Browse the corpus

Background: Virologically confirmed cases of 2019 novel coronavirus (2019-nCoV) in China and other countries have increased sharply (1, 2), leading to concerns regarding its pandemic potential. Viral epidemiology has been characterized sufficiently to permit construction of transmission models that predict the future course of this epidemic (3). Objective: To provide insight into the changing nature of case findings and epidemic growth. Methods: We developed a simple disease-transmission model in which the 2019-nCoV epidemic was modeled as a branching process starting in mid-November 2019, with a serial interval of 7 days (time between cases) and a basic reproduction number (R0) of 2.3 (new cases from each old case), based on available data and assuming no intervention (Figure 1). The epidemic start date aligned our modeled case counts to point estimates from international case exportation data (4). The model estimated plausible values of the effective reproduction number (Re; reproduction number in the presence of control efforts) after implementation of a quarantine in Wuhan and surrounding areas of China on 24 January 2020 (3) (Figure 1). Figure 1. Estimation of cumulative cases with and without implementation of control measures.

Methods: We developed a simple disease-transmission model in which the 2019-nCoV epidemic was modeled as a branching process starting in mid-November 2019, with a serial interval of 7 days (time between cases) and a basic reproduction number (R0) of 2.3 (new cases from each old case), based on available data and assuming no intervention (Figure 1). The epidemic start date aligned our modeled case counts to point estimates from international case exportation data (4). The model estimated plausible values of the effective reproduction number (Re; reproduction number in the presence of control efforts) after implementation of a quarantine in Wuhan and surrounding areas of China on 24 January 2020 (3) (Figure 1). Figure 1. Estimation of cumulative cases with and without implementation of control measures. Serial interval is the average time between cases in a chain of transmission and is used to calculate the number of generations in an epidemic (time since epidemic start ÷ serial interval duration). In the absence of control measures, the total number of cases after t serial intervals depends on R0 (the number of new cases created by an index case in a completely susceptible population in the absence of intervention) and the number of epidemic generations (left-hand equation). Introduction of control is assumed to reduce the reproduction number to Re. The last generation with uncontrolled growth is indicated by tc, with an incident case count of Itc, and we can use the right-hand equation to calculate case numbers in the presence of control. The difference between the 2 curves shows the effect of introducing control measures vs. continued epidemic growth without control. R0 = basic reproduction number; Re = effective reproduction number.

incident case count of Itc, and we can use the right-hand equation to calculate case numbers in the presence of control. The difference between the 2 curves shows the effect of introducing control measures vs. continued epidemic growth without control. R0 = basic reproduction number; Re = effective reproduction number. Re values after intervention can be plotted as epidemic curves in a series of “contours,” similar to altitude values on a map. Because many combinations of model parameters create plausible epidemic trajectories, we have created an interactive tool that produces models with and without control efforts (https://art-bd.shinyapps.io/nCov_control). Findings: Comparison of cumulative case numbers versus model-generated counts shows that reported case numbers remain lower than modeled estimates, but ascertainment is increasingly complete over time (Figure 2). Based on previously published model estimates (4), the fraction of cases reported increased from 2.4% on 12 January 2020 to 11% on 18 January 2020 (4). Our model suggests that (assuming Re remained close to 2.3 after the quarantine on 24 January 2020) reported cases increased to 59% by 31 January 2020 (9930 reported cases vs. 16 860 modeled cases) (1, 2). The fraction of cases reported would be even higher if the reproduction number were lower because of control efforts. Figure 2. Simulated epidemic trajectories and reported cumulative case counts for 2019-nCoV.

Findings: Comparison of cumulative case numbers versus model-generated counts shows that reported case numbers remain lower than modeled estimates, but ascertainment is increasingly complete over time (Figure 2). Based on previously published model estimates (4), the fraction of cases reported increased from 2.4% on 12 January 2020 to 11% on 18 January 2020 (4). Our model suggests that (assuming Re remained close to 2.3 after the quarantine on 24 January 2020) reported cases increased to 59% by 31 January 2020 (9930 reported cases vs. 16 860 modeled cases) (1, 2). The fraction of cases reported would be even higher if the reproduction number were lower because of control efforts. Figure 2. Simulated epidemic trajectories and reported cumulative case counts for 2019-nCoV. The initial growth of the epidemic is based on introduction of the pathogen in mid-November 2019, with R0 = 2.3 and a serial interval of 7 d. The model reproduces estimates of case counts based on volume of internationally exported cases (green squares) (4). Daily cumulative counts of virologically confirmed cases are based on publicly available reports (blue circles) (1, 2). Case counts reported on 3 February 2020 are not compatible with reduction of Re to 1 but could be compatible with reduction to 1.5. If control is achieved, reported case counts will intersect horizontally with the contour lines on this graph. When reported cases move beyond contours vertically, the reproduction numbers represented by those contours become implausible. 2019-nCoV = 2019 novel coronavirus; R0 = basic reproduction number; Re = effective reproduction number.

