This section covers symptom prevalence, click here for Triage Based on Symptom Questionnaires.
Many patients are asymptomatic. Among patients with symptoms, most present with an influenza-like illness (fevers, myalgias, respiratory symptoms), but many do not present with this classic combination. Some may present with less-usual findings such as perniosis (COVID toes) or anosmia. These ranges are pulled from the following articles, and symptom prevalence varies greatly depending on testing and survey methodology (Arentz et al; Chen et al; Guan et al; Li et al; Wu et al; Zhou et al; WHO-China Joint Mission on COVID-19; Young et al; Yan et al; Jiang et al; Huang et al; Tostmann et al).
- Fever, 44-94%
- We recommend using >= 38°C to define fever, taking into account the patient’s age, immune status, medications (steroids, chemotherapy, etc.), and recent use of fever-reducing medications.
- Children are less likely to have fever or cough (Bialek et al).
- Cough, 68-83%
- Anosmia and/or ageusia (loss of sense of taste and/or smell) ~70%
- Upper respiratory symptoms (sore throat, dripping nose, nasal or sinus congestion), 5-61%
- Shortness of breath, 11- 40%
- Fatigue, 23-38%
- Muscle aches 11-63%
- Headache 8-14%
- Confusion 9%
- Gastrointestinal symptoms (nausea, vomiting, diarrhea), 3-17%
Literature Review: University of Washington Literature Report (Clinical Characteristics)
Updated Date: December 19, 2020
Incubation period is the time from exposure to symptom onset. Latency period is the time from exposure to infectiousness (or viral detection, depending on the definition). COVID-19 has a relatively long incubation period, and typically at least 2 days of infectivity before symptoms develop.
- 97.5% of exposed cases will develop symptoms within 11 days and 99% within 14 days. Over 95% of cases develop symptoms within 13 days of infection (Nie et al).
- Incubation periods of up to 24 days are shown in some reports (Nie et al).
Samples taken before symptom onset have high false negative rates, as modeled by (Kurcirka et al). 68% false negatives one day before symptoms, compared to 38% false negatives on the first day of symptoms, based on serial testing. They estimated the window period between exposure and detectability of SARS-CoV-2 RNA on nasopharyngeal sampling at 3-5 days, with peak sensitivity 8 days after exposure or 3 days after symptom-onset in their model. As with incubation, individual cases may show longer delays. Asymptomatic patients should still be tested in certain circumstances, but a negative result does not rule out infection.
- Fever: 12 days
- Shortness of breath: 13 days
- Cough: 19 days
- Multi-System Inflammatory Syndrome in Children (MIS-C): 6 days (range 4-8 days).
- Sepsis: Median onset 9 days (range 7-13 days)
- Acute Respiratory Distress Syndrome (ARDS): median onset 12 days (range 7-15 days)
- Need for Mechanical Ventilation: Median onset 10 days (range 3-12.5 days)
- Acute Cardiac Injury: Median onset 15 days (range 10-17 days)
- Acute Kidney Injury: Median onset 15 days (range 13-19.5 days)
- Secondary Infection: Median onset 17 days (range 13-19 days)
- Death: Median 18.5 days, interquartile range 15-22 days (Zhou et al)
- Illness severity has been noted to have two peaks at ~14 days and ~22 days (Ruan et al)
The majority of patients have only mild symptoms; however, the percentage of patients who develop severe or critical disease is far greater than for most other respiratory viruses, including influenza. See how mild, moderate, and severe cases are defined. Assessing the percentage of patients who develop differing severities of illness is fundamentally challenging, due to the widely variable case definitions and severity definitions, as well as the lack of population-level surveillance testing to estimate asymptomatic and minimally symptomatic cases. All of these are estimates and do not apply to all populations or epidemiologic circumstances.
- Asymptomatic Infection is present in about 20% of cases (Bi et al; Mizumuto et al; Pollan et al). One metanalysis showed asymptomatic infections account for 17% of all infections (Byambasuren et al) but this is difficult to estimate as screenings of entire populations are unavailable.
- Symptomatic Infection: A Chinese CDC report on approximately 72,000 symptomatic COVID cases (1% of the cases included in the study were asymptomatic), documented the following occurrence rates for mild, severe, and critical symptom presentations (Wu et al):
- Mild Symptoms to Mild Pneumonia: approximately 81%
- Severe Symptoms (blood oxygen saturation less than or equal to 93%, respiratory frequency greater than or equal to 30 breaths per minute, and/or lung infiltrates greater than 50% within 48 hours): approximately 14%
- Critical Symptoms (respiratory failure, shock, multiorgan dysfunction): approximately 5%.
- Among critically-ill patients, many receive mechanical ventilation. Median time on a ventilator ranges from 11-17 days (Chen et al; Ling et al).
- Presentation with shock is rare, but vasopressors are eventually used in 67% of critically-ill patients.
- Cardiomyopathy (Heart Tissue Injury) is noted in 33% of critically-ill patients (Ruan et al).
Updated Date: May, 2020
Multiple factors have been associated with worse prognosis in people infected with SARS COV-2.
- Age: increased age is associated with more severe disease and higher rates of death (Wu et al; Chen et al; Yang et al; Qin et al).
- Children are less likely to have severe disease, but pediatric deaths have been reported (Bialek et al).
- Children appear to be as likely to contract the infection as adults, although symptomatic cases of children are more rare (Bi et al).
- Comorbidities and other health factors: Multiple comorbidities and/or health factors are associated with increased risk of severe COVID-19 illness. Evidence-based knowledge on this topic is continuing to develop; for ongoing updates, see the CDC’s living document. The comorbidities and other health factors associated with the strongest bases of evidence for increased risk are listed below. This list is not inclusive of all conditions which may be associated with increased risk; other common conditions which may be associated with increased risk include hypertension, moderate to severe asthma, liver disease, and others (CDC).
- Chronic kidney disease
- Chronic obstructive pulmonary disease (COPD)
- Type 2 diabetes mellitus
- Sickle cell disease
- Down Syndrome
- Immunocompromised status associated with solid organ transplant
- Obesity (BMI of 30kg/M2 or higher)
- Multiple heart conditions, including heart failure, coronary artery disease, and cardiomyopathies
- Race: Please see Health Equity for a discussion on racial differences in COVID infection and severity.
