22nd May - the COVID-19 coronavirus compendium

Vaccine trials, global carbon emissions, and ethnicity as a risk factor for severe disease

Like Comment
Read the paper

It was a pleasure this week to read papers from both of my PhD supervisors: Wendy Barclay at Imperial College London (formerly at the University of Reading), who wrote an excellent review of coronavirus immune responses, and Maria Zambon at Public Health England, who contributed to a study on risk groups in the UK. Both are primarily influenza researchers who have been drawn into researching and commenting on this outbreak. Maria in particular has an influential role as a member of the all-important SAGE, which advises the UK government on the outbreak. I wish them both well!

In less happy news, my internet failed midway through writing this, so many of the links will take you to the paper at PubMed Central, from where you can access the full article on the publisher website. Sorry about that!

This week we learnt that two new DNA vaccines successfully induced an immune response in animal models, that global carbon emissions have fallen due to reduced surface transport, and read yet further evidence from the UK and the USA that some racial and ethnic groups are much more likely to have severe COVID-19, most likely due to existing health inequalities.

Model systems

A human cell culture model for growing SARS-CoV-2 was established and used to track the virus life cycle through proteomics. Several potential drug targets were identified and the cell culture model will be useful to any researcher who wishes to grow the virus.

An intestinal organoid model system derived from horseshoe bats was developed and used to grow SARS-CoV-2. The virus also grew in human intestinal organoids, adding to the evidence that the virus can grow in the gut.

Another animal model has been developed, this time in golden Syrian hamsters. The hamsters infected with SARS-CoV-2 had similar symptoms to humans with mild disease, and spread the virus to other animals via direct contact and aerosols.


The structure of the coronavirus polymerase was solved using cryo electron microscopy, which can be used to model drug interactions, including of remdesivir. The ubiquitination and ISGylation of SARS-CoV-2 PLpro was evaluated, and a potential inhibitor identified.

Fatality rate

There were more than 24,000 excess deaths in New York City in March and April, 5000 of which had not been deemed confirmed or probable COVID-19. A similar study in Italy found an excess of 45,033 deaths between 1st March and 15th April, while the number of official COVID-19 deaths was 21,046. This suggests that using confirmed or probable cases likely underestimates the true number of deaths.

A retrospective analysis of more than 80,000 cases in China estimated a case fatality rate of 8% in Hubei province and 0.9% in the rest of China.


A sequence analysis of the SARS-CoV-2 genome from four cases in Italy was described. Some HLA alleles may be associated with an increased incidence of SARS-CoV-2, according to a study of 82 Chinese individuals.


It's now clear - SARS-CoV-2 can infect dogs. In an excellent piece of technical work, Malik Peiris' lab at the University of Hong Kong definitively showed that 2 out of 15 families that were positive for the virus had also infected their dogs, a Pomeranian and a German Shepherd. None of the dogs were ill, but they shed live virus, and developed antibodies, and so could potentially be infectious to humans, although this has not yet been proven.


In a paper from last week that I missed, an estimated 4% of people in France have been infected with SARS-CoV-2. Social distancing reduced the reproductive number from 2.9 to 0.67 and the case fatality rate ranged from 10% in over 80s, to 0.001% in those under 20 years. As herd immunity has not been achieved, partial lockdown must remain.

The index case of an outbreak in Bavaria was a Chinese visitor, and the virus then spread mainly by asymptomatic or presymptomatic carriers. The outbreak in Italy was more severe in those regions with higher family fragmentation and availability of residential health facilities, contrary to prevailing theories that more intergenerational contact may have helped spread the infection in Italy.

A traveller screening system did not protect California from importing cases of COVID-19, due to incomplete information, and the role of asymptomatic and presymptomatic spread. Most cases in San Francisco came from New York or Europe. Most cases of COVID-19 in Canada came from the US, Europe, and the Middle East, not China. 3-7% of airline passengers travelling to Greece from the UK, Spain and Turkey were infected, despite generally being free from symptoms. The most important routes for importation of COVID-19 into Brazil were predicted.

Prisoners and staff are becoming infected in US correctional facilities, according to data from two studies, allowing spread back to the community. 76 out of 350 guests at a wedding in Jordon became infected with SARS-CoV-2, showing the important role of mass gatherings in transmission. 53 out of 61 members of a choir tested positive for COVID-19, and aerosol transmission was suspected, perhaps due to the effects of singing. SARS-CoV-2 RNA could be detected on surfaces in a quarantine facility for COVID-19 patients.

