Updated June 24, 2020

We are very aware that many patients depend on UCSF Health for specialty care that cannot be found at other hospitals in this region. We are actively working to make sure we can continue to provide that care, while also doing our part in responding to the COVID-19 crisis.

Our clinics are offering appointments and services in person or via telehealth depending on the nature of the appointment or service. For updates on referrals by specific specialty areas, please visit our Referrals By Specialty page.

If you would like to refer a patient to UCSF, please contact the appropriate clinic directly.

For additional questions or assistance, please call the Physician Liaison Service at 800-444-2559.

Updated March 27, 2020

Please contact the appropriate clinic directly. For additional questions or assistance, please call the Physician Liaison Service at 800-444-2559.

Updated March 27, 2020

Video visits are being offered throughout our health system. Each UCSF provider will review their schedules to make the decision about when the patient needs to be seen. Patients may be scheduled in person, if urgent, or by a video or telephone visit.

Updated March 27, 2020

The process for referring patients with suspected respiratory illness to the adult emergency department (ED) is as follows:

1. Call the ED at 415-353-1238 and tell the clerk you are sending a patient.

a. Be ready to give demographic info on the patient.

b. Provide your name and the best callback number.

c. Let the clerk know if you need to speak to the ED attending right away about this patient.

2. If you are on EPIC EHR, please leave a note summarizing the reason for the ED transfer, along with your goals and callback reasons, and sign the note so it’s visible.

3. If you are not on EPIC, wait to speak to an ED provider. We have additional staff answering phones at peak hours to lessen your wait time.

For any patients that require hospitalization but do not need urgent evaluation or stabilization, please follow the usual processes to transfer or directly admit a patient.

Updated March 27, 2020

Teams are working to ensure that the campus and UCSF Health are prepared and that faculty, staff, students, patients and visitors are informed about the outbreak.

The health and safety of our patients and visitors, and our UCSF community, is our most important priority. UCSF remains focused on three primary goals:

• providing compassionate and safe care for patients with COVID-19,
• providing support and training to keep health care workers and other employees safe while caring for these patients, and
• communicating regularly with our community during this fluid situation.

UCSF Health has a proven track record of protecting and safely caring for patients with infectious diseases and severe illnesses and was among the first hospitals to start preparing for a large influx of patients. As a result, UCSF Health has the facilities and protocols in place to help us to care for these patients, while protecting the health of our providers and staff.

UCSF Health is expanding the availability of airborne-infection isolation rooms, which can safely isolate patients with COVID-19. We will adapt additional rooms and hospital areas to care for larger numbers of patients if needed.

In addition, UCSF Health’s Hospital Incident Command System team is working on surge planning for a broader community spread. To this end, UCSF Health has erected structures known as accelerated care units outside its emergency departments at its Parnassus site, as well as outside the two UCSF Benioff Children’s Hospitals at Mission Bay and Oakland. The structures have negative-pressure air flow and are being used to triage patients with respiratory illnesses, to help provide appropriate care for those patients while keeping our other emergency department patients and staff safe.

UCSF Health is also working to increase inpatient capacity by adding additional acute care and ICU beds at its Mount Zion medical campus, and is working with hospitals across the City of San Francisco to open a new COVID-19 facility at Saint Francis Memorial Hospital to help the city prepare for a surge in patients.

UCSF Health has been administering COVID-19 tests to inpatients at our hospitals, to patients with symptoms entering through our Emergency Department, and to outpatients at the UCSF Respiratory Screening Clinic since March 9. UCSF Health continues to work on strategies for even more rapid testing turnaround. See UCSF Health’s statement about current efforts to increase coronavirus testing capacity.

Updated March 27, 2020

UCSF Health cares for patients with some of the most complex health conditions and is highly experienced in infection prevention and control. Its infection prevention practices and protocols are aimed at ensuring that patients and visitors remain safe in the hospital.

UCSF Health has put into place Centers for Disease Control and Prevention (CDC)-recommended infection prevention protocols and procedures to prevent spread of infection from patients with COVID-19.

Specialized structures, known as accelerated care units (ACUs), have been erected outside our hospitals to screen respiratory patients who may have been exposed to COVID-19. These ACUs use specialized air-pressure technology to enhance our infection control and prevention protocols and are fully stocked with medical supplies and diagnostic testing equipment.

If hospitalization for COVID-19 is required, UCSF Health has airborne-infection isolation rooms and can adapt additional rooms to care for a larger number of patients, if needed.

Updated March 27, 2020

UCSF Health has been actively working to expand on-site testing capacity for patients who are suspected of having COVID-19. We do not have the capacity to test the general community at this time.

However, we expect to continue to expand our testing capabilities in the coming days and weeks to include more outpatients with COVID-19 symptoms.

In addition, our researchers are actively working to develop tests that can provide results more quickly. See our full statement on testing capabilities.

Updated March 27, 2020

Due to limited testing capacity, most health systems continue to prioritize COVID-19 testing resources for hospitalized patients and vulnerable, at-risk populations. This strategy will improve the health outcomes of individuals with a higher risk of becoming seriously ill with COVID-19. As testing capacity expands, it will become more commercially available. It is already available by physician referral from some clinical labs, including Quest Diagnostics.

Updated May 1, 2020

There are two main types of tests for COVID-19. Reverse transcriptase polymerase chain reaction (RT-PCR) directly detects active viral infection. Serology tests detect immune response (antibodies) to infection, which typically takes at least two to three weeks to develop.