- Sex: Men appear to be more severely affected by COVID-19 than women. Conclusive evidence related to sex differences is limited by methodology of existing studies (Schiffer et al).
- Smoking: Smoking may offer a small risk reduction for COVID infection, though it is not clear why and this finding may be subject to confounding. It does appear to be associated with worse outcomes. See Smoking for more details.
The most significant laboratory abnormalities associated with severe COVID-19 disease and death include the following:
Normal Ranges (For many US labs, units and values may vary)
White Blood Cell Count (WBC)
> 10 K/uL (K/uL=10^3/uL)
Male and Female- Adults: 3.4-9.6 x10^3/uL
< 1.00 K/uL (K/uL=10^3/uL)
Male and Female- Adults: 0.95-3.07 x10^3/uL
< 150 K/uL (K/uL=10^3/uL)
Male Adults: 135-317 x 10^3/uL
Female Adults: 157-371 x10^3/uL
> 1.5 mg/dL
Male Adults: 0.74-1.35 mg/dL
Female Adults: 0.59-1.04 mg/dL
< 3 g/dL
Alanine transaminase (ALT)
> 40 U/L
Males: 7-55 U/L
Females: 7-45 U/L
Creatinine kinase (CK)
> 185 U/L
Males: 39-308 U/L
Females: 26-192 U/L
Troponin T, high-sensitivity (hs-TnT)
> ~20 ng/L
Male <23 ng/L
Female <15 ng/L
C-reactive protein (CRP)
> 125 mg/L
< or =8.0 mg/L
Lactate dehydrogenase (LDH)
> 245 U/L
Adults: 122-222 U/L
> 300 ug/L (Severe Disease); Ferritin > 1000 ug/L (Death)
Males: 24-336 ug/L
Females: 11-307 ug/L
Interleukin 6 (IL-6)
> 10 pg/mL
< or =1.8 pg/mL
> 1000 ng/mL
< 250 ng/mL
> 0.5 ng/mL
< or =0.15 ng/mL
Determining and reporting the cause of death for patients with COVID-related diseases is complex (as it is with any disease).
- Cause of Death: This is usually the acute medical diagnosis that caused a patient to die, and often relates to a medium-term or long-term diagnosis as well. It will often include other diseases as co-morbid or contributing factors (e.g. pneumonia due to COVID-19 infection or Acute Myocardial Ischemia due to COVID-19 infection and Coronary Artery Disease).
- Mechanism of Death: Defined as the immediate physiologic issue resulting in death (for example, hypoxemia).
A significant number of COVID-related deaths do not have clear delineation of cause of death (CEBM). The majority of people who die from COVID-19 die from respiratory failure. Because definitions of cause of death are reported differently it can be hard to determine exact numbers, but here are estimates (Ruan et al, 68 cases), (Zhang et al, 82 cases):
- Respiratory Failure Alone: 53% - 69%
- Circulatory Failure Alone: 7%-14.6%
- Mixed Respiratory and Circulatory Failure, Sepsis, or Multiorgan Failure: 28-33%
- Hemorrhage: 6.1%
- Renal Failure: 3.1%
- Case Fatality Rate (CFR) is typically the proportion of deaths from a disease relative to the number of people diagnosed with the disease in a specific period of time. Some people define a “case” as showing symptoms.
- Infection Fatality Rate (IFR) is the proportion of deaths from a disease but relative to all infected individuals including asymptomatic people and infections that were missed. It is harder to measure, and thus most places report CFR.
- Case Fatality Rate is variable in different countries. Range around the world seems to be between 0-16%, with most countries in the 1-3% range.
Classification: SARS-CoV-2 is a positive-stranded RNA virus with a nucleocapsid and envelope, belonging to the coronavirus family, of which seven viruses (including the original SARS-CoV in 2003 and MERS in 2013) have crossed from zoonotic origins into humans.
Cell Entry and Replication: For cell entry, the SARS-CoV-2 spike protein binds to the ACE2 receptor, expressed in nasal and bronchial epithelium, pulmonary endothelium, alveolar Type 2 cells, proximal renal tubule cells, cardiac myocytes, gastrointestinal epithelial cells, and others. Cleavage/priming by serine protease TMPRSS2 facilitates SARS-CoV2 cell entry, followed by viral replication using host cell machinery and then exocytosis (Kumar et al).
Cellular Targets and Resulting Lung Injury: The cells that express ACE2 may be the cell populations most injured by infection or targeted by the immune response. Alveolar Type 2 cells secrete surfactant, so injury may result in alveolar collapse at low opening pressures and high PEEP sensitivity, while damage to pulmonary endothelial cells may cause capillary leak and trigger an influx of monocytes and neutrophils, with formation of hyaline membranes. The highly inflamed lung parenchyma can develop microthrombi that help explain some of the thrombotic complications of COVID (Wiersinga et al).
Inflammatory Cascade: Infection with the SARS-Cov-2 virus can cause apoptotic cell death, which triggers an inflammatory cascade of cytokine release, as well as the recruitment of immune cells including macrophages and dendritic cells, and later, antigen-specific T lymphocytes (Bohn et al). If the immune response is not properly checked, a state of hyperinflammation occurs, with the development of Cytokine Storm Syndrome, and sometimes multi-organ failure.
Blood type: There is evidence that A blood type is a risk factor for COVID-19 respiratory failure, and O may be protective. This was based on a genome-wide association study (GWAS) of 835 patients and 1255 control participants from Italy and 775 patients and 950 control participants from Spain. Respiratory failure was defined as a patient requiring supplemental oxygen or mechanical ventilation (Ellinghaus et al).
