Short Answer: Appears to be similar

Roger Paredes, MD, PhD (1/14/2021):

As far as we know, the mutation rate of B117 (UK) appears to follow the same kinetics as other variants.^ECDC Indeed, the high number of mutations in the spike protein and other genomic properties of the variant, combined with the high sequencing coverage in the United Kingdom, suggests that this variant did not arise through gradual accumulation of mutations. The question then is: How did this variant arise? One potential explanation for its emergence is prolonged interhost evolution in a single patient, potentially with reduced immunocompetence.^{Choi 2020} Another is that certain interventions, such as convalescent plasma, might be selecting for mutants with a higher number of mutations. Finally, it is important to note that these coronaviruses can recombine, thereby giving rise to new variants in this manner.^{Bobay 2020} Unfortunately, we may just be seeing the beginning of such events.

Short Answer: Mostly likely

Roger Paredes, MD, PhD (1/14/2021):

The new SARS-CoV-2 variants originating from the United Kingdom and South Africa have been found to be more transmissible.^{CDC New Variants} As such, if a person were exposed to the same amount of virus, it is more likely that he or she would become infected with B117 or B1351 vs the original SARS-CoV-2 variant from Wuhan or other circulating variants. However, there are no studies to date confirming this.

Short Answer: Data are emerging

Roger Paredes, MD, PhD (1/14/2021):

The newest SARS-CoV-2 variant, called P1, was identified at the Haneda airport in Japan in 4 individuals traveling from Brazil to Tokyo upon routine screening.^{CDC New Variants} As with B117 (UK) and B1351 (South Africa) SARS-CoV-2 variants, P1 appears to be more transmissible, but there is currently no evidence that it causes more severe illness or is associated with an increased risk of death. The variant contains many of the mutations present in the B117 and B1351 SARS-CoV-2 variants but also contains a set of additional mutations that may affect its ability to be neutralized by antibodies.^{CDC New Variants} There is suspicion that the rise of this new variant might be the cause of increased cases in Manaus, Brazil.

Short Answer: Not everywhere

Roger Paredes, MD, PhD (1/14/2021):

It is likely that in most parts of the world, we are not sequencing enough to know if the rise in cases is due to the more transmissible variants of SARS-CoV-2. For example, in Spain, we generally see 60 variants if we sequence the virus in 60 patients positive for the infection. However, I believe that is because we are not sequencing enough to see the overlap. In Denmark, where an extensive sequencing program is underway, they are seeing a progressive increase in the prevalence of the UK variant, and they are expecting that this variant will become the dominant strain by mid-February.^ECDC Overall, it is clear that both B117 (UK) and B1351 (South Africa) variants are becoming the most frequent ones in places where they have been introduced, and adequate widespread sequencing would confirm this.

Short Answer: Not likely

Roger Paredes, MD, PhD (1/14/2021):

Although both B117 (UK) and B1351 (South Africa) SARS-CoV-2 variants share the N501Y mutation and several other mutations, current evidence suggests that they emerged completely independently and are not related.^{CDC Emerging SARS-CoV-2 Variant}

Short Answer: Not yet

Vikramjit Mukherjee, MD (12/18/2020):

If there is a small silver lining in this pandemic, it is that influenza has taken a backseat so far. In the southern hemisphere, we saw that their typical influenza season, which occurs during our summer, had extremely low influenza activity.^{Olson 2020} That was largely due to the same mitigation strategies that are in place to prevent the spread of COVID-19, including social distancing, wearing masks, and consistent handwashing.

Although it is still early in the influenza season, here in New York and in many other parts of the United States, the number of patients coming into hospitals and ICUs with influenza has been abnormally low. Regardless, every time we test for COVID-19, we also test for influenza. At my institution, our testing panels include SARS-CoV-2 and influenza A and B together.