Updated May 1, 2020

In addition to providing COVID-19 diagnostic and serologic testing services, UCSF has initiated a number of new research studies. Teams are assessing the performance of commercially available serologic tests, as well as the performance and quality of available diagnostic PCR tests so that health systems and governments are aware which are the most effective testing platforms. 

Updated March 30, 2020

Yes, if the logistics allow for it. Physicians may decide to alter management for patients with a positive PCR test for COVID-19 undergoing these procedures. We have described in prior digests that patients can have high viral loads detected on PCR prior to exhibiting symptoms. It is important to note that the current tests are only 70%-80% sensitive – so using current assays, one is reducing but not eliminating the possibility that the patient has COVID-19 underlying disease. In other words, a positive test confirms disease, but a negative test does not completely rule it out.

Updated March 31, 2020

Yes, this makes sense. Patients in a skilled nursing facility are often older and have multiple comorbidities; therefore, they are at high risk for severe COVID-19. While symptomatic patients will likely be identified, preventing the introduction to these settings of infected patients who are asymptomatic or presymptomatic is also important to prevent spread of this disease, which can become rampant (as noted in the details of the well-publicized outbreak in a Washington state long-term care facility). This approach is not fail-safe, however; false negatives can occur. Therefore, isolating patients for a period of time upon transfer to chronic care facilities should be considered as an alternative or adjunctive strategy.

Updated April 21, 2020

The negative predictive value [(true negatives)/(true negatives + false negatives)] allows us to understand the significance of a negative test, which differs depending on the prevalence of disease in the population being tested. In asymptomatic patients, the prevalence of SARS-CoV-2 in the Bay Area and in the U.S. is not yet known. If we assume that the prevalence in the Bay Area is 1% and the sensitivity/specificity of a nasopharyngeal (NP) swab test is estimated at 75%/98%, then the negative predictive value of the test is 99.7%. In symptomatic patients or those with known exposures, the prevalence rate may be 10% or higher. In this case, the negative predictive value would be lower, at 97.2%.

Conclusion: When interpreting a negative test, one must take into account both the sensitivity of the test and the prevalence of disease.

Updated May 1, 2020

Serology blood tests are designed to identify who has previously been infected with COVID-19.  They look for an immune response to the coronavirus in the form of a specific protein material, or antibodies, in your blood. It takes time for the body to produce antibodies, so antibodies may not be detected early on after infection. Not all patients may develop detectable antibodies, particularly patients who have weakened immune systems due to underlying diseases or certain medications.

Serology tests for COVID-19 are mainly used for diagnostic testing of patients whose symptoms indicate a high suspicion of COVID-19, but who are seeking care more than a week after symptoms began and who do not test positive for COVID-19 RT-PCR test. The tests also are useful for people who want to donate convalescent serum and, once they are more reliable, can be used to determine how broadly the virus has spread in the community in the past.

Updated May 1, 2020

There continues to be significant variability in commercial serologic tests, but we hope to see more accurate tests come online soon. In the meantime, testing of asymptomatic individuals with low suspicion for prior infection should be approached with caution until we understand more about the potential for false positive and false negative results.

Updated May 1, 2020

SARS-CoV-2 serology tests are now available for UCSF patients with a provider order. Recommended uses for this test are described above. Results should be back within 24 hours from the time blood is drawn for the test.

Updated May 1, 2020

A positive test shows that you have antibodies that may have resulted from COVID-19 infection. It could also be a false positive test from a cross reaction with antibodies from a related coronavirus or other virus (there are multiple other common coronaviruses that cause mild upper respiratory infections). We currently don’t know whether antibodies to the SARS-CoV-2 virus offer immunity against future infections and if so, for how long. Several other coronaviruses have limited or short-term immunity after an infection, but it is too early in this epidemic to understand whether or how antibodies will protect people who have had COVID-19. We will need more studies to determine if the presence of antibodies means a person won’t get re-infected. Until those studies are done, decisions about returning individuals to work or school should not be based on antibody test results.

Updated May 1, 2020

Antibodies can appear approximately one week after infection and typically peak four to six weeks after symptom begin.  Most people will develop antibodies within two to three weeks after infection. Some patients, particularly those with immunocompromising conditions, may not produce a detectible antibody response.

The length of time that antibodies remain in the body after infection is not well known for SARS-CoV-2.  For most viruses, antibodies remain detectable for months or sometimes even years after infection.

Updated May 1, 2020

Not necessarily. Antibodies may become detectable before infectious virus is cleared by the body, so a positive antibody test does not mean you cannot spread the virus to other people.

Updated May 1, 2020

A negative test means you probably have not had a previous infection with COVID-19 that has since resolved. However, you could still have a current infection if antibodies have not formed yet (it takes 1 to 3 weeks to develop antibodies), or you could have had a previous infection but not developed an antibody response (particularly if you have a weak immune system from underlying medical conditions or medications).

Updated April 10, 2020

The following treatments are available through clinical trials, expanded access programs (EAPs) or compassionate use for intensive care unit (ICU) patients at UCSF Health and Zuckerberg San Francisco General Hospital and Trauma Center (ZSFG): (1) remdesivir (which blocks viral replication), (2) mesenchymal stromal cells (for ARDS) and (3) convalescent COVID-19 plasma (coming soon). We will be reporting on additional studies as they become available.