- Autopsy studies indicate universal damage to pulmonary tissue (Falasca et al; Elsoukkary et al). Pulmonary histology of COVID-19 shows bilateral diffuse alveolar damage, desquamation of pneumocytes, pulmonary edema, hyaline membrane formation, inflammatory cell infiltrates, and multinucleated giant cells, as well as some evidence of direct viral injury (Xu et al; Geng et al). Vascular involvement in the lung is also quite common, with microthrombi, endotheliitis, severe capillaritis, vascular complement deposition, and pulmonary thromboemboli, often in small and mid-sized vessels (Calabrese et al).
- Cardiac injury and thrombotic complications are widely prevalent, including cardiac inflammatory infiltrates, epicardial edema, and pericardial effusion in some autopsies (Falasca et al; Elsoukkary et al; Geng et al).
- Acute kidney injury, while common in hospitalized COVID patients, was found to be mild in post-mortem patients with theoretical potential for recovery (Santoriello et al).
- Neurologic lesions in autopsy series of 43 patients (not necessarily with neurologic manifestations) showed fresh ischemic lesions in 14%, and neuroinflammatory changes with infiltration of cytotoxic T lymphocytes most pronounced in the brainstem (also cerebellum and meninges) (Matschke et al). In patients with significant neurologic decline, more severe findings have been noted including hemorrhagic lesions through the cerebral hemispheres, marked axonal injury, areas of necrosis, and pathology similar to Acute Disseminated Encephalomyelitis (ADEM). (See e.g. Reichard et al).
Literature Review: University of Washington Literature Report (Geographic Spread)
Literature Review: University of Washington Literature Report (Modeling and Prediction)
Prevalence estimates depend significantly on testing availability and percentage of the population that has asymptomatic infection as well as on the severity of the epidemic in a specific location. Seroprevalence studies, measuring antibodies across an entire population, can help give a better estimate of true prevalence. In one meta-analysis of 47 studies on seroprevalence covering 399,265 people from 23 countries, the SARS-CoV-2 seroprevalence in the general population varied from 0.37% to 22.1%, with a pooled estimate of 3.38% (Rostami et al).
COVID-19 transmission is primarily human-to-human following a suspected animal-to-human initiating event (Li et al). It is thought that it may have emerged from racoon dogs or civets, but this is still being investigated (Mallapaty). The virus was initially recognized in December 2019 by Chinese authorities in the setting of cases of pneumonia that seemed to be clustered around a seafood market in Wuhan, Hubei Province (Wuhan Municipal Health Commission, 2019). Laboratory samples collected in December 2019 yielded evidence of a novel betacoronavirus, genetically-distinct from previously identified SARS-CoV and MERS-CoV but genetically-similar to previously-published coronavirus strains collected from bats from southwestern China (Zhu et al).
Updated Date: February 3, 2021
Literature Review (Viral Genetics) Gallery View, Grid View
Tool: Viral genomes have been published to GenBank from diverse geographies.
Tool: Reports on real-time phylogenetic tracking of the viral genome can be found at NextStrain (Hadfield et al).
Tool: CDC emerging variants.
Frequency of new mutations: Mutation of RNA viruses is expected and common, though less common in coronaviruses than many other RNA viruses due to “proofreading” capacity (Robson et al). Several new mutations occurred in SARS-CoV-2 in the fall of 2020 (CDC), and more are likely to occur over time. The meaning of these mutations for transmission and severity is as yet unclear, and likely depends on the exact mutation. Many viruses tend to mutate over time to more transmissible but less virulent strains, though this is not always the case.
Testing, Vaccine, and Antibody efficacy: As new strains emerge, some tests may be able to detect the new strains and some may not: for example, most NAAT testing can be changed rapidly to include novel strains, but some antigen RDTs may not recognize new strains and cannot be altered once they have been manufactured. Most vaccines are targeted at a part of the spike protein that is highly conserved (common between strains) and so vaccination may still provide some protection against many novel strains and complete protection against some strains, but the extent of this protection will depend on the exact mutations in question (which also may evolve over time) and is not yet known. Similarly, prior infection, convalescent plasma, or monoclonal antibodies may provide partial protection against new strains, but the extent is not yet known and it will depend on the exact variant.
Lineage B.1.1.7. was first identified in the UK in September, 2020 and had been found globally by December, 2020. This particular strain rapidly became the dominant strain in London where it emerged, and has been now noted in multiple cities around the globe. It is likely more transmissible than prior strains (15% of contacts get infected relative to 10% for the prior strains, meaning 50% increase in infectivity, Public Health England). As of January, 2020 there is concern that it could also cause more severe disease (BBC), but this has not yet been confirmed. Though this mutation effects the spike protein of the virus, vaccines targeted at the spike protein so far appear to be effective against this strain, but this may change over time.
Lineage B.1.351. first emerged in South Africa in October, 2020 and shares some mutations with B.1.1.7. It also appears to have increased transmission (ECDC).
Lineage P.1. appears to have emerged from Brazil in January, 2021. Little is known at this time, though there is anecdotal concern that it may evade antibodies from prior infection.
Lineage B.1.429 first emerged in Los Angeles, USA. in Summer, 2020 and appears to be growing as a percentage of total cases, implying perhaps greater transmissibility relative to other strains.
Updated Date: December 19, 2020
Patients who are infected with SARS COV-2 and who have higher levels of virus in their respiratory tracts and oropharynx are the most infectious, regardless of their level of symptoms (Bullard et al). Symptom status does not seem to correlate predictably with viral load (Walsh et al; Lee et al; Zou L et al).
Upper airway viral load peaks within ~5 days of symptom onset, followed by decline (Wölfel et al; Young et al). Consequently, patients appear to be most infectious in the 2-3 days before symptom onset and the 2-3 days after (Ferretti et al). PCR detection continues for a median of 20 days from time of symptom onset, with an interquartile range 17-24 days (Zhou et al). There are rare cases that remain positive up to ~60 days after infection (McKie et al). However, the virus is very rarely culturable (our closest proxy to infectivity) after 9 days (Cevik et al). The culture data underlies the newer guidance (after November 2020) about Quarantine time. See Testing for a diagram of test positivity compared with infectivity and symptoms.