Short Answer: Determines proportion of population infected, but utility may decrease over time

David Heymann, BA, MD, DTM&H (10/28/2020):

Serological testing is very useful for determining where infections have occurred. However, a recent study from the United Kingdom has shown that antibodies detectable by currently available tests rapidly decrease over time.^{Ward 2020} Therefore, after 3 or 4 months, it is often no longer possible to detect antibodies that were detectable closer to the time of infection. In addition, in people who have had very mild illness, it is sometimes not possible to detect antibodies at all, even soon after the infection has occurred or after the PCR test becomes negative. As a result, there are questions as to whether or not we will be able to continue detecting previous infections moving forward. In some countries such as Sweden, the percentage of the general population that appears to have been infected is approximately 15%,^{Ioannidis 2020; Rostami 2020} but in many countries, it is < 10% based on antibody testing.^{Pollán 2020; Anand 2020} Regarding what this means for SARS-CoV-2 immunity, it is important to remember that cellular immunity is not assessed by antibody detection, and we are just beginning to understand the role of cellular immunity in protection against infection. Cellular immunity is very difficult to measure, so it is not used to determine where infection has occurred in a community—rather the antibody tests are the best tool for that assessment.

Short Answer: Not understood, studies underway; social determinants of health and resulting comorbidity differences may play a role

David Heymann, BA, MD, DTM&H (10/28/2020):

The reason for these differences is not yet understood. There are many cohort studies underway in the United Kingdom and North America to understand the causative factors. One hypothesis is that inequitable access to health promotional messages and preventive care have resulted in disproportionate rates of comorbidities across racial groups. It is important to consider the multitude of possible contributing factors, including social determinants of health, a potential underlying genetic association, and comorbidity prevalence, among others. But the answer is not yet known. What we do know is that everyone should take care to protect themselves and others, especially those who have comorbidities. Regarding social determinants of health, we clearly must do a better job in equalizing healthcare access in many countries. That means making sure that people have access not only to clinical medicine when they need it but also to the public health and health promotion messages that are so important to help people have a healthier lifestyle—because it is through healthy lifestyles and access to preventive care that we can prevent the comorbidities associated with increased mortality risk in people with COVID-19.

Short Answer: Speed, cost, and sensitivity

Vikramjit Mukherjee, MD (12/18/2020):

There are advantages and disadvantages to each test. RT-PCR tests in general have high sensitivity but require more time to obtain results.^{Premraj 2020} Antigen tests provide quicker turnaround time but have lower sensitivity compared with PCR. My suggestion is that if you have a high pretest probability of a positive SARS-CoV-2 test—such as patient with acute respiratory distress syndrome, cough, leukopenia, etc—interpret a negative antigen test with caution.

David Heymann, BA, MD, DTM&H (10/28/2020):

What is currently known from testing validation studies conducted by the Foundation for Innovative New Diagnostics, a global nonprofit, nongovernmental health organization based in Geneva, Switzerland, is that the antigen detection tests are a bit less sensitive than PCR testing,^FIND which is the gold standard. However, there may be an important role for antigen tests for surveillance in areas where PCR testing is not available. Algorithms are being studied to evaluate how the 2 types of testing may complement each other. Antigen detection is being used in countries that cannot afford PCR testing because the lateral flow antigen detection tests are less expensive and can provide more rapid results. In countries or regions where PCR results are taking 24 hours or longer, an antigen test may provide an advantage even with reduced sensitivity because it would allow identification of many cases sooner to facilitate more rapid case isolation and reduce transmission risk. Using antigen and PCR testing in tandem would help to overcome the reduced sensitivity of antigen testing because PCR results would follow, thereby identifying additional cases that were initially negative by antigen testing.

Short Answer: Limited data thus far

David Heymann, BA, MD, DTM&H (10/28/2020):

There is a limited amount of evidence right now. Countries that have assessed households of children who have gone back to school are not finding that there is an increased level of infection.^{Ismail 2020; Russell 2020} That information provides a possible early sign that children are not bringing COVID-19 home from schools at substantial rates. Some schools that have opened have had very low case numbers among enrolled children, even over the course of several months.^{Buonsenso 2020; Russell 2020; Otte Im Kampe 2020} However, because data are limited, the impact of schools on overall community transmission is not fully clear. I think most countries now feel that it is important for children to socialize and to be back in the school setting rather than attempting distanced learning with only some having Internet/computer resources and parents available to assist. For that reason, many countries are trying to move toward getting children back in school as much as possible while monitoring very closely to see if transmission is increasing in their families and communities.