, reported case counts will intersect horizontally with the contour lines on this graph. When reported cases move beyond contours vertically, the reproduction numbers represented by those contours become implausible. 2019-nCoV = 2019 novel coronavirus; R0 = basic reproduction number; Re = effective reproduction number. Figure 2 shows a narrowing (horizontal distance) between case counts generated by the model and those reported by public health authorities over time. This suggests decreasing reporting times (from >10 days on 27 January 2020 to approximately 4 days by 3 February 2020). Contours generated by the model with intervention give us information about which (average) reproduction numbers may be plausible and which are implausible (Figure 2). If Re had fallen to 1.0 after 24 January 2020, the model predicts fewer cases than are currently being reported (as of 3 February 2020), making this level of control implausible. By contrast, reduction to an Re of 1.5 is plausible on the basis of reported cases and model estimates up to 3 February 2020, but it would also imply complete reporting.

r 24 January 2020, the model predicts fewer cases than are currently being reported (as of 3 February 2020), making this level of control implausible. By contrast, reduction to an Re of 1.5 is plausible on the basis of reported cases and model estimates up to 3 February 2020, but it would also imply complete reporting. Discussion: Using a simple model of epidemic growth that includes the representation of control efforts can provide helpful insights into the growth of the 2019-nCoV epidemic that are not directly observable in publicly reported data. Comparison of modeled and reported case counts suggests that reporting lags are decreasing and case ascertainment increasing over time. The narrowing gap between modeled and confirmed cases shows that the massive public health effort under way in China is increasing ascertainment of 2019-nCoV cases. Large leaps in reported case counts represent both disease activity and a surveillance effort that is “catching up” with an epidemic. Contour plots can be used to indirectly estimate Re after introduction of control efforts, because case counts exceeding a given contour suggest that an Re value is implausible. Potential limitations of this model include underrepresentation of mild infections and its focus on an epidemic currently centered in China. If this epidemic becomes a pandemic, epidemiology in individual countries may diverge. Nonetheless, the tool may help policymakers by allowing inferences about likely underlying dynamics of the epidemic, even when available disease data are delayed or incomplete.

nd its focus on an epidemic currently centered in China. If this epidemic becomes a pandemic, epidemiology in individual countries may diverge. Nonetheless, the tool may help policymakers by allowing inferences about likely underlying dynamics of the epidemic, even when available disease data are delayed or incomplete. We will continue to plot case counts against such projections moving forward (with updated counts incorporated into our online tool). If cumulative case counts flatten and intersect with contour lines horizontally, either control is improving and the mean reproduction number is decreasing or (a pessimistic interpretation) case ascertainment efforts are flagging because of limited laboratory or human resources. Conversely, if reported case counts cross the contour lines above them, that would imply an ever higher minimum value for Re. This article was published at Annals.org on 5 February 2020. Disclaimer: The tool, available at https://art-bd.shinyapps.io/nCov_control, was developed by the authors for this article using a third-party application, which may have limited access and functionality. Neither Annals of Internal Medicine nor the American College of Physicians is responsible for the content and functionality of this online application. Questions regarding the use of the application should be addressed to the corresponding author (e-mail, david.fisman@utoronto.ca). Disclosures: Authors have disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M20-0358.

Disclaimer: The tool, available at https://art-bd.shinyapps.io/nCov_control, was developed by the authors for this article using a third-party application, which may have limited access and functionality. Neither Annals of Internal Medicine nor the American College of Physicians is responsible for the content and functionality of this online application. Questions regarding the use of the application should be addressed to the corresponding author (e-mail, david.fisman@utoronto.ca). Disclosures: Authors have disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M20-0358. Reproducible Research Statement: Study protocol and statistical code: Described in Methods. Data set: Derived from reference 1 and available at https://docs.google.com/spreadsheets/d/19qC9EK2ydaSoKDMkmbbarXBo8Ism_1_6zeMrJh5kZ9Y/edit?usp=sharing.

In December 2019, a cluster of severe pneumonia cases of unknown cause was reported in Wuhan, Hubei province, China. The initial cluster was epidemiologically linked to a seafood wholesale market in Wuhan, although many of the initial 41 cases were later reported to have no known exposure to the market (1). A novel strain of coronavirus belonging to the same family of viruses that cause severe acute respiratory syndrome (SARS) and Middle East respiratory syndrome (MERS), as well as the 4 human coronaviruses associated with the common cold, was subsequently isolated from lower respiratory tract samples of 4 cases on 7 January 2020 (2). Infection with the virus, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), can be asymptomatic or can result in mild to severe symptomatic disease (coronavirus disease 2019 [COVID-19]) (3). On 30 January 2020, the World Health Organization declared that the SARS-CoV-2 outbreak constituted a Public Health Emergency of International Concern, and more than 80 000 confirmed cases had been reported worldwide as of 28 February 2020 (4, 5). On 31 January 2020, the U.S. Centers for Disease Control and Prevention announced that all citizens returning from Hubei province, China, would be subject to mandatory quarantine for up to 14 days (6).