Social distancing

COVID-19 prevention strategies in Tsimane forager-horticulturalists in the Bolivian Amazon requires collaboration between different groups of specialists, with implications for the health of other indigenous people.

The outbreak on the Diamond Princess was estimated to have a reproductive number of almost 15, far higher than seen elsewhere. Isolation and quarantine were therefore highly effective at reducing the potential impact, despite 17% of the crew and passengers testing positive.

China's response to the outbreak resulted in sub-exponential growth, with less cases than would have been expected early on, showing that effective containment works. Isolation of those with disease in institutions, rather than at home, would be more effective at controlling the outbreak, according to a model, although this policy may be deemed unacceptable in many countries.

Restrictive measures should remain in Lombardy until July, to prevent an increase in hospitalisations, whereas in Emilia-Romagna a higher contact rate can be permitted, according to a model of the disease. The reproductive number in Iran has declined to less than 1, after 2 months of interventions. The outbreak in Charlotte, North Carolina, was modelled.

10 options for social distancing in the UK were assessed and presented to government rapidly, as described in this paper. Dynamic cycles of 50-day suppression followed by a 30-day relaxation will effectively reduce spread and allow hopsitals to cope with increased demand, according to a model of the outbreak in 16 countries. Another model suggests that stepping down social isolation measures every 80 days will prevent a second peak, compared to a sudden removal of all non-pharmaceutical interventions.

The effects of wind on viral transmission were examined in a fluid particle dynamics model. Distancing should be increased, the authors argue.


Global CO2 emissions decreased by 17% in April, mainly due to reductions in surface transport, and comparable to the levels in 2006. Individual countries saw their emissions reduce by as much as 26%, but not all at the same time. Annual emissions could be reduced by between 3 and 13% for 2020, if some restrictions remain worldwide until the end of the year.

Low solar irradiance and high population density were independently associated with larger COVID-19 outbreaks. Another study found that low 9am humidity was associated with increased incidence of COVID-19 in New South Wales, Australia. Rainfall and temperature did not correlate with incidence, and both studies are correlations that require further study.


Neutralising antibodies against SARS-CoV-2 were generated, with one showing therapeutic effect in a mouse model of infection. A structure of this antibody bound to the spike protein allowed further antibodies to be rationally designed.

Patients who started Lopinavir/Ritonavir treatment within 10 days showed shorter viral shedding. There was no effect if treatment was started from 11 days.

Sunlight inactivates SARS-CoV-2, with 90% inactivated in 7 minutes in simulated saliva. Early use of methylprednisolone, a corticosteroid, improved clinical outcome in COVID-19 patients in a small study in Michigan. Narcissoside can inhibit SARS-CoV-2 protein 6W63, according to molecular docking. A computational study identified a ligand-binding domain in the viral protease.


A synthetic DNA-based vaccine candidate targeting SARS-CoV-2 spike protein generated immune responses in mice and guinea pigs. A different DNA vaccine, again using the spike protein, or fragments of it, was tested in rhesus macaques and shown to reduce titres of virus when the animals were challenged. Over 90 vaccines have been developed and are being tested in animals and humans.


Knowledge about COVID-19 was generally good in a surveyed population in Jimma, Ethiopia, and also in Malaysia. Whatsapp is an effective tool for sharing health information about COVID-19 in challenging conditions in North West Syria.

Psychological factors that influence hygiene practices were examined in a survey of more than 2000 people in the UK, and found that reflective motivation was the greatest predictor of good hygiene. Perception of risk varied by age in Israel, in a survey.

Recommendations about whether to wear masks varied greatly between countries, with some recommending and others not.


Google searches related to buying firearms has surged to unprecedented levels – higher than seen after mass shootings in the US, with 2.1 million searches over 34 days. Previous spikes in gun purchasing during national emergencies led to a spike in firearms injuries and deaths.


The costs for social protection measures for those who depend on informal work in Mexico were estimated, to allow them to stay at home and not work during COVID-19.


Half of the infected health care workers in China were nurses and 72% were women. The infection rate of healthcare workers was higher than in the general population, but the fatality rate was significantly lower.

A portfolio of nine research studies in the UK were immediately activated as a result of the COVID-19 outbreak. Originally intended for an influenza pandemic, the studies will allow researchers to quickly monitor the outbreak, looking at at risk groups, modelling, and other questions.