Updated July 21, 2020

A few publications have described clinical features in patients with acute respiratory distress syndrome (ARDS) from COVID-19 that appear to have physiologic differences from classical ARDS, particularly in terms of a phenotype of severe hypoxemia but minimal alterations in respiratory compliance. This resulted in controversial recommendations to change ventilator management strategies in patients with ARDS from COVID-19. However, a recent scholarly review of reports from several major centers in the U.S., China and Europe indicates that, on average, the respiratory compliance (tidal volume/plateau airway pressure-the level of PEEP) and the severity of hypoxemia (Pa0₂/Fi0₂) in COVID-19 ARDS patients is similar to classical ARDS (Fan et al.). While there is heterogeneity around these mean values, on balance the differences from classical ARDS are minimal. Thus, the authors recommend adherence to an evidence-based approach to ventilatory management of COVID-19 ARDS, including lung protective ventilation with a tidal volume of 6 mL/kg predicted body weight, and a plateau airway pressure of <30 cm H₂O (Matthay et al.). PEEP should be individualized to maintain oxygen delivery and reduce the risk of ventilator-associated lung injury. As in classical ARDS, prone positioning is recommended in mechanically ventilated COVID-19 patients with moderate to severe ARDS ((Pa0₂/Fi0₂) <150 mm Hg) (Matthay et al.).

Conclusion: Intensivists should approach ventilatory management the same way in COVID-19-associated ARDS as with ARDS in other conditions.

Updated March 31, 2020

A case series of five mechanically ventilated patients in Wuhan, China with persistently positive COVID-19 PCR who were treated with large-volume infusions of convalescent sera (passive antibody therapy) reported clinical improvement (Shen et al.). The donor plasma had IgG and IgM antibodies to the COVID-19 spike protein and neutralized the virus in the cell culture infection model. The treated patients demonstrated significant clinical improvement after 1 week, which included an increase in neutralizing antibody titers, and PCR became negative 1-12 days after transfusion. Possible benefit of convalescent plasma was confounded by concurrent use of antivirals and steroids. An additional study reports treatment of 19 patients with COVID-19 with convalescent sera and subsequent clinical improvement (Duan et al.). Optimal dosage, timing of administration and whether hyperimmune or convalescent sera may be more effective remain to be determined. Based on these preliminary data and past use of convalescent sera for other viral epidemics, the Food and Drug Administration (FDA) approved emergent use of this therapy on March 24, 2020 (FDA). The National COVID-19 Convalescent Plasma Project is planning trials across the country.

Updated April 10, 2020

Extracorporeal membrane oxygenation (ECMO) has been used as a rescue therapy for severe acute respiratory distress syndrome (ARDS); however, its role in COVID-19 remains unclear (Hong et al.). Data on its use with respiratory viral infections is encouraging. In the H1N1 influenza epidemic, ECMO reduced hospital mortality rate (23.7% ECMO-referred vs. 52.5% non-ECMO-referred), and a retrospective study of 35 patients with severe Middle East Respiratory Syndrome (MERS) showed decreased in-hospital mortality (65% with ECMO vs. 100% without) (Alshahrani et al.). To date, ECMO has been used infrequently for the care of COVID-19 patients in intensive care units (only 1% in the largest study from Italy) (Grasselli et al.), and outcomes are largely unknown.

Interim guidance: the WHO and ELSO (Extracorporeal Life Support Organization) recommend administering venovenous ECMO to eligible patients not improving by conventional methods, at experienced ECMO centers.

Updated April 7, 2020

A hypercoagulable state is well described in patients with pneumonia and sepsis (Levi and van der Poll). Elevated D-dimer and elevated IL-6 (mediator of cytokine-induced coagulation) are correlated with poor outcomes in COVID-19 (Zhou et al.). Among 21/183 non-survivors hospitalized with COVID-19 pneumonia in China, 71% met criteria for disseminated intravascular coagulation (Tang et al.). In a retrospective study of severe COVID-19 cases with coagulopathy, heparin was not associated with a benefit in reduction of 28-day mortality except in a subset of patients with very high D-dimer levels (Tang et al.). The bottom line: More data are needed to inform these clinical decisions. In the meantime, in light of patient isolation and limited mobility, we agree with the American Society of Hematology (ASH) recommendation that “all hospitalized patients with COVID-19 should receive pharmacologic thromboprophylaxis with LMWH [low-molecular-weight heparin] or fondaparinux (suggested over unfractionated heparin to reduce contact) unless the patient is judged to be at increased bleeding risk.”

Updated March 27, 2020

It is not known if infection with SARS-CoV-2 results in long-lasting immunity. However, a recent study in rhesus monkeys suggests that at least short-term immunity is likely (Bao et al.). Researchers infected rhesus monkeys with SARS-CoV-2 and documented clinical disease and viral shedding. Twenty-eight days after primary infection, monkeys were re-exposed to the virus. None of the re-exposed monkeys demonstrated clinical disease nor were found to have viral shedding. However, studies of common coronaviruses that cause upper respiratory tract infection in humans have been shown to result in only partial, short-term immunity. Whether natural infection with SARS-CoV-2 will result in short-term or durable immunity in humans remains unknown and an area for further research.