Short Answer: Multifactorial

David Heymann, BA, MD, DTM&H (10/28/2020):

That is a good question, and some of the reasons suggested are as follows. When the lockdown occurred in many countries or many states in the United States in February and March, countries really had no exit strategies. Therefore, when they decided to stop these measures, they just stopped them across the board. The result was that people were pleased to be out of the lockdown situation and began congregating in night clubs, restaurants, and other areas where they could have a drink together and socialize, especially younger people. It is speculated that we are now seeing the result of increased transmission that began when that happened and has now spread elsewhere into their families and communities. That is one hypothesis. Another hypothesis is that as people move indoors because of colder weather, transmission is easier, especially if they are not wearing protective equipment such as masks and especially in household settings. Until recently, it was thought that it took quite a substantial amount of time in proximity for transmission to occur. But it has since become clear that it may take much less time, as has been advised by the CDC.^{CDC Contact Tracing}

Short Answer: No

David Heymann, BA, MD, DTM&H (10/28/2020):

There is no confirmed explanation for this observation, and many questions and hypotheses have emerged about why this might be occurring. One suggestion is that men may be more likely to have comorbidities that increase risk for more severe disease than women. However, this is only speculation at this time.

Short Answer: Not at this time

Lynora Saxinger, MD (9/15/20):

Investigators are tracking the degree of mutation over time to monitor for changes in viral genome. The mutation rate is relatively slow (estimated at approximately 23 substitutions per year) compared with some other viruses, such as influenza and HIV,^GISAID and the mutations that are accruing do not appear to have had any major effects on clinical outcomes. There has been a small suggestion that there may have been mutations in the gene encoding the spike protein that might end up affecting disease, but I do not think the potential effect is significant enough to change our management or our impression of COVID-19 at the moment.^{Young 2020; Volz 2020}

Short Answer: 10 days, generally repeat testing not recommended

Lynora Saxinger, MD (9/15/20):

Except for critically ill and immunocompromised patients who might shed viable virus for longer, most people do not shed virus beyond 8 days. Indeed, available data suggest that most transmission occurs within the first 6 days of symptoms, as well as several days before the onset of symptoms.^{CDC Isolation; Cheng 2020} However, prolonged viral load detectability does lead to some questions. I think most healthcare facilities have now shifted away from repeat testing because viral load does not provide any useful information about transmissibility. Therefore, I routinely discourage retesting. Most guidelines in Canada advise that 10 days after the onset of symptoms or symptom resolution, people can return to normal activities without concern for transmission.^{CA Public Health Agency}

There is a bit of a nuance, however, for patients who have been critically ill and/or are experiencing a prolonged convalescence. These patients might not be asymptomatic, but their symptoms of active infection—malaise, fever, acute cough, runny nose, etc—should be resolved. Patients who have been intubated, for example, might still have symptoms 10 days out. This nuance needs to be considered when applying the guidelines. Thus, if there are no longer symptoms of active infection and it has been more than 10 days since the onset of symptoms, the likelihood of still being able to transmit infection is very low. I think retesting is probably not the best use of resources right now, and I would discourage retesting because it does not provide useful information in the case of prolonged convalescence.

Short Answer: No

Paul E. Sax, MD (9/1/2020):

The saliva test detects SARS-CoV-2 RNA^{FDA Saliva} and avoids the nucleic acid extraction step, making it a simpler test that shows very comparable test performance (~ 93% positive agreement) to PCR (modified CDC assay). It has a limit of detection of 6-12 SARS-CoV-2 copies/µL.^{Vogels 2020}

The PCR test, which is the most sensitive test, should continue to be used for sick patients in the hospital or individuals with symptoms in the outpatient setting. However, I consider there to be 2 types of testing needs. The first is the one I just described—testing for symptomatic individuals with suspected COVID-19—and the other is testing as a public health tool to find asymptomatic cases in the community. A highly sensitive test could be used in both cases, but we do not have a sufficient number of PCR tests or PCR testing capacity for widespread public health testing. For widespread community testing of asymptomatic individuals, where the likelihood of having infection is low, even a test that is less sensitive than PCR would be very useful because when the probability of infection is already low, a negative test further lowers that probability. The ideal test for this approach would be a point-of-care option with rapid turnaround—a 15-minute pregnancy test equivalent. We have home testing for HIV; we have home testing for lots of things. 