Stroke patients took an hour longer than usual to reach hospital in Hong Kong during COVID-19 than before, and less patients presented at hospital, despite no decrease in incidence. 1% of COVID-19 patients had a stroke, possibly related to hypercoagulability. Admissions to Italian hopsitals for acute myocardial infarction were also reduced during the COVID-19 outbreak, causing an increase in the fatality rate.

Many paediatric cancer units in Latin American hospitals are understaffed because of COVID-19, with implications for cancer treatment in this region.

A reusable N95 mask with pores of just 5 nm was developed. An A4 over head projector sheet was converted into a face shield, for healthcare use. Hydrogel patches can be used to prevent skin damage from face mask wearing. Cloth masks were as effective as medical masks in preventing particle spread (live virus was not used), in an innovative model using a car with air conditioning.

An increase in scalds from steam inhalation was seen in the UK, mainly in those of Indian, Pakistani, or Bangladeshi ethnicity, most likely due to the belief that steam has health benefits during colds.


Rhesus macaques previously infected with SARS-CoV-2 were protected from a subsequent infection. Viral loads were 100,000-fold lower during re-infection, and the monkeys mounted an immune response. If true in humans, this suggests that a past infection provides immunity. 

Antibodies isolated from a patient infected with SARS-CoV-1 during the SARS outbreak bind to and neutralise the SARS-CoV-2 virus, and so could potentially be used to treat COVID-19. Cryo-electron microscopy was used to further characterise the binding of one of the monoclonal antibodies to the spike protein. Another study found that although cross-reactivity was seen between SARS and COVID-19, cross-neutralisation was rare.

Single domain antibodies from llamas inoculated with the coronavirus spike protein were isolated and crystallised with the receptor. Llama antibodies are of interest as they contain just a heavy chain, and so are more stable than conventional antibodies. They can also be mass produced in CHO cells, and nebulised as a potential treatment, which will no doubt now be explored for COVID-19.

The varied serological responses in individuals exposed during an outbreak in a German hospital were described, as was the immune response to infection. A commercial antibody test was assessed.

At risk groups

The proportion of African Americans living in each of the 2886 US counties was associated with more confirmed COVID-19 cases, more confirmed deaths and a higher case mortality, most likely due to existing health disparities. Parts of Chicago with high numbers of African Americans were also more likely to show social vulnerability and risk factors for COVID-19, as well as an increased death rate.

Risk factors for a positive SARS-CoV-2 test were analysed from primary care data in the UK, with deprivation, population density, ethnicity, and chronic kidney disease all being risk factors. Black people were almost five times more likely to test positive as white people, whereas smokers were less likely to be infected.

22% of the population of Sweden have a risk factor for severe COVID-19, according to an analysis of healthcare records and population registers from more than 9 million people, although the prevalance varied by region. 25% of severe COVID-19 patients in a France hospital were obese, compared to 15% of the general population, again highlighting this as a risk factor.

Critically ill patients in New York City were more likely to be male, and have a pre-existing health condition such as hypertension, diabetes, or obesity. Mortality was associated with several factors and biomarkers, including older age, chronic cardiac disease, chronic pulmonary disease, high levels of IL-6, and high levels of D-dimer. IL-6 and D-dimer are markers of immune dysregulation and thrombosis, respectively. In a study from the Bronx, New York City, severe obesity, increasing age, and male sex were associated with an increased risk of death. More evidence that men were more likely to have severe disease, as were older people, was published in a study of more than 5000 in Israel.

Diabetes is a risk factor for severe COVID-19 disease, and glycosylated hemoglobin (HbA1c) was associated with inflammation and hypercoagulability. In addition, patients with COVID-19 who are older, male, or diabetic are at higher risk of requiring intubation in hospital. Dementia is also a risk factor for severe disease, according to an Italian study, as is metabolic associated fatty liver disease.


The levels of ACE-2 and TMPRSS2 in various cells of the mouse lung and in the human airway were measured. ACE-2 expression was elevated in COPD patients who are overweight, and its expression is modulated by oestrogen in airway epithelial cells.

Binding sites for antibodies, and ACE-2, were identified, and another study looked at the affinity of binding to ACE-2.

Markers of severe disease

A machine learning model was developed to predict risk of death from COVID-19, using a dataset of blood biomarkers from almost 500 patients in Wuhan. High lactic dehydrogenase (LDH), low lymphocytes, and high levels of high-sensitivity C-reactive protein (hs-CRP), can together be used to predict death from COVID-19 at 90% accuracy. The tests can identify high risk patients more than 10 days in advance of death, allowing these patients to be prioritised. Another study also found that plasma CRP was associated with disease severity.

