfì 2020).
  • In Paris, persistent symptoms and QoL were assessed in 120/222 patients discharged from a COVID-19 ward unit, at a mean of 111 days after their admission. The most common persistent symptoms were fatigue (55%), dyspnea (42%), loss of memory (34%), concentration and sleep disorders (28% and 31%, respectively) and hair loss (20%). Of note, ward and ICU patients showed no differences with regard to these symptoms. In both groups, EQ-5D (mobility, self-care, pain, anxiety or depression, usual activity) showed a slight difference in pain in the ICU group (Garrigues 2020).
  • The only US data to date, including a random sample of adults testing positive at an outpatient visit (Tenforde 2020). Telephone interviews were conducted at a median of 16 (14–21) days after the test date. Among 292 respondents, 94% reported experiencing one or more symptoms at the time of testing; 35% of these reported not having returned to their usual state of health by the date of the interview, increasing from 26% (those aged 18–34 years) to 47% (≥ 50 years).
  • Physical fitness before and after infection in 199 young, predominantly male military recruits (Crameri 2020) from Switzerland. Recruits had had a “baseline” fitness test, performed 3 months prior to a large COVID-19 outbreak in the company, including a progressive endurance run. Baseline fitness values were compared with a fitness test at a median of 45 days after SARS-CoV-2 diagnosis. Participants were grouped into convalescent recruits with symptomatic COVID-19 (n=68), asymptomatic cases (n=77) and a naive group without symptoms or laboratory evidence of SARS-CoV-2 infection (n=54). Results: neither of the strength tests differed significantly between the groups. However, there was a significant decrease in VO2 max among convalescents compared with naive and asymptomatically infected recruits. Around 19% of the COVID-19 convalescents had a decrease of more than 10% in VO2 max, while none of the naive recruits showed such a decrease.
  • The best study to date on cardiac issues, including 100 COVID-19 patients at a mean age of 49 years (Puntmann 2020). The median time between diagnosis and cardiac MRI (CMR) was 71 (64-92) days. Most patients recovered at home (n=67), with only minor or moderate (n=49) or without any symptoms (n=18). Compared with pre-COVID-19 status, 36% reported ongoing shortness of breath and general exhaustion, of whom 25 noted symptoms during less-than-ordinary daily activities, such as household chores. CMR revealed cardiac involvement in 78% and ongoing myocardial inflammation in 60%, independent of pre-existing conditions, severity of COVID-19 or from the time of diagnosis. The authors concluded that “participants with a relative paucity of pre-existing cardiovascular condition and with mostly home-based recovery had frequent cardiac inflammatory involvement, which was similar to the hospitalized subgroup”.
  • A comprehensive CMR examination in 26 competitive athletes, among them 14 asymptomatic and 12 with only mild symptoms. CMR was performed 11-53 days after recommended quarantine (Rajpal 2020). In total 4/26 (15%) had CMR findings suggestive of myocarditis and 8/26 (31%) exhibited changes suggestive of prior myocardial injury. In 7/12 of patients with pathological findings, CMR had been performed at least three weeks after the positive SARS-CoV-2 test result.
  • MRI in 60 COVID-19 patients (47 classified as mild), performed at a mean of 97 days from symptom onset. Compared with 39 age- and sex-matched non-COVID-19 volunteers, recovered COVID-19 patients showed volumetric and micro-structural abnormalities that were detected mainly in the central olfactory cortices and partially in the white matter in the right hemisphere. According to the authors, these abnormalities might cause long-term burden to COVID-19 patients after recovery (Lu 2020).

Taken together, clinical data is still scarce. However, it is dismissive to solely attribute persisting symptoms after mild or moderate COVID-19 to anxiety or to depression or to label them as anecdotal. “COVID-19 long haulers” are not hypochondriacs. There is an urgent need to quantify long-term complications properly and accurately, including non-hospitalized patients with mild disease, and several prospective studies are underway (Reviews: Alwan 2020, Greenhalgh 2020, Marshall 2020, Yelin 2020).


Over the coming months, serological studies will give a clearer picture of the true number of asymptomatic patients and those with unusual symptoms. More importantly, we have to learn more about risk factors for severe disease, in order to adapt prevention strategies. Older age is the main but not the only risk factor. Recently, a 106-year-old COVID-19 patient recently recovered in the UK. The precise mechanisms of how co-morbidities (and co-medications) may contribute to an increased risk for a severe disease course have to be elucidated. Genetic and immunological studies need to reveal susceptibility and predisposition for both severe and mild courses. Who is really at risk, who is not? Quarantining only the old is too easy.

By Christian Hoffmann  &
  Bernd Sebastian Kamps