Asymptomatic, minimally symptomatic (paucisymptomatic), and pre-symptomatic patients can all transmit the virus (Bai et al; Rothe et al; Furukawa et al), though presence of symptoms is probably associated with increased frequency of transmission. Though it is hard to estimate the prevalence of asymptomatic cases due to testing bias and few population-level studies, one metanalysis found that asymptomatic patients represented 17% or cases, and were 42% less likely to transmit than symptomatic cases (Byambasuren et al). In one study in Beijing, face masks worn by family members of pre-symptomatic COVID-19 patients were shown to be 79% effective (OR = 0.21) at reducing transmission, suggesting that presymptomatic transmission is an important mode of transmission and that masks can be effective at preventing it (Wang et al).
Patients who have recovered from COVID sometimes will have fragments of viral RNA that continue to test positive by PCR. Shedding of viral RNA is longer in more severe disease, or in patients who are immunocompromised. However, recent data shows that this viral RNA does not likely represent infectious virions, but rather parts of the virus that are unable to replicate. As such, the U.S. CDC has changed it’s recommendations on the duration of isolation and quarantine as well as releasing patients from isolation (Cevik et al).
We do not yet know how all available and pending vaccines will perform with respect to asymptomatic infection. This is an evolving area of research, but the data suggests that being vaccinated does NOT preclude asymptomatic infection or transmission from one person to another, and therefore most global agencies recommend that people who have been vaccinated continue transmission reduction measures like masks and distancing (CDC).
Updated Date: December 18, 2020
Literature Review: University of Washington Literature Report (Transmission)
Basic Reproduction Number (R0)
Tool: For global estimates of R0, See Here. For the United States, state-by-state estimates of R0 are available Here (Data from The COVID Tracking Project). Please keep in mind these are merely estimates and all models are fallible.
R0 (R-naught) is a measure of transmissibility. It represents the theoretical number of secondary infections from an infectious individual. This is a property both of the infectiousness of the virus and the behaviors of humans to decrease spread.
- An R0 > 1 is consistent with sustained outbreak.
- An R0 < 1 means an epidemic is declining.
The R0 for COVID-19 is likely similar to, or slightly higher than, many other respiratory viruses, but because it is so highly influenced by human behavior, it can be changed. The initial R0 of COVID in Wuhan in the absence of containment measures was thought to be about 2.5 (Majumder et al). However, R0 declines with control measures (Zhao et al; Riou et al; Flaxman et al; Read et al; Shen et al).
The pre-control R0 of 2.5 is:
- Roughly comparable with the 2002-2003 Severe Acute Respiratory Syndrome (SARS) outbreak (Lipsitch et al; Bauch et al; Wallinga et al).
- Higher than the 1918-1919 influenza pandemic (R0 1.5) (Petersen et al).
- Higher than typical seasonal influenza (average R0 1.3) (Coburn et al).
It is believed that COVID-19 transmission primarily occurs through liquid respiratory particles that travel through the air between people who are within a distance of about 2 meters of one another. Growing evidence indicates that transmission beyond a distance of 2 meters is also possible, especially in poorly-ventilated spaces and with periods of exposure exceeding 30 minutes (Lancet Editorial).
Modes of transmission have been a contested topic. The discussion has been complicated by lack of standardized definitions and usage of terms such as large droplets, small droplets, and aerosols. It is believed that most transmission occurs with droplets produced when a person coughs or sneezes, though other modes also occur.
Droplet Transmission: Liquid respiratory particles vary in size and are produced during breathing, talking, singing, coughing, and sneezing (CDC). Larger particles of 60-100 micrometers typically do not travel through the air farther than 2 meters (Lancet Editorial).
Airborne/Aerosol Transmission: Very small respiratory droplets, often called aerosols, may remain suspended in the air and travel a distance exceeding 2 meters (Lancet Editorial). The risk of producing aerosols is heightened during coughing, sneezing, and certain medical procedures (WHO-China Joint Mission on COVID-19). Aerosolized particles appear to remain in the air for at least 3 hours (Van Dorelmalen et al).
Fomite (Objects and Surfaces) Transmission: Transmission may occur through touching contaminated objects before touching the mouth, nose, or eyes, but this is an inefficient mode of transmission (Kampf et al). While SARS-CoV-2 can persist for days on some surfaces, attempts to culture from surfaces have been unsuccessful. Viral particles have been shown to survive < 24h on cardboard and < 72h on plastic or steel (Van Dorelmalen et al). In cases of suspected transmission through fomites and direct contact, full exclusion of respiratory transmission as the actual mode has not been possible. Transmission through the handling of contaminated objects is presumed to be unusual (Meyerowitz et al). Adherence to standard precautions and disinfection of equipment and surfaces is still indicated (Mondelli et al).
Water and Sewage: Persistence of SARS-CoV-2 virus in drinking-water is possible; indeed, some organizations and public health departments are tracking COVID infection rates by measuring waste-water RNA (see CDC Wastewater Testing, and Larsen et al). There is no evidence to date about survival of the virus in water or sewage, but it is likely to become inactivated significantly faster than non-enveloped human enteric viruses with known waterborne transmission (such as adenoviruses, norovirus, rotavirus and hepatitis A).
- Feces and whole blood have been shown to contain viral ribonucleic acid (RNA) on PCR studies (Wölfel et al; Young et al). Significance for transmission is unclear (Chen et al), though in one systematic review of smaller studies, replication-capable virus was found in 35% of samples (van Doorn et al), meaning that fecal transmission may be possible.
- Urine does not appear to contain viral ribonucleic acid (Wölfel et al).
- Semen and vaginal secretions: COVID-19 virus has not been detected in vaginal secretions (Qiu et al). It is detectable in semen, but transmissibility is unclear. Likelihood of transmission via respiratory secretions during sexual encounters, however, is likely (Sharun et al).
- Tears: a few studies have indicated presence of COVID-19 virus in tears, while others have not. Current evidence is limited, but risk of transmission through tears is thought to be low (Seah et al).
- Cerebrospinal Fluid: Rarely, CSF has been noted to be positive by PCR (in 2 of 578 samples in one study, but not at levels that are infectious) (Destras et al).