International Concern, and more than 80 000 confirmed cases had been reported worldwide as of 28 February 2020 (4, 5). On 31 January 2020, the U.S. Centers for Disease Control and Prevention announced that all citizens returning from Hubei province, China, would be subject to mandatory quarantine for up to 14 days (6). Our current understanding of the incubation period for COVID-19 is limited. An early analysis based on 88 confirmed cases in Chinese provinces outside Wuhan, using data on known travel to and from Wuhan to estimate the exposure interval, indicated a mean incubation period of 6.4 days (95% CI, 5.6 to 7.7 days), with a range of 2.1 to 11.1 days (7). Another analysis based on 158 confirmed cases outside Wuhan estimated a median incubation period of 5.0 days (CI, 4.4 to 5.6 days), with a range of 2 to 14 days (8). These estimates are generally consistent with estimates from 10 confirmed cases in China (mean incubation period, 5.2 days [CI, 4.1 to 7.0 days] [9]) and from clinical reports of a familial cluster of COVID-19 in which symptom onset occurred 3 to 6 days after assumed exposure in Wuhan (1). These estimates of the incubation period of SARS-CoV-2 are also in line with those of other known human coronaviruses, including SARS (mean, 5 days; range, 2 to 14 days [10]), MERS (mean, 5 to 7 days; range, 2 to 14 days [11]), and non-SARS human coronavirus (mean, 3 days; range, 2 to 5 days [12]).

posure in Wuhan (1). These estimates of the incubation period of SARS-CoV-2 are also in line with those of other known human coronaviruses, including SARS (mean, 5 days; range, 2 to 14 days [10]), MERS (mean, 5 to 7 days; range, 2 to 14 days [11]), and non-SARS human coronavirus (mean, 3 days; range, 2 to 5 days [12]). The incubation period can inform several important public health activities for infectious diseases, including active monitoring, surveillance, control, and modeling. Active monitoring requires potentially exposed persons to contact local health authorities to report their health status every day. Understanding the length of active monitoring needed to limit the risk for missing SARS-CoV-2 infections is necessary for health departments to effectively use limited resources. In this article, we provide estimates of the incubation period of COVID-19 and the number of symptomatic infections missed under different active monitoring scenarios. Methods Data Collection We searched for news and public health reports of confirmed COVID-19 cases in areas with no known community transmission, including provinces, regions, and countries outside Hubei. We searched for reports in both English and Chinese and abstracted the data necessary to estimate the incubation period of COVID-19. Two authors independently reviewed the full text of each case report. Discrepancies were resolved by discussion and consensus.

, including provinces, regions, and countries outside Hubei. We searched for reports in both English and Chinese and abstracted the data necessary to estimate the incubation period of COVID-19. Two authors independently reviewed the full text of each case report. Discrepancies were resolved by discussion and consensus. For each case, we recorded the time of possible exposure to SARS-CoV-2, any symptom onset, fever onset, and case detection. The exact time of events was used when possible; otherwise, we defined conservative upper and lower bounds for the possible interval of each event. For most cases, the interval of possible SARS-CoV-2 exposure was defined as the time between the earliest possible arrival to and latest possible departure from Wuhan. For cases without history of travel to Wuhan but with assumed exposure to an infectious person, the interval of possible SARS-CoV-2 exposure was defined as the maximum possible interval of exposure to the infectious person, including time before the infectious person was symptomatic. We allowed for the possibility of continued exposure within known clusters (for example, families traveling together) when the ordering of transmission was unclear. We assumed that exposure always preceded symptom onset. If we were unable to determine the latest exposure time from the available case report, we defined the upper bound of the exposure interval to be the latest possible time of symptom onset. When the earliest possible time of exposure could not be determined, we defined it as 1 December 2019, the date of symptom onset in the first known case (1); we performed a sensitivity analysis for the selection of this universal lower bound. When the earliest possible time of symptom onset could not be determined, we assumed it to be the earliest time of possible exposure. When the latest time of possible symptom onset could not be determined, we assumed it to be the latest time of possible case detection. Data on age, sex, country of residence, and possible exposure route were also collected.

ptom onset could not be determined, we assumed it to be the earliest time of possible exposure. When the latest time of possible symptom onset could not be determined, we assumed it to be the latest time of possible case detection. Data on age, sex, country of residence, and possible exposure route were also collected. Statistical Analysis Cases were included in the analysis if we had information on the interval of exposure to SARS-CoV-2 and symptom onset. We estimated the incubation time using a previously described parametric accelerated failure time model (13). For our primary analysis, we assumed that the incubation time follows a log-normal distribution, as seen in other acute respiratory viral infections (12). We fit the model to all observations, as well as to only cases where the patient had fever and only those detected inside or outside mainland China in subset analyses. Finally, we also fit 3 other commonly used incubation period distributions (gamma, Weibull, and Erlang). We estimated median incubation time and important quantiles (2.5th, 25th, 75th, and 97.5th percentiles) along with their bootstrapped CIs for each model.

detected inside or outside mainland China in subset analyses. Finally, we also fit 3 other commonly used incubation period distributions (gamma, Weibull, and Erlang). We estimated median incubation time and important quantiles (2.5th, 25th, 75th, and 97.5th percentiles) along with their bootstrapped CIs for each model. Using these estimates of the incubation period, we quantified the expected number of undetected symptomatic cases in an active monitoring program, adapting a method detailed by Reich and colleagues (14). We accounted for varying durations of the active monitoring program (1 to 28 days) and individual risk for symptomatic infection (low risk: 1-in-10 000 chance of infection; medium risk: 1-in-1000 chance; high risk: 1-in-100 chance; infected: 1-in-1 chance). For each bootstrapped set of parameter estimates from the log-normal model, we calculated the probability of a symptomatic infection developing after an active monitoring program of a given length for a given risk level. This model conservatively assumes that persons are exposed to SARS-CoV-2 immediately before the active monitoring program and assumes perfect ascertainment of symptomatic cases that develop under active monitoring. We report the mean and 99th percentile of the expected number of undetected symptomatic cases for each active monitoring scenario.