Short Answer: Interesting information, still evolving

Paul E. Sax, MD (9/1/2020):

The first case of reinfection that I heard about involved a completely healthy person returning to Hong Kong. Everyone who comes through the Hong Kong airport is tested and this person tested positive. A comparison of the viral sequences obtained from this relapse and his first episode showed that the viruses were phylogenetically different, and the conclusion was made that he had a reinfection. He subsequently developed a nice antibody response, and that is exactly how you want the immune system to work—your immune system comes in contact with an infection that you have had before, you fight it off, and you do not get sick. This was kind of good news.

However, another case of reinfection was reported in Nevada where a 25-year-old immunocompetent man contracted COVID-19. Approximately 2 months later, he was in the household with his parent who contracted COVID-19. Although we do not know whether he practiced social distancing or he cared for his sick parent thinking he was immune, his second illness was actually more severe. He developed hypoxemia and pneumonia and was admitted to the hospital. We do not know the patient’s current condition, but we do know that his second episode was worse than the first.

As of the time of this writing, at least 10 cases of confirmed reinfection have occurred in Hong Kong, the United States, India, Ecuador, Netherlands, and Belgium. Reinfection does happen and it raises many questions: (1) How often does reinfection happen? (2) Will the second episode be mild, moderate, or severe? (3) Does severity depend on inoculum size? (4) What does reinfection mean for vaccine development and how often a vaccine needs to be given?

It is important to remember that we are really in the infancy stage of understanding this infection. There is a lot still unknown about the immune system’s response to this entirely new disease. 

Short Answer: Likely not infectious

Vikramjit Mukherjee, MD (12/18/2020):

Current guidance from the CDC recommends against retesting individuals who have tested positive for SARS-CoV-2.^{CDC Isolation} Instead, they recommend a symptom-based approach to ending isolation, such that individuals can return to normal activities 10 days after symptom onset and 24 hours after fever resolution. This is based on a body of evidence suggesting that patients with COVID-19 rarely shed infectious virus after 10 days—or up to 20 days in some severe cases. Although viral RNA has been detected up to 12 weeks after symptom onset, the data suggest that this is replication-incompetent viral fragments that are not capable of transmitting infection to others.^{Korea CDC 2020; Li 2020a; Xiao 2020}

Leo Yee-Sin, MBBS, MPH, MRCP, FRCP, FAMS (8/17/20):

Prolonged, detectable viral shedding is a general phenomenon that has been reported in Korea and around the world. In a report of 378 cases, the median duration of viral RNA PCR shedding was found to be 53.5 days (interquartile range: 47.75‐60.5) in respiratory samples of known COVID-19 cases, including a case in which the last positive PCR result was 83 days post-infection.^{Li 2020a} The Ct value of the respiratory samples was on the high side, indicating that the viral load was relatively low in this group of individuals. In Singapore, we have also observed prolonged viral shedding by PCR detection in some patients. We have also done multiple virus cultures from these patients, and we were not able to culture virus from samples with very high Ct values. This finding suggests that patients with such prolonged detection of low levels of virus are likely no longer infectious. 

Short Answer: Median 5 days

Vikramjit Mukherjee, MD (7/8/2020):

We know that the incubation period for COVID-19—that is the time from exposure to disease onset—can range from 2 and 14 days,^{CDC Symptoms} but the median is approximately 5 days.^{Lauer 2020} If someone is exposed to an infected individual, he or she may acquire infection at an approximate median of 5 days later. At this point, a nasal swab for SARS CoV-2 would start to be positive. In some reports, swabs for SARS-CoV-2 were shown to yield false negative results in up to approximately 30% of cases.^{Woloshin 2020} So, a nasal swab can be negative in someone with a true infection, particularly if they are tested too early in the course of infection. If other signs and symptoms suggest active COVID-19—for example, exposure to a confirmed case, upper respiratory tract infection, fever, myalgia, or other key symptoms­—a negative nasal swab PCR test should be interpreted with caution.

Short Answer: Yes

Vikramjit Mukherjee, MD (7/8/2020):

Estimates from evolving data suggest that approximately 20% to 25% of people infected with SARS-CoV-2 will be asymptomatic, but some models suggest that the range may be higher.^{CDC Pandemic Planning} These individuals may be unaware of their infection. Public health guidance on community-related exposure from the CDC recommends that if you are exposed to SARS-CoV-2, you should self-isolate for 14 days from the last exposure.^{CDC Community Exposure} Self-isolation means staying at home and keeping a distance of at least 6 feet from other people. You should avoid contact with people who are at higher risk of severe illness due to COVID-19, such as older patients and people with other medical conditions. Everyone, regardless of exposure, should follow personal infection control practices, such as washing hands often and wearing a mask.^{CDC How to Protect} Masks primarily control the source of infection. In other words, if I am infected, I reduce the chance of giving that infection to a neighbor if I wear a mask.