Household contacts of an index case appear more likely to contract the virus than other contacts (Bi et al). Most transmission events occur within households (Luo et al). The household secondary attack rate (e.g. number of people who get infected from an index case) is very variable, thought to be about 17.2% in one meta analysis (Fung et al), though very few studies tested more than once, so many cases may not have been missed The results ranged from 10.3-32.4% when contacts were tested at least twice. One recent study that did daily testing estimated SAR at 35% excluding those who had positive tests at enrollment, 53% including cases positive on enrolment. 75% of secondary cases occurred within 5 days of the index patient’s symptom onset (Grijalva et al). However, when prevalence increases, more community (meaning with no known exposure) transmissions tend to occur, highlighting the necessity of non-pharmaceutical interventions (e.g. masks) coupled with public health strategies such as sentinel and syndromic surveillance.
Super-Spreading Events are when an individual directly spreads an infection to an unusually large number of others. Several cases of superspreading have occurred at choirs (Hamner et al), weddings (including a Maine wedding that led to 177 linked cases, including seven deaths), churches Daegu, South Korea, where “Patient 31” infected at least 40 others (Ryall), and even within the White House. SSEs are believed to be disproportionately responsible for COVID-19 cases globally, with several studies suggesting that ≈80% of secondary transmissions have been caused by a small fraction (≈10%) of initially infected individuals. (Althouse et al; Endo et al). SSEs are heavily dependent on sociobiological mechanisms, including individual viral load, numbers of susceptible contacts per person, residence or employment in congregate settings, and ‘opportunistic’ scenarios including temporary clustering of individuals in mass gathering events. Environmental factors also are very important with closed places, crowded places, and poor ventilation playing a significant role in SSEs. Because SSEs play such an outsized role in fueling the pandemic, they amount to a significant concern, but also serve as an opportune area for public health interventions, particularly the prevention of transmission events where over 10 people are infected (Althouse et al).
Schools are unique settings and are likely to contribute to COVID-19 transmission between households and within communities. However, sustained closure of in-person schooling is expected to have an adverse effect on life outcomes for children and to worsen existing inequalities.
The American Academy of Pediatrics advocates that children should be physically present in school where possible (AAP Guidance). However, large-scale studies of over 500,000 contacts of 85,000 infected cases in India have noted that children are a significant source of spread, even despite school closures (Laxminarayan et al). This thorough Review of the Literature on School Transmission and Safety summarizes some of the unique challenges and recommendations (Massachusetts General Hospital COVID-19 Resource Library). Decisions on whether or not to open schools depends significantly on local policy and local epidemiology.
The risk of contracting COVID-19 on airplanes is low. 50% of the air circulated in the cabin is brought in from the outside, and the remaining 50% is filtered through HEPA filters. Air enters the cabin from overhead inlets and flows downwards toward floor-level outlets. There is relatively little airflow forward and backward between rows, making it less likely to spread respiratory particles between rows (Pombal et al). To avoid transmission, it is advised to avoid moving up and down the aisles as much as possible, and to wear a mask for the duration of the flight.
While transmission risk from pets is low, the United States Centers for Disease Control now recommend that social distancing rules should apply to pets as well as to humans (CDC). Dogs showed low susceptibility. Pigs, chickens, and ducks were deemed not susceptible according to early data. Evidence of viral replication was noted in inoculated ferrets and cats, with viral transmission occurring between cats (Chen et al). There is no current evidence of transmission to humans from cats and ferrets, though minks can transmit to humans (Meyerowitz et al). Virologist cited in Nature News suggests cat owners should not yet be alarmed, noting deliberate high-dose inoculation of said cats - unrepresentative of day-to-day pet/owner interactions, and that none of the infected cats developed symptoms in the aforementioned study (Mallapaty et al).
Experimental data suggest that the persistence of SARS-CoV-2, either on surfaces or while airborne, is somewhat sensitive to environmental conditions such as temperature, humidity, and ultraviolet radiation. Comparable environmentally-sensitive respiratory viruses often demonstrate seasonality, with greater numbers of infections during winter, and so it seems plausible that SARS-CoV-2 might demonstrate a similar pattern (Carlson et al). However, further studies suggest minimal (≈1%) reductions in SARS-CoV-2 transmission linked to environmental UV radiation (Carleton et al), and the current consensus on such environmental effects is that they are minor in real-world circumstances.
Other respiratory infections such as influenza manifest seasonal oscillations; ‘cold and flu season’ occurs when population susceptibility is high and environmental drivers such as lower temperatures, humidity, and solar radiation conspire to increase transmission, often by changing human behaviors (forcing people indoors). But current levels of immunity to SARS-CoV-2 in most countries are low enough that summer weather is not likely to be protective (Baker et al). If the virus ultimately becomes endemic, it is likely that seasonal oscillations will be observable in temperate regions, with recurrent wintertime outbreaks likely (Kissler et al).
Updated Date: December 18, 2020
The majority of patients with RT-PCR-confirmed COVID-19 develop antibodies against the virus (Zhao et al; Wang et al). These two large series of serial samples found antibodies in 161/173 and 308/310 patients respectively. Time to seroconversion, correlation with protection, and durability of immunity, continue to be studied.
When assessing research studies, details may depend on exactly which antibodies are being assessed(e.g,. IgA/IgM/IgG or total antibody, antibodies to the nucleocapsid vs those to the spike protein, or whether the antibodies are “neutralizing” -- for example, antibodies directed against the receptor-binding domain (a component of the spike protein) may appear earlier than antibodies to other antigens (To et al; Okba et al).
Seroconversion (detection of circulating antibodies) typically occurs 7-14 days after symptom onset (Deeks et al; Huang et al). In one study of 173 patients, 100% were seropositive (total antibody) at 15 days (Zhao et al).
- Although IgM seroconversion is often thought of as occurring before seroconversion for IgG, this has not been consistently observed for SARS-CoV-2 (e.g., Qu et al; Xiang et al; Wang et al; Zhao et al).
- IgA antibodies are important in mucosal immunity and may play an important role in the response to SARS-CoV-2 (Sterlin et al; Wang et al), but data are currently limited (Deeks et al).