assumes that persons are exposed to SARS-CoV-2 immediately before the active monitoring program and assumes perfect ascertainment of symptomatic cases that develop under active monitoring. We report the mean and 99th percentile of the expected number of undetected symptomatic cases for each active monitoring scenario. All estimates are based on persons who developed symptoms, and this work makes no inferences about asymptomatic infection with SARS-CoV-2. The analyses were conducted using the coarseDataTools and activemonitr packages in the R statistical programming language, version 3.6.2 (R Foundation for Statistical Computing). All code and data are available at https://github.com/HopkinsIDD/ncov_incubation (release at time of submission at https://zenodo.org/record/3692048) (15). Role of the Funding Source The findings and conclusions in this manuscript are those of the authors and do not necessarily represent the views of the U.S. Centers for Disease Control and Prevention, the National Institute of Allergy and Infectious Diseases, the National Institute of General Medical Sciences, and the Alexander von Humboldt Foundation. The funders had no role in study design, data collection and analysis, preparation of the manuscript, or the decision to submit the manuscript for publication.

ntion, the National Institute of Allergy and Infectious Diseases, the National Institute of General Medical Sciences, and the Alexander von Humboldt Foundation. The funders had no role in study design, data collection and analysis, preparation of the manuscript, or the decision to submit the manuscript for publication. Results We collected data from 181 cases with confirmed SARS-CoV-2 infection detected outside Hubei province before 24 February 2020 (Table 1). Of these, 69 (38%) were female, 108 were male (60%), and 4 (2%) were of unknown sex. The median age was 44.5 years (interquartile range, 34.0 to 55.5 years). Cases were collected from 24 countries and regions outside mainland China (n = 108) and 25 provinces within mainland China (n = 73). Most cases (n = 161) had a known recent history of travel to or residence in Wuhan; others had evidence of contact with travelers from Hubei or persons with known infection. Among those who developed symptoms in the community, the median time from symptom onset to hospitalization was 1.2 days (range, 0.2 to 29.9 days) (Figure 1). Figure 1. SARS-CoV-2 exposure (blue), symptom onset (red), and case detection (green) times for 181 confirmed cases. Shaded regions represent the full possible time intervals for exposure, symptom onset, and case detection; points represent the midpoints of these intervals. SARS-CoV-2 = severe acute respiratory syndrome coronavirus 2.

Figure 1. SARS-CoV-2 exposure (blue), symptom onset (red), and case detection (green) times for 181 confirmed cases. Shaded regions represent the full possible time intervals for exposure, symptom onset, and case detection; points represent the midpoints of these intervals. SARS-CoV-2 = severe acute respiratory syndrome coronavirus 2. Table 1. Characteristics of Patients With Confirmed COVID-19 Included in This Analysis (n = 181)* Fitting the log-normal model to all cases, we estimated the median incubation period of COVID-19 to be 5.1 days (CI, 4.5 to 5.8 days) (Figure 2). We estimated that fewer than 2.5% of infected persons will show symptoms within 2.2 days (CI, 1.8 to 2.9 days) of exposure, and symptom onset will occur within 11.5 days (CI, 8.2 to 15.6 days) for 97.5% of infected persons. The estimate of the dispersion parameter was 1.52 (CI, 1.32 to 1.72), and the estimated mean incubation period was 5.5 days. Figure 2. Cumulative distribution function of the COVID-19 incubation period estimate from the log-normal model. The estimated median incubation period of COVID-19 was 5.1 days (CI, 4.5 to 5.8 days). We estimated that fewer than 2.5% of infected persons will display symptoms within 2.2 days (CI, 1.8 to 2.9 days) of exposure, whereas symptom onset will occur within 11.5 days (CI, 8.2 to 15.6 days) for 97.5% of infected persons. Horizontal bars represent the 95% CIs of the 2.5th, 50th, and 97.5th percentiles of the incubation period distribution. The estimate of the dispersion parameter is 1.52 (CI, 1.32 to 1.72). COVID-19 = coronavirus disease 2019.

ymptom onset will occur within 11.5 days (CI, 8.2 to 15.6 days) for 97.5% of infected persons. Horizontal bars represent the 95% CIs of the 2.5th, 50th, and 97.5th percentiles of the incubation period distribution. The estimate of the dispersion parameter is 1.52 (CI, 1.32 to 1.72). COVID-19 = coronavirus disease 2019. To control for possible bias from symptoms of cough or sore throat, which could have been caused by other more common pathogens, we performed the same analysis on the subset of cases with known time of fever onset (n = 99), using the time from exposure to onset of fever as the incubation time. We estimated the median incubation period to fever onset to be 5.7 days (CI, 4.9 to 6.8 days), with 2.5% of persons experiencing fever within 2.6 days (CI, 2.1 to 3.7 days) and 97.5% having fever within 12.5 days (CI, 8.2 to 17.7 days) of exposure.