Updated April 14, 2020

Discontinuation of isolation precautions may be done using clinical criteria or a test-based strategy, depending on the clinical scenario and test availability. At UCSF and Zuckerberg San Francisco General Hospital and Trauma Center (ZSFG), hospitalized patients can come off precautions when they have met all of the following milestones: (1) ≥14 days from symptom onset, (2) ≥72 hours fever-free without antipyretics, (3) improving respiratory symptoms (e.g., cough, dyspnea) and (4) two consecutive negative RNA swabs (nasopharyngeal/mid-nasal turbinate plus oropharyngeal) collected ≥24 hours apart. For patients returning home, we recommend following the CDC criteria for discontinuation of home isolation. All of the following criteria must be met: (1) ≥72 hours fever-free without antipyretics, (2) improving respiratory symptoms (e.g., cough, dyspnea) and (3) ≥7 days since the first symptoms appeared. For patients with immunosuppression or who have protracted symptoms, a test-based approach should be considered if testing supplies are available.

Updated March 27, 2020

SARS-CoV-2 utilizes the angiotensin-converting enzyme 2 (ACE2) receptor for viral cell entry, the expression of which is increased when taking renin-angiotensin-aldosterone system (RAAS) inhibitors (ACE/ARB blockers that block the ACE1 receptor). The clinical significance of ACE2 modulation in human SARS-CoV-2 infection is unknown. RAAS inhibition has been argued to pose a theoretical increased risk of a higher susceptibility to infection due to increased ACE2 expression (Kuster et al.). On the other hand, animal models suggest that increased ACE2 expression is protective from severe acute lung injury induced by aspiration, sepsis and SARS-CoV infection (Imai et al., Kuba et al.). Currently, there is no data proving a causal relationship between ACE2 activity and human SARS-CoV-2-associated mortality.

We do not recommend stopping or starting a RAAS inhibitor because of SARS-CoV-2 infection. There are interventional and observational studies planned to address this association.

Updated May 1, 2020

Convalescent serum is the cell-free part of blood containing antibodies that is taken from patients who have recovered from a certain illness and then delivered to patients with the same illness to help their immune response. It is a treatment that has been used for many infections over the years, but its effectiveness still needs to be studied for patients with COVID-19.

Updated May 1, 2020

Hydroxychloroquine is in a class of drugs that are primarily used to treat malaria, but also are used to treat discoid or systemic lupus erythematosus and rheumatoid arthritis in patients whose symptoms have not improved with other treatments.

Updated May 1, 2020

We currently don’t know. There have been a few small studies that have not answered key questions and have shown evidence of harm by using chloroquine and hydroxychloroquine in patients with COVID-19. This demonstrates the importance of not using these drugs off-label and instead testing them in clinical trial settings where we can monitor patients.

Updated May 1, 2020

Remdesivir is an investigational antiviral compound undergoing clinical trials in China, the United States, and the United Kingdom as a potential treatment for COVID-19, including at UCSF Health. It is not yet licensed or approved anywhere globally.

Updated May 1, 2020

Published studies to date have not demonstrated the benefit of using remdesivir for the treatment of COVID-19.  However, a recent press release from the National Institutes of Allergy and Infectious Diseases reported improved clinical outcomes in an interim analysis of a large randomized control trial in patients taking remdesivir compared to those taking placebo.  Publication of this study through a process of peer review is pending as are several other studies using remdesivir.

Updated May 1, 2020

UCSF is leading clinical trials of hydroxychloroquine and the antibiotic azithromycin, and of remdesivir.

Updated June 24, 2020

In a preliminary DSMB analysis, the multi-arm UK RECOVERY trial found no difference in 28-day mortality in 1,542 COVID-19 patients randomized to hydroxychloroquine vs. 3,132 randomized to standard of care; 25.7% vs. 23.5% (HR 1.1 [95% CI, 0.98-1.26], p=0.1). There was no evidence of a beneficial effect of hydroxychloroquine on length of hospital stay or other outcomes.

Conclusion: These results continue to build on the body of literature suggesting that hydroxychloroquine may not be effective for prevention or treatment of COVID-19.

Updated May 1, 2020

In response to the global COVID-19 pandemic, the global vaccine research and development effort has moved at an unprecedented speed. Based on the scale and speed at which vaccine development is proceeding, there is an indication that a vaccine could be available at some time in 2021 under emergency use protocols.

Updated July 21, 2020

The San Francisco Department of Public Health has a detailed health directive on best practices for dental health care. Elements include posting signage, social distancing, symptom screening for patients and personnel, screening for close exposures within 14 days, face coverings for all (including patients when not receiving care) and hand hygiene protocols. The directive is to minimize aerosol-generating procedures (AGPs); and polymerase chain reaction (PCR) testing within 7 days of a nonemergent AGP is “strongly recommended.” For AGPs, four-handed dentistry, high-evacuation suction and dental dams are recommended. CDC guidance additionally recommends that personnel wear surgical masks plus eye protection plus gloves plus gowns if there may be splashing or splattering, with N95 or higher-level respirators used for AGPs in those assumed to be noninfectious. Only emergent procedures should be performed if the patient has tested positive for COVID-19.

Updated April 3, 2020

The CDC now recommends “wearing cloth face coverings in public settings where other social distancing measures are difficult to maintain (e.g., grocery stores and pharmacies) especially in areas of significant community-based transmission” (CDC). We support the use of face masks or face coverings in addition to present guidance on social distancing and hand hygiene for three reasons: the highly contagious nature of the virus, the potential for asymptomatic transmission, and empiric evidence from Asia, where masks are routinely used. Supplies of hospital-grade masks for health care workers must be prioritized, but excess surgical masks or homemade masks of multilayered cotton likely provide more protection compared to nothing during the epidemic. An increasing number of cities in the U.S. are embracing this recommendation as an adjunct measure: Stay in place, keep your space and cover your face.