Data from 326 COVID-19 patients in Shanghai found that age, lymphocyte count, co-morbidities, and gender were the strongest predictors of disease severity, from asymptomatic, to mild and severe. Levels of inflammatory cytokine, IL-6 and IL-8, were also higher in those with more severe illness. Genetic variation in the virus was observed, but did not affect virulence. 

A number of other studies also looked at markers of disease severity. IL-6 and IL-8 elevation, plus lymphopenia was associated with severe disease. Another study from Wuhan found that co-infection, lymphocyte count and levels of D-dimer were associated with severity of COVID-19. Levels of IL-6 and C-reactive protein were predictive of the need for mechanical ventilation, and IL-6 was a marker of severity in another study. Some symptoms were also associated with severe disease.

Mental health

SARS-CoV-2 might cause delirium in a significant minority of cases, according to a systematic review. Sleep patterns were disrupted during the outbreak in Italy. Patients with existing severe mental health conditions experienced more pandemic-related stress than the general population in Italy during the lockdown. 

Young adults are worried about the impact that COVID-19 is having on their older relatives, rather than on themselves, according to a survey of more than 1000 in Italy. News exposure was associated with depressive symptoms in 500 surveyed adults. Concern over COVID-19 also caused somatic symptoms in Chinese students.

Older men were the least worried about COVID-19, compared to women and younger men, according to a small survey in the US, and were less likely to show behavioural changes, despite them being most at risk of severe disease.

Post-traumatic stress was seen in a small proportion of healthcare workers in China, and anxiety and depression were seen in pregnant women. Exacerbation of obsessive compulsive disorder in a patient, due to media reports of COVID-19, was described.

Social media

123 million tweets about COVID-19 have been written and collected in a database, and 10,000 tweets from medical professionals were analysed for content.


12% of pregnant women in New York tested positive for COVID-19, but none had severe disease.


An artificial intelligence algorithm was able to accurately diagnose patients with COVID-19, based on CT scans, clinical symptoms and exposure history. The AI algorithm performed as well as a senior thoracic radiologist, and gave a positive result in some cases where there was no obvious CT abnormality, in cases that proved to be positive for the virus by RT-PCR.

A colorimetric assay to detect SARS-CoV-2, based on gold nanoparticles, was developed. This allows detection of the virus using the naked eye. An ultra-fast single-tube nucleic acid isothermal amplification detection assay was also developed.

SARS-CoV-2 RNA was detected in waste water in Spain before any official cases had been recorded. Antibody tests against NP were more sensitive than those against the spike protein.

Saliva is an alternative to nasopharangeal swabs for COVID-19 testing, as are oropharyngeal/nares swabs. Mutations in the primer-binding site seen in some cases of COVID-19 in Nepal may affect RT-PCR diagnostics. Rates of false-negative tests in China were presented and discussed.

Clinical findings

Infectious virus was confirmed in the faeces of a COVID-19 patient, and in the eye. 5% of patients at Tongji Hospital has conjunctival congestion. The clinical findings of paediatric COVID-19, with and without gastrointestinal symptoms, was described. Diffuse alveolar damage was seen in the lungs of patients who died of COVID-19.

Patients with severe disease often lose their sense of taste or smell, but do not report this as a symptom, suggesting that a previously seen associated between anosmia and mild disease is not true.

More cases of deep vein thrombosis associated with COVID-19 were recorded in China.  More acral lesions were described.

Patients with rheumatic disease are not at increased risk of COVID-19, according to a survey from Spain. Children with renal disease also have a normal infection. Disease in those with cirrhosis in China was described.

And finally, the Royal Flying Doctor Service in Australia may need to increase its number of aircraft to cope with excess demand from COVID-19. The brave pilots and crew have so far conducted 291 potentially life-saving transfers.

Ben Johnson

Magazine Editor, Nature Medicine, Springer Nature

I trained as a virologist, starting with an undergraduate degree in virology from the University of Warwick, UK. My PhD, in influenza virus genetics and immunoevasion, was from Public Health England and the University of Reading, UK, with Maria Zambon and Wendy Barclay. My research interests then moved to smallpox vaccines, viral ion channels and cell adhesion, while a postdoc at Imperial College London with Geoffrey Smith, FRS. I then joined open-access publisher BioMed Central in 2011 as an editor and then associate publisher and was Head of Communities & Engagement at Springer Nature from 2016, running the Nature Research Communities and other online engagement activities for researchers. I joined Nature Medicine in 2021, with responsibility for news and opinion content, and am based in the London office.