using the time from exposure to onset of fever as the incubation time. We estimated the median incubation period to fever onset to be 5.7 days (CI, 4.9 to 6.8 days), with 2.5% of persons experiencing fever within 2.6 days (CI, 2.1 to 3.7 days) and 97.5% having fever within 12.5 days (CI, 8.2 to 17.7 days) of exposure. Because assumptions about the occurrence of local transmission and therefore the period of possible exposure may be less firm within mainland China, we also analyzed only cases detected outside mainland China (n = 108). The median incubation period for these cases was 5.5 days (CI, 4.4 to 7.0 days), with the 95% range spanning from 2.1 (CI, 1.5 to 3.2) to 14.7 (CI, 7.4 to 22.6) days. Alternatively, persons who left mainland China may represent a subset of persons with longer incubation periods, persons who were able to travel internationally before symptom onset within China, or persons who may have chosen to delay reporting symptoms until they left China. Based on cases detected inside mainland China (n = 73), the median incubation period is 4.8 days (CI, 4.2 to 5.6 days), with a 95% range of 2.5 (CI, 1.9 to 3.5) to 9.2 (CI, 6.4 to 12.5) days. Full results of these sensitivity analyses are presented in Appendix Table 1. Appendix Table 1. Percentiles of SARS-CoV-2 Incubation Period From Selected Sensitivity Analyses* We fit other commonly used parameterizations of the incubation period (gamma, Weibull, and Erlang distributions). The incubation period estimates for these alternate parameterizations were similar to those from the log-normal model (Appendix Table 2).

of SARS-CoV-2 Incubation Period From Selected Sensitivity Analyses* We fit other commonly used parameterizations of the incubation period (gamma, Weibull, and Erlang distributions). The incubation period estimates for these alternate parameterizations were similar to those from the log-normal model (Appendix Table 2). Appendix Table 2. Parameter Estimates for Various Parametric Distributions of the Incubation Period of SARS-CoV-2 Using 181 Confirmed Cases* Given these estimates of the incubation period, we predicted the number of symptomatic infections we would expect to miss over the course of an active monitoring program. We classified persons as being at high risk if they have a 1-in-100 chance of developing a symptomatic infection after exposure. For an active monitoring program lasting 7 days, the expected number of symptomatic infections missed for every 10 000 high-risk persons monitored is 21.2 (99th percentile, 36.5) (Table 2 and Figure 3). After 14 days, it is highly unlikely that further symptomatic infections would be undetected among high-risk persons (mean, 1.0 undetected infections per 10 000 persons [99th percentile, 4.8]). However, substantial uncertainty remains in the classification of persons as being at “high,” “medium,” or “low” risk for being symptomatic, and this method does not consider the role of asymptomatic infection. We have created an application to estimate the proportion of missed COVID-19 cases across any active monitoring duration up to 100 days and various population risk levels (16).

ersons as being at “high,” “medium,” or “low” risk for being symptomatic, and this method does not consider the role of asymptomatic infection. We have created an application to estimate the proportion of missed COVID-19 cases across any active monitoring duration up to 100 days and various population risk levels (16). Figure 3. Proportion of known symptomatic SARS-CoV-2 infections that have yet to develop symptoms, by number of days since infection, using bootstrapped estimates from a log-normal accelerated failure time model.

ersons as being at “high,” “medium,” or “low” risk for being symptomatic, and this method does not consider the role of asymptomatic infection. We have created an application to estimate the proportion of missed COVID-19 cases across any active monitoring duration up to 100 days and various population risk levels (16). Figure 3. Proportion of known symptomatic SARS-CoV-2 infections that have yet to develop symptoms, by number of days since infection, using bootstrapped estimates from a log-normal accelerated failure time model. Table 2. Expected Number of Symptomatic SARS-CoV-2 Infections That Would Be Undetected During Active Monitoring, Given Varying Monitoring Durations and Risks for Symptomatic Infection After Exposure* Discussion We present estimates of the incubation period for the novel coronavirus disease (COVID-19) that emerged in Wuhan, Hubei province, China, in 2019. We estimated the median incubation period of COVID-19 to be 5.1 days and expect that nearly all infected persons who have symptoms will do so within 12 days of infection. We found that the current period of active monitoring recommended by the U.S. Centers for Disease Control and Prevention (14 days) is well supported by the evidence (6). Symptomatic disease is frequently associated with transmissibility of a pathogen. However, given recent evidence of SARS-CoV-2 transmission by mildly symptomatic and asymptomatic persons (17, 18), we note that time from exposure to onset of infectiousness (latent period) may be shorter than the incubation period estimated here, with important implications for transmission dynamics.

ty of a pathogen. However, given recent evidence of SARS-CoV-2 transmission by mildly symptomatic and asymptomatic persons (17, 18), we note that time from exposure to onset of infectiousness (latent period) may be shorter than the incubation period estimated here, with important implications for transmission dynamics. Our results are broadly consistent with other estimates of the incubation period (1, 7–9). Our analysis, which was based on 181 confirmed COVID-19 cases, made more conservative assumptions about the possible window of symptom onset and the potential for continued exposure through transmission clusters outside Wuhan. Of note, the use of fixed times of symptom onset, as used in 3 of the 4 prior analyses, will truncate the incubation period distribution by either decreasing the maximum possible incubation period (if the earliest possible time of symptom onset is used) or increasing the minimum possible incubation period (if the midpoint or latest possible time of symptom onset is used). Therefore, using a symptom onset window more accurately accounts for the full distribution of possible incubation periods.