Updated April 10, 2020

Given the present shortage of personal protective equipment (PPE), strategies for extended use (wearing PPE continuously without doffing it between multiple patient interactions) and reuse (doffing PPE and storing it in a clean, dry place such as a paper bag labeled with the user’s name) have been developed.

Regularly updated guidelines on extended use and reuse of PPE at UCSF Medical Center can be found here.

Extended wearing of masks and face shields in areas where COVID-19 patients are cohorted is easiest to implement. N95 masks can be reused if contaminated hands have not touched the inside of the mask, the mask is not wet/soiled/damaged, and appropriate fit is maintained (a fit check should be performed each time the N95 is donned). If a face shield is worn over the N95 mask during aerosol-generating procedures, reuse is also acceptable. If masks become contaminated, they must be discarded. Some institutions are decontaminating N95 respirators using UV germicidal irradiation, vaporous hydrogen peroxide or moist heat (Liao et al.), and UCSF Health is investigating these options. Extended use and reuse guidance is also available for goggles and face shields.

Updated July 21, 2020

Our present understanding of the prevalence of SARS-CoV-2 infection in Southern California is based mostly on polymerase chain reaction (PCR) testing. Antibody testing is helpful in that it can detect people with prior asymptomatic or milder infection who may not have been tested for the virus itself. In order to better understand the seroprevalence of SARS-CoV-2 in Los Angeles County, researchers conducted a study offering antibody (IgM/IgG) testing to a random sample of patients at six study sites as well as in-home testing (Sood et al.). Of 1,952 people who were invited to participate in antibody testing, 1,702 provided consent and 865 of these (50.9%) were tested. Thirty-five individuals tested positive, resulting in an overall adjusted seroprevalence of 4.65%. Among different racial/ethnic groups, the prevalences were Black, 6.94%; white, 4.42%; and Hispanic, 2.10%. The study respondents were disproportionately white, affluent and female in this investigation; the authors attempted to adjust for differing characteristics of study participants versus the general Los Angeles County population.

Conclusion: This study suggests that a much higher number of COVID-19 cases have occurred in Los Angeles County than had been reported (367,000 versus 8,430 as of April 10, 2020). Ongoing surveillance is needed to understand the impact of COVID-19 throughout the state and in particular among vulnerable populations.

Updated July 21, 2020

Debate continues around the extent to which SARS-CoV-2, the virus that causes COVID-19, is transmitted via respiratory droplets versus aerosols, highlighted in several recent commentaries and a scientific brief released by the World Health Organization (Morawska and Milton, Klompas et al., WHO). Although experimental data that show that speaking and coughing can produce a mixture of both, the balance of evidence to date continues to suggest droplet transmission as the main mode of spread. An exception may be prolonged exposure to an infected person in a poorly ventilated space that allows otherwise insignificant amounts of virus-laden aerosols to accumulate. This risk is minimized in health care facilities by ventilation standards. WHO continues to recommend droplet plus contact precautions (outside of aerosol-generating procedures) for suspected and confirmed COVID-19 patients as part of a comprehensive strategy that includes universal masking and physical distancing.

Conclusion: Droplets are the primary mode of transmission of COVID-19; in health care settings, outside of aerosol-generating procedures, the impact of aerosols on the spread of COVID-19 is likely minimal.

Updated July 21, 2020

A recent article reported results of a telephone survey in 204 patients with COVID-19. Patients reported a reduction in taste (in 55%) and smell (in 42%) at a median of 4 days before diagnosis; 40% had both, and it was severe in 35%-40% (Mercante et al.). Only 15%-17% of patients with severe reduction in taste/smell reported severe nasal obstruction, suggesting that the pathophysiology may be a direct effect of the virus on the olfactory nerve rather than solely due to nasal inflammation and obstruction. Overall, this study adds to a growing body of (mostly retrospective) studies showing that disorders of taste and smell are very common in COVID-19, ranging from 34% to 89% depending on the study (Giacomelli et al., Lechien et al.).

Conclusion: Disorders of taste and smell are common presenting symptoms of COVID-19 and can occur without nasal congestion.

Updated July 21, 2020

COVID-19-associated multisystem inflammatory syndrome has been previously reported in children (MIS-C). CDC and WHO case definitions include only patients aged less than 21 years and 19 years, respectively. Two recent case reports, both from New York, describe a similar clinical syndrome occurring in adults with laboratory-confirmed (PCR) SARS-CoV-2 infection (Shaigany et al., Sokolovsky et al.). Both patients were previously healthy adults in their 30s and 40s presenting with fevers and multisystem features of Kawasaki disease, including nonexudative conjunctivitis, cracked lips, diffuse rash and cervical lymphadenopathy. Both patients presented with prominent gastrointestinal symptoms but lacked significant respiratory symptoms.

Conclusion: These reports suggest that the multisystem inflammatory syndrome previously described in children may also occur rarely in adults. Future research is needed to further delineate the pathophysiology and epidemiology of these conditions and improve current approaches to diagnosis and treatment.