Short Answer: Yes

David Heymann, MD (7/21/2020):

Some data indicate that the level of antibody production may be related to the severity of COVID-19 illness. People with an asymptomatic SARS-CoV-2 infection have been observed to have a weaker antibody response than symptomatic cases and to lose their antibody response after a period of time.^{Long 2020a} The general understanding is that the more serious the signs and symptoms, the more likely it is that antibody will be detected through existing antibody detection systems.

In addition, there is a delay between infection and the production of antibodies. An antibody test performed in this period could be negative despite a positive PCR test.

Short Answer: Yes

David Heymann, MD (7/21/2020):

The WHO recommends that people can be discharged from isolation (regardless of the setting) 10 days after the onset of symptoms plus at least 3 more days if their infection was symptomatic, and 10 days after a positive SARS-CoV-2 test if asymptomatic.^{WHO 2020} According to the WHO advice, these patients do not require retesting. If retesting is conducted, patients should have 2 negative PCR tests at least 24 hours apart to demonstrate clearance of infection.

Short Answer: Yes

David Heymann, MD (7/21/2020):

Pooled testing has been proposed as a way of conserving PCR assays. The process involves batching a number of samples before testing the pooled sample for SARS-CoV-2. If the test is positive, all samples in the batch are retested to identify the positive(s). If the test is negative, then all samples in the batch are negative. This approach is only recommended in areas where there are expected to be a low number of positive results.^{CDC Pooling} Therefore, it is essential to know the incidence of infection locally before employing this strategy.

Short Answer: No

Sharon R. Lewin, AO, FRACP, FAHMS (06/23/2020):

In general, antibody testing is not necessary after PCR-confirmed diagnosis. Once you have a positive PCR test, you have COVID-19.

Short Answer: Still unknown

Sharon R. Lewin, AO, FRACP, FAHMS (6/23/2020):

First of all, it is important to define “immunity.” We know that the immune response to COVID-19 includes generating antibodies; IgG or IgM antibodies are detected in nearly all patients by Day 14-21.^{Long 2020b} However, the key factor for immunity is actually the development of neutralizing antibodies that neutralize and eliminate the virus. There are now several studies showing that nearly everyone infected with SARS-CoV-2 develops neutralizing antibodies, although individual titers vary widely and there is a direct relationship between total antibodies and neutralizing antibodies.^{Robbiani 2020; Wang 2020a} It remains unknown what level of neutralizing antibodies is critical to prevent reinfection or protect from initial infection, which is very important for vaccine design.

In my own experience, the patient’s antibody titer increases progressively and maximizes at Day 21. I have not seen evidence that the titer level actually declines. Large studies evaluating serology would be needed to assess if immunity wanes. That being said, from work on other coronaviruses, especially in animal models, and also from our knowledge of measles, it is not a guarantee that immunity is lifelong.

Short Answer: Early detection and testing

Sharon R. Lewin, AO, FRACP, FAHMS (6/23/2020):

I strongly believe that the key difference was detection and testing early in the course of the outbreak.

In Australia, we were testing and isolating infected people and quarantining all their contacts very aggressively from about mid to late January. We tested anyone with clinical symptoms who had an epidemiologic risk factor: at the time, coming from China or at close contact with a case. We also adjusted those epidemiologic risk factors over time, as we saw outbreaks in Iran, northern Italy, and the United States. Testing rates in Australia in February were extremely high and allowed us to identify and quarantine individuals early in their course of infection.

I have summarized some of the key milestones and timing involved in Australia’s response below, but globally, I think countries that have done well have had high testing rates with early and aggressive testing, effective public health measures of isolation and quarantine (as well as travel bans) in place very early in the course of the outbreak, and physical distancing with good community compliance.