e incubation period (if the earliest possible time of symptom onset is used) or increasing the minimum possible incubation period (if the midpoint or latest possible time of symptom onset is used). Therefore, using a symptom onset window more accurately accounts for the full distribution of possible incubation periods. Although our results support current proposals for the length of quarantine or active monitoring of persons potentially exposed to SARS-CoV-2, longer monitoring periods might be justified in extreme cases. Among those who are infected and will develop symptoms, we expect 101 in 10 000 (99th percentile, 482) will do so after the end of a 14-day monitoring period (Table 2 and Figure 3), and our analyses do not preclude this estimate from being higher. Although it is essential to weigh the costs of extending active monitoring or quarantine against the potential or perceived costs of failing to identify a symptomatic case, there may be high-risk scenarios (for example, a health care worker who cared for a COVID-19 patient while not wearing personal protective equipment) where it could be prudent to extend the period of active monitoring.

ing or quarantine against the potential or perceived costs of failing to identify a symptomatic case, there may be high-risk scenarios (for example, a health care worker who cared for a COVID-19 patient while not wearing personal protective equipment) where it could be prudent to extend the period of active monitoring. This analysis has several important limitations. Our data include early case reports, with associated uncertainty in the intervals of exposure and symptom onset. We have used conservative bounds of possible exposure and symptom onset where exact times were not known, but there may be further inaccuracy in these data that we have not considered. We have exclusively considered reported, confirmed cases of COVID-19, which may overrepresent hospitalized persons and others with severe symptoms, although we note that the proportion of mild cases detected has increased as surveillance and monitoring systems have been strengthened. The incubation period for these severe cases may differ from that of less severe or subclinical infections and is not typically an applicable measure for those with asymptomatic infections.

gh we note that the proportion of mild cases detected has increased as surveillance and monitoring systems have been strengthened. The incubation period for these severe cases may differ from that of less severe or subclinical infections and is not typically an applicable measure for those with asymptomatic infections. Our model assumes a constant risk for SARS-CoV-2 infection in Wuhan from 1 December 2019 to 30 January 2020, based on the date of symptom onset of the first known case and the last known possible exposure within Wuhan in our data set. This is a simplification of infection risk, given that the outbreak has shifted from a likely common-source outbreak associated with a seafood market to human-to-human transmission. Moreover, phylogenetic analysis of 38 SARS-CoV-2 genomes suggests that the virus may have been circulating before December 2019 (19). To test the sensitivity of our estimates to that assumption, we performed an analysis where cases with unknown lower bounds on exposure were set to 1 December 2018, a full year earlier than in our primary analysis. Changing this assumption had little effect on the estimates of the median (0.2 day longer than for the overall estimate) and the 97.5th quantile (0.1 day longer) of the incubation period. In data sets such as ours, where we have adequate observations with well-defined minimum and maximum possible incubation periods for many cases, extending the universal lower bound has little bearing on the overall estimates.

n for the overall estimate) and the 97.5th quantile (0.1 day longer) of the incubation period. In data sets such as ours, where we have adequate observations with well-defined minimum and maximum possible incubation periods for many cases, extending the universal lower bound has little bearing on the overall estimates. This work provides additional evidence for a median incubation period for COVID-19 of approximately 5 days, similar to SARS. Assuming infection occurs at the initiation of monitoring, our estimates suggest that 101 out of every 10 000 cases will develop symptoms after 14 days of active monitoring or quarantine. Whether this rate is acceptable depends on the expected risk for infection in the population being monitored and considered judgment about the cost of missing cases (14). Combining these judgments with the estimates presented here can help public health officials to set rational and evidence-based COVID-19 control policies. This article was published at Annals.org on 10 March 2020. * Dr. Lauer and Ms. Grantz share first authorship. Acknowledgment: The authors thank all who have collected, prepared, and shared data throughout this outbreak. They are particularly grateful to Dr. Kaiyuan Sun, Ms. Jenny Chen, and Dr. Cecile Viboud from the Division of International Epidemiology and Population Studies, Fogarty International Center, National Institutes of Health; Dr. Moritz Kraemer and the open COVID-19 data working group; and the Johns Hopkins Center for Systems Science and Engineering.

ul to Dr. Kaiyuan Sun, Ms. Jenny Chen, and Dr. Cecile Viboud from the Division of International Epidemiology and Population Studies, Fogarty International Center, National Institutes of Health; Dr. Moritz Kraemer and the open COVID-19 data working group; and the Johns Hopkins Center for Systems Science and Engineering. Grant Support: By the U.S. Centers for Disease Control and Prevention (NU2GGH002000), the National Institute of Allergy and Infectious Diseases (R01 AI135115), the National Institute of General Medical Sciences (R35 GM119582), and the Alexander von Humboldt Foundation. Disclosures: Dr. Lauer reports grants from the National Institute of Allergy and Infectious Diseases and the U.S. Centers for Disease Control and Prevention during the conduct of the study. Ms. Grantz reports a grant from the U.S. Centers for Disease Control and Prevention during the conduct of the study. Dr. Reich reports grants from the National Institute of General Medical Sciences and the Alexander von Humboldt Foundation during the conduct of the study. Dr. Lessler reports a grant from the U.S. Centers for Disease Control and Prevention during the conduct of the study. Authors not named here have disclosed no conflicts of interest. Disclosures can also be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M20-0504.