Updated July 21, 2020

Based on the current literature, most neonates born to mothers with SARS-CoV-2 infection do not develop symptoms/sequelae (Prabhu et al.). A limited number of reports demonstrate that neonates can become infected and develop symptoms. For example, symptomatic early-onset SARS-CoV-2 infection has been described in a case series of 33 neonates born to women with COVID-19; three of these infants were infected with SARS-CoV-2 (Zeng et al.). The symptoms of the infected neonates included fever, lethargy, vomiting, pneumonia, leukocytosis, lymphopenia and thrombocytopenia. Another recent case report of a late preterm baby born to a symptomatic COVID-19-positive mother was consistent with vertical transmission given positive COVID-19 polymerase chain reaction (PCR) test results from the placenta/amniotic fluid/vaginal fluid and from the baby’s blood/nasopharyngeal (NP) swab; this baby developed neurologic symptoms on day 3 of life and had central nervous system radiographic abnormalities (Vivanti et al.). There have also been reports of neonates/young infants presenting with late-onset sepsis, fever/hypothermia, lethargy, feeding difficulty, pneumonia/respiratory failure, hypoglycemia, leukopenia, neutropenia and thrombocytosis (Feld et al.).

Conclusion: Although uncommon, neonates can develop symptomatic COVID-19. More studies are needed to further establish how common COVID-19 is in neonates and what symptoms are most common.

Updated July 21, 2020

A study in New York revealed significant disparities in hospitalization and death rates across the city’s five boroughs, with the highest rates in Queens and the Bronx (Wadhera et al.). And a cross-sectional study of pregnant women investigated associations between the “built environment” (large households, household crowding and low socioeconomic status) and neighborhood among pregnant women delivering at NewYork-Presbyterian hospitals after implementation of a universal SARS-CoV-2 polymerase chain reaction (PCR) testing campaign (Emeruwa et al.). Of 396 women tested, 71 (17.9%) were infected with SARS-CoV-2. Odds of infection (all odds ratios are presented as interdecile) were lower among women living in buildings with more residential units (OR, 0.34 [95% CI, 0.16-0.72]) and higher appraised values (OR, 0.29 [95% CI, 0.10-0.89]) and in neighborhoods with higher median incomes (OR, 0.32 [95% CI, 0.12-0.83]). Odds of infection were higher among women residing in neighborhoods with higher unemployment rates (OR, 2.13 [95% CI, 1.18-3.83]) and in dwellings with large household membership (OR, 3.16 [95% CI, 1.58-6.37]) and greater household crowding (OR, 2.27 [95% CI, 1.12-4.61]). Limitations of the study include small sample size and specialized patient population (pregnant women).

Conclusion: This study provides empirical evidence for the hypothesis that variation in the urban environment, including overcrowding and poverty, are important social determinants of SARS-CoV-2 transmission. Support for those unable to socially distance under these conditions will be important for curbing the COVID-19 pandemic in the U.S.

Updated June 24, 2020

A recent international multicenter cohort study examined 30-day mortality and rate of pulmonary complications in a cohort of 1,128 patients diagnosed with COVID-19 in the 7 days before (26.1%) or 30 days after (71.5%) undergoing surgery (Archer et al.). Most (74%) operations were emergent, and 74.6% were classified as major. Thirty-day mortality was 23.8%, and pulmonary complications occurred in 51.2%, accounting for 82.6% of deaths. Risk factors for mortality included male sex, age ³70, malignancy and emergent or major surgery. Limitations of this study are as follows: (1) there was no control group; (2) lack of universal testing may have biased the cohort toward those with more severe or symptomatic COVID-19; (3) many centers limited lower-risk/elective surgeries during this time period, biasing the cohort toward higher risks of morbidity and mortality.

Conclusion: Patients with perioperative COVID-19 appear to have a higher risk of poor outcomes than is generally reported for most surgeries. Decisions around delaying surgery must account for this risk but should also balance risks of progression of underlying disease. Studies of outcomes in patients with asymptomatic infections undergoing surgery will be informative.

Updated June 24, 2020

Early data from China suggested that co-infection with other respiratory pathogens was rare in patients with COVID-19 (Chen et al.). A recent systematic review and meta-analysis (preproof, not peer reviewed) of 30 studies including 3,834 patients with COVID-19 demonstrated that bacterial co-infection was present in 7% of hospitalized patients and that rates were higher in the ICU compared to the non-ICU ward setting (14% vs. 4%) (Lansbury et al.). The pooled proportion of viral co-infections in this study was 3%, with respiratory syncytial virus (RSV) and influenza A being the most common pathogens. Another recent study examined nasopharyngeal specimens from 1,206 patients in Northern California and found that among 116 laboratory-confirmed cases of COVID-19, viral co-infection was present in 24 (20.7%) specimens (Kim et al.). RSV and rhinovirus/enterovirus were the most common co-pathogens.

Conclusion: Rates of bacterial co-infection in patients with COVID-19 are overall low, which has important implications for antibiotic stewardship in the hospital. Rates of viral co-infection differ widely between studies, which likely reflects differences in sample size as well as seasonal and geographic trends.

Updated June 24, 2020

A recent Spanish study surveyed a range of neurologic complications among 841 inpatients with COVID-19 (Romero-Sánchez et al.). In this cohort, 57.4% were found to have some neurologic symptom or disorder, though nonspecific problems such as myalgias were also counted. Of greater concern, 1.3% of participants suffered an ischemic stroke, and 0.4% developed intracranial hemorrhage. Approximately 5% developed taste disturbance or loss of smell, which has been described previously. Other complications such as seizure or encephalitis were diagnosed in <1% of patients. In 4.1% of cases, death was attributed to neurologic complications.