- The sensitivity of serology (IgM or IgG) may be higher than that of PCR by the second week of illness (day 8 in Zhao et al, day 6 in Guo et al), based on studies with serial samples from individual patients.
- Antibody detection may identify cases with negative upper airway PCR but high clinical suspicion when timed appropriately found positive IgM in 54 of 58 probable cases without detectable nucleic acid (Guo et al).
Neutralizing antibodies prevent viral replication, usually by binding the spike glycoprotein that SARS-CoV-2 uses to enter cells. Not all antibodies are neutralizing; some bind to the virus but do not stop its activity.
- Most diagnostic tests detect antibodies without specifying whether they are neutralizing. The first test to receive a U.S. FDA EUA for specifically detecting neutralizing antibodies is the cPass SARS-CoV-2 Neutralization Antibody Detection Kit, by GenScript, USA.
- Understanding which antibodies are neutralizing is critical for Vaccine Development, Monoclonal Antibody Therapy, studying Convalescent Plasma, and determining whether seropositive individuals are Immune from Reinfection.
- Fortunately, over 90% of people seropositive for SARS-CoV-2 appear to have detectable neutralizing antibody responses (Wajnberg et al).
Cross-reactivity occurs when a pre-existing immune response to seasonal human coronaviruses recognizes SARS-CoV-2. It has been seen in both T-cell (e.g., Mateus et al) and humoral (e.g., Ng et al) arms of the immune system. This presence or absence of this response has been hypothesized to contribute to variable outcomes in COVID-19 (Beretta et al). A possible altered immune response due to pre-existing antibodies has also raised the possibility of antibody-dependent enhancement after SARS-CoV-2 vaccination or convalescent plasma therapy. Fortunately, such enhancement has not been reported to date (Wen et al).
Reinfection refers to individuals who have been infected and cleared the original virus, but again show evidence of viral replication after exposure to a new SARS-CoV-2 virus (Falahi et al).
- Reinfection is different from post-COVID-19 syndrome, relapse/reactivation (also called recrudescence) and repositivity (NICE guidelines 2020-12-18, Yahav et al.)
A person who is asymptomatic but tests positive after resolution of COVID-19 may have (1) residual shedding of RNA fragments or viral particles (not necessarily infectious; see Infectivity) from the initial infection, or (2) reinfection after exposure to another SARS-CoV-2 virus. A person who has new or prolonged symptoms and a positive test after resolution of an initial COVID-19 diagnosis could either have (1) ongoing or post- COVID-19 physiological injury in the absence of replicating virus, (2) recrudescence of residual virus not fully cleared after the initial infection, or (3) reinfection after exposure to another SARS-CoV-2 virus.
- Conclusive demonstration of reinfection is difficult outside of research settings, since confirmation requires analysis of paired viral whole genome sequences taken during both initial and subsequent infections (ECDC Threat Assessment Brief 2020-09-21).
- Analysis of over 130000 RT-qPCR positive tests in Qatar, looking for repeat positive swabs ≥45 days after an initial test with genomic confirmation where possible, estimated a reinfection risk of 0.02% over the course of their study for an incidence rate of 0.36 per 100000 person weeks. Subsequent comparison of antibody positive and negative cohorts estimated that antibodies from natural infection conferred ~95% protection (Abu-Raddad et al, Abu-Raddad et al).
- Another large study looking at over 160000 RT-qPCR positive tests in Denmark looking for paired positive results, but without genomic confirmation found a lower estimated protection from natural infection of ~81%, dropping to ~47% in those aged 65 years and older (Hansen et al).
- Many factors make comparison of these large studies difficult, including the populations being studied, indications for testing, and how both cohorts and reinfections are defined. Decreased protection after asymptomatic infection or in the absence of antibody development may explain some of these differences.
Tool: Track Cases of Reinfection.
The duration of immunity after infection is not yet conclusively known (Iwasaki). Relevant host factors may include immune status, age, and severity of initial infection. Studies documenting decay of IgG antibodies may underestimate immunity, since T-cell responses likely also play a significant role (Karlsson et al).
- Patients with mild infection lose detectable antibodies more quickly but may have an immune memory that allows them to rapidly produce antibodies on re-exposure (Stephens et al).
- Longer term analysis of 2003 SARS-CoV-1 infection showed IgG and neutralizing antibody titers peaked at 4 months and diminished over a 3-year follow up period (Cao et al). In a cohort of 23 patients with a history of SARS-CoV-1 infection, specific IgG Ab to SARS-CoV-1 was undetectable in 91.3% of former patients after 6 years (Tang et al).
- The duration of immunity after vaccination and the breadth of the immune response between different viral variants are actively being studied (Fergie and Srivastava).
By the end of 2020, there were approximately 75 vaccines worldwide in phases I through III of clinical research, 5 in limited use, and ~2 holding approvals (depending on country) for full use after completion of phase III (not counting trials underway for repurposing of pre-existing vaccines). Most of these vaccines fall into one of 4 categories:
- Genetic Vaccines (typically lipid envelopes carrying SARS-CoV-2 genetic material into cells)
- Viral Vector Vaccines (repurposed viruses such as adenovirus carrying SARS-CoV-2 genetic material into cells)
- Protein-based Vaccines (delivering coronavirus proteins only)
- Traditional “inactivated/attenuated” Coronavirus Vaccines
The best-studied vaccines are the mRNA vaccines produced by Pfizer/BioNTech and Moderna, and key “viral vector” vaccines include adenovirus-vector examples produced by Oxford/AstraZeneca, CanSino Biologics (China), the Russian Ministry of Health, and Johnson & Johnson/Beth Israel Deaconess Medical Center.
- The mRNA Vaccines: Code for the coronavirus “spike” protein to induce an immune response mediated by antibodies and T cells. Both Moderna and Pfizer versions have a greater than 90% efficacy after 2 doses (94.1 and 95%, respectively), although longer-term study is needed to determine how immunity and infectivity change over time. Due to the temperature-sensitivity of genetic material (Simmons-Duffin), these vaccines require somewhat different but very cold storage and transportation environments as well as restrictions on how long they can be kept at room temperature.