ing the conduct of the study. Dr. Lessler reports a grant from the U.S. Centers for Disease Control and Prevention during the conduct of the study. Authors not named here have disclosed no conflicts of interest. Disclosures can also be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M20-0504. Editors' Disclosures: Christine Laine, MD, MPH, Editor in Chief, reports that her spouse has stock options/holdings with Targeted Diagnostics and Therapeutics. Darren B. Taichman, MD, PhD, Executive Editor, reports that he has no financial relationships or interests to disclose. Cynthia D. Mulrow, MD, MSc, Senior Deputy Editor, reports that she has no relationships or interests to disclose. Jaya K. Rao, MD, MHS, Deputy Editor, reports that she has stock holdings/options in Eli Lilly and Pfizer. Christina C. Wee, MD, MPH, Deputy Editor, reports employment with Beth Israel Deaconess Medical Center. Sankey V. Williams, MD, Deputy Editor, reports that he has no financial relationships or interests to disclose. Yu-Xiao Yang, MD, MSCE, Deputy Editor, reports that he has no financial relationships or interest to disclose. Reproducible Research Statement: Study protocol: Not applicable. Statistical code and data set: Available at https://github.com/HopkinsIDD/ncov_incubation. Corresponding Author: Justin Lessler, PhD, Department of Epidemiology, Bloomberg School of Public Health, Johns Hopkins University, 615 North Wolfe Street, Baltimore, MD 21205; e-mail, justin@jhu.edu.

Reproducible Research Statement: Study protocol: Not applicable. Statistical code and data set: Available at https://github.com/HopkinsIDD/ncov_incubation. Corresponding Author: Justin Lessler, PhD, Department of Epidemiology, Bloomberg School of Public Health, Johns Hopkins University, 615 North Wolfe Street, Baltimore, MD 21205; e-mail, justin@jhu.edu. Current Author Addresses: Drs. Lauer, Meredith, and Lessler; Ms. Grantz; Ms. Bi; Mr. Jones; and Ms. Zheng: Department of Epidemiology, Bloomberg School of Public Health, Johns Hopkins University, 615 North Wolfe Street, Baltimore, MD 21205. Dr. Azman: Médecins Sans Frontières, Rue de Lausanne 72, 1202 Genève, Switzerland. Dr. Reich: Department of Biostatistics and Epidemiology, Amherst School of Public Health and Health Sciences, University of Massachusetts, 715 North Pleasant Street, Amherst, MA 01003-9304. Author Contributions: Conception and design: S.A. Lauer, K.H. Grantz, F.K. Jones, N.G. Reich, J. Lessler. Analysis and interpretation of the data: S.A. Lauer, K.H. Grantz, Q. Bi, F.K. Jones, N.G. Reich, J. Lessler. Drafting of the article: S.A. Lauer, K.H. Grantz, Q. Bi, F.K. Jones, A.S. Azman, N.G. Reich. Critical revision of the article for important intellectual content: Q. Bi, F.K. Jones, A.S. Azman, N.G. Reich, J. Lessler. Final approval of the article: S.A. Lauer, K.H. Grantz, Q. Bi, F.K. Jones, Q. Zheng, H.R. Meredith, A.S. Azman, N.G. Reich, J. Lessler. Statistical expertise: Q. Bi, N.G. Reich, J. Lessler. Collection and assembly of data: S.A. Lauer, K.H. Grantz, Q. Bi, F.K. Jones, Q. Zheng, H.R. Meredith.

Critical revision of the article for important intellectual content: Q. Bi, F.K. Jones, A.S. Azman, N.G. Reich, J. Lessler. Final approval of the article: S.A. Lauer, K.H. Grantz, Q. Bi, F.K. Jones, Q. Zheng, H.R. Meredith, A.S. Azman, N.G. Reich, J. Lessler. Statistical expertise: Q. Bi, N.G. Reich, J. Lessler. Collection and assembly of data: S.A. Lauer, K.H. Grantz, Q. Bi, F.K. Jones, Q. Zheng, H.R. Meredith. Previous Posting: This manuscript was posted as a preprint on medRxiv on 4 February 2020. doi:10.1101/2020.02.02.20020016

Background: Although many studies have demonstrated the epidemiologic characteristics of SARS–CoV-2 disease (COVID-19), details of pathologic changes in the lung are still lacking. Objective: To describe the histopathologic changes in the lung of a patient with COVID-19. Case Report: A 72-year-old man with a history of diabetes and hypertension presented with fever and cough. His throat and pharyngeal swabs were positive for SARS–CoV-2 by day 6 after the initial symptoms. Rapidly progressive respiratory failure required endotracheal intubation and mechanical ventilation 1 week after presentation. Lung tissue was obtained by transthoracic 14-gauge needle biopsy from the left upper anterior segment (Figure 1, A, arrow), left upper lingular segment (Figure 1, B, arrow), and left lower lobe (Figure 1, C, arrow), coinciding with ground-glass opacities on chest computed tomography (CT). Two throat swab samples were collected from the tonsils and posterior pharyngeal wall. Figure 1. Computed tomographic images obtained from the patient 3 weeks after initial clinical manifestations of COVID-19 and 2 weeks before transthoracic biopsy, demonstrating ground glass–like opacifications. Pleural thickening and enlarged mediastinal lymph nodes were present. Arrows indicate the approximate locations of the subsequently obtained postmortem transthoracic needle biopsy samples. A. Left upper anterior segment. B. Left upper lingular segment. C. Left lower lobe.