Conclusion: Persons with COVID-19 may experience a range of neurologic complications such as stroke. Determining which are directly related to COVID-19 as opposed to general physiologic derangements, concomitant medications or critical illness can be challenging. Prospective registries (Román et al.) will provide greater clarity.

Updated June 24, 2020

Cancer patients appear to be at increased risk of COVID-19 as well as severe complications. Earlier reports from China demonstrated an increased prevalence of malignancy among COVID-19 patients (Liang et al.) and COVID-19 among cancer patients (Yu et al.). Subsequent larger studies from China, the United States (including New York City) and the United Kingdom have also found that patients with cancer and COVID-19 are at increased risk for severe disease (requiring intensive care or mechanical ventilation) and death (Yang et al., Kuderer et al., Mehta et al., Lee et al.), with reported case fatality rates ranging from 20% to 29% (Tian et al., Zhang et al.). Among this population, the highest risk for severe disease and mortality included patients with other known risk factors for COVID-19 complications (older age, male gender, hypertension, cardiovascular disease), advanced malignancy, recent chemotherapy and hematologic malignancy (Yang et al.).

Conclusion: Patients with cancer and COVID-19 are at increased risk for severe disease. Further studies are needed to determine whether behavioral (frequent health care exposure) or clinical (myelosuppression/immunosuppression) risk factors are most associated with these poor outcomes.

Updated April 21, 2020

We are unsure. Cutaneous findings were rarely reported (<1%) in large studies from China (Guan et al.). A recent study from Italy found that 18 of 88 (20%) hospitalized patients had skin findings: erythematous rash (78%), diffuse urticaria (17%) and vesicles resembling varicella (5%) (Recalcati). Another report from Italy described 22 patients with papulovesicular eruption resembling varicella (Marzano et al.). In both reports, the trunk was most commonly involved, and itching was uncommon. Individual case reports of patients with COVID-19 and a diffuse erythematous rash (Hunt and Koziatek), diffuse urticaria (Henry et al.), petechial rash (Joob and Wiwanitkit) and violaceous lesions in the toes (Mazzotta and Troccoli) have been described as well. The American Academy of Dermatology has launched a COVID-19 dermatology registry to better understand the cutaneous manifestations of COVID-19.

Conclusion: Cutaneous findings with COVID-19 seem uncommon, and work is underway to better characterize dermatologic manifestations.

Updated June 24, 2020

Researchers in Spain have now completed the largest prospective cohort study to date of 51 consecutive HIV-positive patients admitted for COVID-19 to a single hospital in Madrid (Vizcarra et al.). Compared with a cohort of 1,288 people living with HIV (PWH) without COVID-19, those who were diagnosed with COVID-19 were more likely to have a higher body mass index (BMI) (25.5 vs. 23.7 kg/m², p<0.05); to have hypertension, diabetes, chronic kidney disease and chronic liver disease; and to have been on tenofovir-based antiretroviral therapy (ART) prior to their COVID-19 diagnosis, although the majority of patients were on tenofovir. There were no differences in clinical features, laboratory abnormalities and radiographic changes among PWH and those without HIV from other series. PWH with CD4 counts <200 cells/mm³ did not exhibit differences in clinical outcomes compared to those with higher CD4 counts. ART regimen did not change the risk of mild, moderate or severe COVID-19.

Conclusion: Most data to date suggest persons with HIV are not at increased risk of COVID-19 susceptibility or severe disease; however, the studies are small like this one. Whether any ART regimen protects from COVID-19 is still an open question.

Updated March 30, 2020

We do not yet know the answer for the U.S. We do have some data from other countries. Among 72,314 cases in mainland China (through 2/11/20), 3.8% were in health care personnel (Guan et al.). Seventy-five percent (75%) of these cases were diagnosed in Hubei Province. In a separate study of a selected group of 1,099 patients from mainland China with laboratory-confirmed COVID-19, the percentage of health care workers among the cases was 3.5% (ICN). In Italy, which continues to experience widespread community transmission, 9% of COVID-19 cases are in health care workers. Currently the California Department of Public Health is reporting 42 of 3,006 (1.4%) in their latest stats that include health care workers. This percentage is likely to vary depending on local outbreaks. We can say that health care worker transmission of COVID-19 is lower than that seen with SARS. In the SARS outbreak in Canada in 2002-2003, 43% of the cases were in health care workers.

Updated March 27, 2020

SARS-CoV-2 transmissions have occurred predominantly within a temperature range of 3°C-17°C and a humidity of 4-9 g/m3, suggesting that SARS-CoV-2 transmission may be less efficient in a warmer, humid climate (Bukhari and Jameel). Similarly, in temperate climates, infection by other coronaviruses has been shown to occur primarily in winter (Gaunt et al.). While it may be reasonable to expect a decline in the contagiousness of SARS-CoV-2 in warmer weather, the degree to which weather will impact transmission is not known and will need further monitoring.

Updated April 7, 2020

This remains an area of uncertainty. Several reports of children shedding SARS-CoV-2 have raised concern about the potential impact of asymptomatic shedding from children in the spread of COVID-19 (Kelvin and Halperin). In one study of 36 children admitted with COVID-19 in China, 28% were asymptomatic and 19% had only mild upper respiratory infection (URI) symptoms (Qiu et al.). A report from the U.S. Centers for Disease Control and Prevention (CDC) found that children were less likely (73%) to have symptoms of fever, cough or shortness of breath compared to adults (93%) (CDC). A recent report documented contamination in the hospital room of an asymptomatic infected 6-month-old (Yung et al.). A preprint release of a study from Singapore suggests that children represented the source in ~10% (3/31) of household transmission clusters; this is lower than the 54% (30/56) for H5N1 (Zhu et al.). More testing of children with little or no symptoms and investigations of transmission events in the U.S. and the San Francisco Bay Area are needed to answer this question.