- Adenovirus-vector Vaccines: Use a variety of engineered adenoviruses (common viruses that cause colds and related symptoms) as vectors to expose human cells to the SARS-CoV-2 spike. These are generally cheaper than genetic vaccines (Knoll et al), both due to ease of transport as they only require refrigeration to protect the virus vector, as well as due to less costly supply chains than those required for mRNA vaccine technology.
Efficacy of Vaccines. Efficacy data is currently available for four vaccines. The Pfizer/BioNTech vaccine in its EUA cited efficacy of 95% at preventing symptomatic infection. Asymptomatic infection was not studied in this trial (Pfizer EUA). The Moderna vaccine showed 94% efficacy, and asymptomatic infection rates are not yet known after two doses, but after a single dose it reduces asymptomatic infection by 63% (data after the second shot is forthcoming, Moderna EUA). The Oxford/AstraZeneca vaccine had an efficacy of 90% at one dosing schedule (the now intended schedule) though they had two different dosing schedules and the other one did not perform as well. In the intended dosing schedule, it demonstrates 59% efficacy in preventing asymptomatic infection (Ledford; Knoll et al). Janssen/Johnson & Johnson report an efficacy of 77% in preventing severe/critical COVID-19 at 14 days after their single dose vaccine, rising to 85% at 28 days; efficacy at preventing moderate disease was lower (Janssen EUA).
Adverse Events/Reactogenicity: Most observed adverse events during vaccine trials were injection-related or reflected the an expected immune response. Many people feel ill following vaccine administration for about 1-3 days, especially after the second dose of the vaccine This is not a sign of infection by the coronavirus.
- There have been reports of rare venous thrombotic disease -- and particulary cerebral venous sinus thrombosis -- in recipients of the widely deployed Oxford/AstraZeneca and Janssen/Johnson & Johnson adenovirus vector vaccines.
- Cerebral (also “central” or “dural”) venous sinus thrombosis refers to a blood clot in the veins that drain blood flow from the brain. Obstruction of outflowing blood can lead to increased intracranial pressure and, depending on the anatomy of the clot, focal neurological symptoms that are a type of stroke.
- Although the mechanism of this clotting dysfunction is not certain, it appears to be a vaccine cross-reaction that causes an auto-immune thrombocytopenia.
- There appears to be an increased risk in women under the age of sixty, possibly related to estrogen levels.
- For both vaccines, the frequency of these events appears to be lower than the risk of thromboembolic complications of COVID-19. As of April 15, 2021, the benefits of both Oxford/AstraZeneca and Janssen/Johnson & Johnson vaccines in a pandemic context are still thought to outweigh potential risks.
- Given the appropriately high bar for safety of vaccines, however, deployment of both of these vaccines has been paused in many jurisdictions, as agencies try to better understand any causal link between the vaccines and these thrombotic complications.
Ability to Transmit to Others: We do not yet have information about viral transmission, but based on asymptomatic infection rates it may still be possible, though very unlikely, to transmit the virus to others even if vaccinated and asymptomatic. Please see Transmission.
Prior Infection or Antibody Therapies: People who have been previously infected and/or received antibody therapies (monoclonal antibodies, convalescent plasma) can receive the vaccine. The U.S. CDC recommends waiting 90 days before vaccination. If a vaccinated person becomes infected, they can still get antibody treatment after vaccination. (CDC)
Contraindications: The U.S. CDC considers the following contraindications: severe allergy (e.g. anaphylaxis) to a prior dose of an mRNA COVID vaccine or any of its components, immediate allergic reaction of any severity to a previous dose or any of its components (including PEG), immediate allergic reaction of any severity to polysorbate. (CDC) Reactions to non-COVID vaccines are considered a “precaution” but not a contraindication.
Routine Vaccinations: The U.S. CDC is currently recommending that there be no other vaccinations within 14 days of COVID vaccination (before or after) so that side effects due to the COVID vaccination can be more effectively tracked.
- Obstetrics: There are currently no data on the safety and efficacy of vaccines in pregnant people or lactating people. No safety concerns have been demonstrated in animal models of the Moderna vaccine. Because the vaccine is not a live virus and the mRNA is degraded very quickly, many experts believe that mRNA vaccines are unlikely to pose a risk to pregnant people or fetuses. mRNA vaccines are not thought to be a risk to a breastfeeding infant. Pregnant women and their doctors may discuss the risks and benefits depending on the individual’s risk of acquiring COVID, although this is not required. Side effects, such as fever, can sometimes cause adverse pregnancy outcomes.
- Pediatrics: Currently the Pfizer vaccine is authorized for adolescents aged 16-17, Moderna is restricted to people over 18 years. No vaccines are currently available for children under 16.
- Immunosuppressed: Many immunosuppressed people can get the vaccine as it does not include live virus, but it may not be as efficacious as in those who are not immunosuppressed. Antibody testing is not necessarily reliable in predicting if there has been an immune response. Revaccination once immunocompetent might be recommended in the future, though there is currently no data on this.
- Autoimmune Conditions and History of Guillain-Barre: Insufficient data is available on these populations, though people with autoimmune conditions were included in trails and did not seem to have increased symptoms. To date no cases of Guillain Barre have been found with the mRNA vaccines, and it is not a contraindication to vaccination
Tool: Vaccine Allocation Planner (helps states and countries plan vaccine allocation)
Tool: COVID Vaccine Development Tracker
Tool: FDA COVID Vaccine FAQs
Tool: NEJM Vaccine Resources
Tool: NEJM Vaccine FAQ
While approval of the first vaccine marked the culmination of a tremendous scientific effort, the fight against COVID-19 now faces a new challenge: a massive worldwide vaccination campaign. The same embedded structural forces driving inequities in the burden of COVID-19 must also be considered within the context of vaccine access and distribution.
Vaccine Prioritization: It is essential that COVID-19 vaccines be distributed equitably. People who should be prioritized for vaccination include (adapted from the National Academies of Sciences, 2020).