Summary of Australia’s Pandemic Timeline and Response

• Mid-January: The Doherty Institute led by Dr. Mike Catton developed a test for the novel coronavirus and immediately shared the test with the country’s major public health reference laboratories to prepare for broad-scale testing.
• January 24: The first COVID case in Australia was diagnosed (2 days after the first diagnosis was made in the United States), and by this time, we had a test up and running and were at heightened awareness of clinical illness in people returning from China.
• February 1: Australia stopped flights from China to Australia, which I believe was about 24 hours later than the United States.
• By mid-March: Australia’s numbers were doubling every 3 days similar to other developed countries across the globe. However, by this time, we already had a lot of testing in place, so we were diagnosing a lot more people, and therefore, we were aware of who was infected in order to appropriately isolate and quarantine.
• March 29: Australia went into lockdown with physical distancing playing a major role and fairly good compliance in the community. At my work, we have only been working at 30% capacity. Restaurants also remained closed until about mid-June.

It is also worth noting that, in Australia, our government policies have been developed hand-in-hand with scientists. Key scientists, modelers, epidemiologists, and clinicians are all members of the major government committees. Despite the economic hardships, the nation’s response has been very much driven by the science and I am very proud of that.

Short Answer: Probably

Jens D. Lundgren, MD, DMSc (6/9/2020):

In these cases, it is likely that the patient was previously infected. Cross-reactivity with other coronaviruses has been noted with some antibody tests, resulting in a false-positive result, but it is most likely indicative of a previous infection.

Short Answer: No

Jens D. Lundgren, MD, DMSc (6/9/2020):

COVID-19 antibodies are not useful for determining infectiousness. We have also learned that PCR positivity is not necessarily indicative of infectiousness as patients may continue to be PCR positive up to 5 weeks after their symptoms have resolved. It is not known whether this is due to the persistence of live virus because it is difficult to culture and identify live virus. It is also possible that the positive result is due to remnants of bound RNA that are still recognized and amplified by the PCR test.

Data regarding the infectious period of COVID-19 have shown that people with SARS-CoV-2 are infectious before they develop symptoms^{Rothe 2020}; this presymptomatic stage occurs 1-3 days before symptom onset. Some patients remain asymptomatic during infection, which can last 5-7 days, and it is unknown whether they are infectious during this time. However, if patients do develop symptoms, they are considered infectious while they have the symptoms and probably as long as 48 hours after the symptoms are resolved.

For clinical purposes, one can assume a patient is no longer infectious 48 hours after their symptoms have resolved.

Short Answer: It’s complicated

Jens D. Lundgren, MD, DMSc (6/9/2020):

There is much debate on this topic. The World Health Organization had previously announced that asymptomatic carriers rarely transmit SARS-CoV-2 but clarified later that the actual rates of asymptomatic transmission are not yet known. The topic is complicated by our inability to know if someone is asymptomatic or presymptomatic.

We know that in the presymptomatic phase (1-3 days before symptoms develop), if patients later develop symptoms, they were very likely infectious in that presymptomatic stage.^{Rothe 2020} This is particularly important regarding contact tracing as anyone persons would have been in close contact with 1-3 days before they developed symptoms was potentially exposed.

Persons who have detectable virus but never develop symptoms are known as asymptomatic carriers. Although the existence of asymptomatic carriers is clear, little is known about the prevalence, duration of active viral replication, and whether these carriers have a different immunologic response to the infection compared with symptomatic carriers.

It remains unknown how many of the total number of infections worldwide are asymptomatic. It is thought that the prevalence is higher in younger persons compared with older individuals.

Short Answer: No, with exception

Jens D. Lundgren, MD, DMSc (6/9/2020):

In the beginning of the pandemic, a limitation on testing capacity guided whom and when to test. But even in a setting of increased testing capacity, I would argue that, in the absence of ongoing transmission in the local population, routine testing is not necessary except in persons who develop symptoms.

However, in the case of an outbreak, particularly in a hospital or nursing home setting, it is imperative to conduct a proper immunologic investigation and test everyone. This must be done to fully understand who is infected, to make sure sufficient contact tracing occurs, and to ensure the outbreak is handled.

Cases of nosocomial outbreaks affecting both patients and hospital personnel have occurred, typically because there was not a high enough suspicion that a patient had COVID-19 or the patient was in the presymptomatic stage.

Short Answer: It depends

Jens D. Lundgren, MD, DMSc (6/9/2020):

From an epidemiologic point of view, it is interesting and important to understand what proportion of the population has been infected. Also, there are many people who did not get a PCR test while they were sick and, therefore, do not know if they actually had the virus. To satisfy one’s curiosity, this could be answered with serologic testing.