Updated March 31, 2020

Gastrointestinal (GI) symptoms do occur in COVID-19, although not in the majority of patients. We know the virus causing COVID-19 can infect the gut, so it is not surprising that this would occur. A large case series reported 1%-10% of patients with COVID-19 presented with diarrhea or nausea/vomiting (Guan et al.). Other studies reported diarrhea in 14%-27% of patients at the time of diagnosis (Qin et al.). Emerging reports are now describing patients presenting with GI symptoms (i.e., diarrhea, nausea or vomiting) without respiratory symptoms in 3%-10% of cases (Pan et al., Luo et al.). We continue to learn about the clinical presentation of COVID-19 and expect that we will learn more about GI presentations and this disease over the coming months.

Updated April 7, 2020

The main ocular manifestation of COVID-19 is conjunctivitis. A recent report described the ocular findings in a cohort of 38 patients from China: 12 patients (38%) had chemosis, conjunctival hyperemia and epiphora (watery eyes) (Wu et al.). All 12 patients had moderate, severe or critical illness, suggesting that ocular findings may be found in more severe disease. Notably, one patient had epiphora as the first symptom of COVID-19. Two out of the 12 patients had a positive conjunctival swab for COVID-19, suggesting the possibility of transmission directly from the eye. Prior to this report, there were also a few case reports where conjunctivitis was described as part of the clinical syndrome of COVID-19 (Guillen et al.).

Updated March 27, 2020

In published studies to date, the survival rate after intubation is quite low (0%-14%). Yang et al. reported on 52 critically ill patients in Wuhan and found that of 22 intubated patients, only 3 (14%) survived. Zhou et al. reported on the clinical course of 191 patients in Wuhan and found that among 32 intubated patients, only 1 (3%) survived. Similarly, Wu et al. reported on risk factors for acute respiratory distress syndrome (ARDS) and death in 201 patients from Wuhan; only 6 patients were intubated and none survived. Lastly, a report on 21 critically ill patients from Evergreen Hospital in Washington was published on March 19 in JAMA (Arentz et al.). They reported that 15 patients required intubation (71%), and of the 21 patients, 67% have died, 24% remain critically ill, and 10% have been discharged from the ICU; they do not specify in the report a specific mortality rate for the intubated group.

Updated April 3, 2020

To date, most studies have examined conditions associated with severe COVID-19 outside the United States. Findings in the U.S. are similar to those in other countries. Among 7,162 cases reported to the CDC with data on comorbidities and other risk factors, 37.6% had at least one comorbidity or risk factor for severe disease (CDC). The most common included diabetes (10.9%), chronic lung disease (9.2%) and cardiovascular disease (9.0%). A higher proportion of persons admitted to the ICU had at least one comorbidity or risk factor (78%) versus persons hospitalized but not admitted to the ICU (71%) or those not hospitalized (27%). Of 184 fatal cases, 94% were among persons with one or more major comorbidity. These findings are not surprising in light of what is observed with other respiratory viral infections, such as influenza, but we have much to learn regarding COVID-19’s impact on other patient populations, such as persons living with HIV.

Updated June 24, 2020

After early data showed an increased risk of maternal complications in pregnant women with COVID-19 (Chen et al.), researchers in Spain sought to determine whether mode of delivery was associated with maternal or neonatal complications. In a recent study, 82 pregnant women with laboratory-confirmed COVID-19 were stratified by symptom severity at admission (Martínez-Perez et al.). Of the 78 patients who presented with no or mild symptoms of COVID-19, 41 (53%) patients delivered vaginally, and 37 (47%) delivered by C-section. In the latter group, 8 (21.6%) had increased oxygen needs after delivery, and 5 (13.5%) required ICU admission. Among the infants, only 2 developed symptomatic, laboratory-confirmed COVID-19; symptoms for both resolved within 48 hours. After adjusting for potential confounding factors, the authors found that C-section was independently associated with both increased maternal oxygen requirements after delivery and increased risk of NICU admission.

Conclusion: While C-section delivery was independently associated with worse maternal and neonatal outcomes, the conclusions that can be drawn from this small study are limited. More research is needed to better understand the consequences of maternal COVID-19 on maternal and neonatal outcomes.

Additional Resources:

UCSF has launched a nationwide registry for pregnant women with suspected or confirmed COVID-19 and their infants called PRIORITY.

Updated March 30, 2020

Perhaps. This is an evolving story. We know patients with underlying cardiovascular disease are at higher risk for severe COVID-19. However, COVID-19 itself may be associated with cardiac injury. Studies of patients hospitalized with COVID-19 in China reported cardiac injury in 17%-20% of cases (Zhou et al., Shi et al.). One study found that patients with cardiac injury were older, had more comorbidities and had a higher risk of death compared to those without cardiac injury (Shi et al.). Another study reported that 7% of 68 COVID-19-related deaths were due to myocardial damage/heart failure (Ruan et al.). Whether COVID-19-associated cardiac injury is due to direct viral injury or other mechanisms requires further research (Driggin et al.). Awareness of this potential complication, particularly when considering medications with potential cardiac toxicities, should be noted while researchers study potential mechanisms.