- High-risk of COVID-related Morbidity and Mortality
- Medical Comorbidities
- Over the age of 65
- High-risk of Contracting COVID-19
- Residents of Long-term Care Facilities and Group Homes
- Healthcare Workers
- Front Line Workers (e.g. Supermarkets, Factories, Schools, Agriculture, and Meat-processing Plants)
Due to generations of structural racism and socioeconomic inequalities, people of color, people with disabilities, immigrants and migrants, indigenous peoples, and the poor are all disproportionately represented in many of these groups.
Global Distribution: The distribution of vaccines among nations should follow similar principles. No country should have enough vaccines to vaccinate their entire population before another country has enough to vaccinate their high-risk populations. As of December, 2020 there is significant imbalance: Canada has ordered enough vaccines to inoculate six-times its population, the United Kingdom and the United States four-times their populations, and the European Union twice its population (New York Times).
COVAX, a global coalition including the WHO to assure vaccination, has proposed that all countries receive an adequate supply to inoculate at least 20% of their population before any nation receives additional vaccines. This will ensure that high-risk groups are vaccinated regardless of where they live. Following this initial roll-out, vaccines should be distributed based on the vulnerability of the country’s health system and the impact of COVID-19 on the country, prioritizing countries most in need (COVAX, 2020).
Vaccine hesitancy: In countries like the US, vaccine hesitancy and distrust of the medical system may further exacerbate inequity (Warren et al). This is shaped by the legacy of exploitation and oppression of marginalized groups in the name of science (for example, the Tuskegee Experiment). Meaningful community engagement and promotion of informed decision-making requires an acknowledgment that these historical and contemporary forces contribute to a rational distrust of the health system among marginalized communities (Burgess et al).
Updated Date: January 24, 2021
The minimum or critical level of population immunity - acquired either through immunization or previous infection and subsequent recovery - that is required to stymie transmission of a particular communicable disease is colloquially referred to as ‘herd immunity’. When ‘herd immunity’ is achieved, susceptible individuals are indirectly protected from infection because sufficient numbers of immune individuals serve to prevent the circulation of the pathogen to immunologically-naive individuals. The percentage of immune individuals required to achieve ‘herd immunity’ against a particular pathogen varies dramatically depending on factors such as the baseline reproduction rate of the pathogen (R0), the effective reproductive number for a given population (Rt) - which is itself influenced by the efficacy of (and societal adherence to) non-pharmaceutical interventions, population density, therapeutics, immunological factors like the length of immunity, etc.
Current estimates suggest that achieving ‘herd immunity’ against SARS-CoV-2 will not be possible without an absolute minimum of 50% of population immunity (Fonanet et al), and as high as 85% in countries with higher Rt values (On Kwok et al). Because of the significant case fatality rate of COVID, and the ancillary consequences of unnecessary cases and deaths, the WHO recommends that ‘herd immunity’ against SARS-CoV-2 be achieved through immunization campaigns and not by needlessly exposing populations to the pathogen.
Updated Date: December 17, 2020
Equity focuses on the fair and just treatment of all people. By extension, addressing inequities involves eliminating avoidable, unfair or changeable differences among groups, whether these are defined socially, economically, demographically, or otherwise. Upholding equity in health allows prioritization of fair opportunities for everyone to attain their full health potential (WHO Health Systems: Equity).
The COVID-19 pandemic has disproportionately affected historically oppressed populations around the world. Due to long-standing structural inequities, people from these communities are: 1) more likely to be exposed to disease, working essential jobs and living in crowded conditions; 2) less likely have to have access to quality healthcare, including COVID-19 testing and treatment; and 3) more likely to suffer from preexisting health conditions, as a result of adverse social determinants of health, putting them at increased risk of complications and death (Warren et al).
Not all of these can be included here, but we will address several major concerns. Inclusive data collection, while important, needs to be followed by evidence-driven steps to create an inclusive pandemic response and to be the foundation for equitable public health emergency planning (Reed et al).
Providers should screen for and Address Social Determinants of Health (SDOH): SDOH are the conditions under which people are born, grow, live, work, and age (AAFP's The EveryONE Project) which act to shape the health and well-being of people in complex ways. In the context of COVID-19, living situations coupled with job insecurity increase the risk of infection and then make safe isolation and quarantine difficult. In some neighborhoods in the United States, as many as 70% of positive cases required social support to safely isolate and quarantine (Kerkhoff et al).
Updated Date: January 20, 2021
Wealth inequality functions as both a cause and effect of health inequity. The global imbalance of wealth among and within nations is the result of historical (and current) forces including colonialism, racism, structural readjustment, and extractive capitalism. This has left many countries with chronically underfunded health systems lacking in infrastructure, equipment, and adequate staffing.
Historically, in the face of these challenges, containment measures are often emphasized over provision of treatment and supportive care. As was seen in the Ebola epidemic, this strategy backfires by ignoring the human toll of weak treatment systems, and downplaying the impact that effective treatment has on containment: When treatment and supportive care are not available or are not high quality, it undermines confidence in public health institutions and messaging; people understandably avoid seeking care when they need it, and may not trust public education campaigns encouraging social distancing, isolation, and other precautionary measures.
It should be noted, despite facing significant barriers to containment and treatment, a number of low- and middle-income countries have prevented COVID-19 cases and fatalities from reaching the astronomical levels seen in many wealthier nations.
The COVID pandemic has led to a global income drop for workers and exacerbated existing health gaps between rich and poor countries (AP News). Disruptions to food supplies and economies risk worsening malnutrition worldwide, and will be a severe setback to the effort to achieve the United Nations Sustainable Development Goals (Ekwebelem et al). Additionally, the pandemic will leave fragile health systems with a legacy of death and attrition in the workforce and shrinking budgets driven by unstable financial outlooks.
In the United States and United Kingdom multiple sources have demonstrated that Black and Latinx populations are disproportionately likely to be infected and/or die from COVID-19 (Garg; NYSDOH Fatalities, NYC DOH). A systematic review and meta-analysis of over 18 million patients across 50 studies from these two countries found higher COVID infection rates within Black, Latinx, and Asian communities (Sze et al). As of late November, 2020 Black and Latinx Americans hav