However, it is critical to understand that a positive serology test does not mean patients can now relax and assume they will not get infected again. Whether reinfection is possible is still not known as we have not had a second wave of the virus.

We know that in the short term, there is some degree of immunity/protection as the immune system was able to control the virus’s replication and recover. However, we do not know how long that immunity is maintained or what degree of protection is needed to prevent reinfection.

For this reason, I caution against the use of serologic testing as we do not want people to use the antibody results as a reason to change their behavior.

Everyone should continue to keep physical distance and be careful with hand hygiene. Clinicians also need to continue to use personal protective equipment even if they had a documented infection early on. No one should assume that they are protected. This advice may be different 6 months from now, but for now, this is my best advice.

Short Answer: It’s complicated.

Vikramjit Mukherjee, MD (5/28/2020):

Unfortunately, it is still lower than we would prefer. When respiratory tract samples collected from patients hospitalized with COVID-19 in China were used to detect SARS-CoV-2 by RT-PCR, investigators found that nasopharyngeal swabs have a sensitivity of approximately 60%, tracheal aspirates approximately 70%, and bronchoalveolar lavage in the 90% to 95% range.^{Wang 2020b; Yang 2020} This is where your clinical suspicion as a healthcare provider and how badly your community is affected comes into play.

We probably faltered in the beginning of our experience here—you should not be making big changes in de-escalating isolation based on a negative test. You must consider the clinical picture, including imaging, and how badly your community is affected. Testing on nasopharyngeal or oropharyngeal samples complements these other diagnostic criteria and should not be taken in isolation.^{Woloshin 2020}

Short Answer: It’s complicated.

Vikramjit Mukherjee, MD (5/28/2020):

Most viruses have an incubation period of 7-14 days. COVID-19 has an earlier onset, between 2 and 14 days from exposure to symptoms.

In most places in New York, we are testing patients at 21 days after their first positive test to see if they are still PCR positive by nasopharyngeal or oropharyngeal swab. If the patient is positive, we do not know if the PCR detected viral RNA from dead virus particles in the patient’s respiratory tract or if the patient still has active viral shedding despite recovery from COVID-19.

The only way of distinguishing these possibilities would be to do viral cultures in a lab with a biosafety level 3 or higher, as dead virus will not grow in culture. This research is ongoing.

Short Answer: Treat as if it is COVID-19.

Vikramjit Mukherjee, MD (5/28/2020):

This goes back to the sensitivity of RT-PCR testing on nasopharyngeal and the oropharyngeal swabs. There is a lot that can go wrong in the testing process that can lead to false-negative results, for example, poor sample collection and low viral load.^{Li 2020b}

If your community is being hit hard by COVID-19 and you have a patient who has a fever and a cough and either a gastrointestinal illness or a respiratory illness, that patient should be treated as a patient with COVID-19 until proven otherwise and isolation precautions should be followed. A negative PCR test on a swab means nothing at this point.

Short Answer: Yes

Vikramjit Mukherjee, MD (5/28/2020):

This question is important for clinicians and for our public health colleagues where source identification and isolation in a densely populated city like New York is such a challenge.

Data from Europe and from the United States show that asymptomatic/presymptomatic transmission occurs, meaning that transmission from human to human happens before the index patient has a fever or a cough.^{Rothe 2020} Of course, this is similar to other coronaviruses where virus shedding also happens before a patient becomes febrile or dyspneic (short of breath).^{Gandhi 2020}

Short Answer: Yes

Vikramjit Mukherjee, MD (5/28/2020):

My perspective is informed by New York, and as most of you know, New York is a densely populated place with unique aspects, such as the overcrowded public transport system and a large immigrant population with poor access to conventional healthcare. These factors make it more prone to another wave of COVID-19.

Antibody testing across the city shows that approximately 20% of the population has been infected and has recovered. It is estimated that to achieve herd immunity to COVID-19, 70% to 80% of the population will need to be infected and recover.^{Randolph 2020; Kwok 2020} However, I hope that we do not have a second wave—we have seen the ugliness of this disease, and so we are preparing and are very cognizant of the fact that as the social distancing parameters are relaxed and people re-engage in business activities, eat at restaurants, and resume social activities, the virus will likely spread again.