Figure 1. Computed tomographic images obtained from the patient 3 weeks after initial clinical manifestations of COVID-19 and 2 weeks before transthoracic biopsy, demonstrating ground glass–like opacifications. Pleural thickening and enlarged mediastinal lymph nodes were present. Arrows indicate the approximate locations of the subsequently obtained postmortem transthoracic needle biopsy samples. A. Left upper anterior segment. B. Left upper lingular segment. C. Left lower lobe. Biopsy lung sections were analyzed with hematoxylin–eosin staining, and immunostaining for SARS–CoV-2 was conducted as reported elsewhere (1). Throat swabs were assessed for SARS–CoV-2 by using real-time reverse transcriptase polymerase chain reaction assays (2). The CT scans revealed patchy bilateral ground glass–like opacifications (Figure 1 A-C, arrows). Despite antiviral therapies, respiratory and hemodynamic instability continued and the patient died 3 weeks after diagnosis. Permission for postmortem transthoracic needle biopsy, but not autopsy, was obtained from the patient's family.

revealed patchy bilateral ground glass–like opacifications (Figure 1 A-C, arrows). Despite antiviral therapies, respiratory and hemodynamic instability continued and the patient died 3 weeks after diagnosis. Permission for postmortem transthoracic needle biopsy, but not autopsy, was obtained from the patient's family. Histopathologic examination of lung biopsy tissues revealed diffuse alveolar damage, organizing phase. Denuded alveolar lining cells (Figure 2, A-1, arrow 1), with reactive type II pneumocyte hyperplasia, were noted (Figure 2, A-1, arrow 2). Intra-alveolar fibrinous exudates were present (Figure 2, A-2, arrow 3), along with loose interstitial fibrosis and chronic inflammatory infiltrates (Figure 2, A-2, arrow 4). Intra-alveolar loose fibrous plugs of organizing pneumonia were noted (Figure 2, A-3, arrow 5), with presence of intra-alveolar organizing fibrin seen in most foci (Figure 2, A-4, arrow 6). Figure 2. Histopathologic examination of lung biopsy tissues and immunostaining from a patient who died of COVID-19 (×100 magnification).

Histopathologic examination of lung biopsy tissues revealed diffuse alveolar damage, organizing phase. Denuded alveolar lining cells (Figure 2, A-1, arrow 1), with reactive type II pneumocyte hyperplasia, were noted (Figure 2, A-1, arrow 2). Intra-alveolar fibrinous exudates were present (Figure 2, A-2, arrow 3), along with loose interstitial fibrosis and chronic inflammatory infiltrates (Figure 2, A-2, arrow 4). Intra-alveolar loose fibrous plugs of organizing pneumonia were noted (Figure 2, A-3, arrow 5), with presence of intra-alveolar organizing fibrin seen in most foci (Figure 2, A-4, arrow 6). Figure 2. Histopathologic examination of lung biopsy tissues and immunostaining from a patient who died of COVID-19 (×100 magnification). A. Histopathologic examination revealing diffuse alveolar damage, organizing phase (A-1); denudation of alveolar lining cells (arrow 1), with presence of reactive type II pneumocyte hyperplasia (arrow 2) (A-2); intra-alveolar fibrinous exudates (arrow 3) and interstitial loose fibrosis with chronic inflammatory infiltrates (arrow 4) (A-3); and intra-alveolar loose fibrous plugs (arrow 5) (A-4). In most foci, intra-alveolar organizing fibrin is seen (arrow 6). B. Immunostaining of SARS–CoV-2 in lung sections. Images were taken under light and fluorescent conditions, respectively (f×100 magnification). Merged images were also generated. Blue arrows indicate interstitial areas between the alveoli, and green arrows indicate injured epithelial cells desquamated into the alveolar spaces. The dashed black lines indicate the blood vessel. Immunostaining of SARS–CoV-2 was done by using a rabbit polyclonal antibody (made in house, 1:100) against the Rp3 NP protein, which is highly conserved between SARS-CoV and SARS–CoV-2, followed by probing with a Cy3-conjugated goat antirabbit IgG (1:50, Abcam, ab6939). C. Positive and negative controls for immunostaining. For the positive control, the Huh7 cells were infected with SARS–CoV-2 at multiplicity of infection of 0.5 for 48 hours. After extensive washes, the cells were then fixed with 2.5% (wt/vol) glutaraldehyde. The infected cells were stained in red, and nuclei were stained with DAPI (Beyotime, Wuhan, China) in blue. For the negative control, biopsy lung sections derived from a patient with HIV who died of fungal infection were stained in parallel with lung sections from the patient with COVID-19 as above.

ol) glutaraldehyde. The infected cells were stained in red, and nuclei were stained with DAPI (Beyotime, Wuhan, China) in blue. For the negative control, biopsy lung sections derived from a patient with HIV who died of fungal infection were stained in parallel with lung sections from the patient with COVID-19 as above. Immunostaining of lung sections with an antibody to the Rp3 NP protein of SARS–CoV-2 revealed prominent expression on alveolar epithelial cells (Figure 2, B, top panel), including damaged, desquamated cells within the alveolar space (Figure 2, B, bottom panel, green arrows). In contrast, viral protein expression was minimally detectable on blood vessels (Figure 2, B, dashed black line) or in the interstitial areas between alveoli (Figure 2, B, bottom panel, blue arrows). Immunostaining of Huh7 cells infected with SARS–CoV and of lung sections from an HIV-positive patient who died of fungal infection served as positive and negative staining controls, respectively (Figure 2, C). Discussion: The histopathologic changes seen on postmortem transthoracic needle biopsies from a patient with COVID-19 who had respiratory failure and radiographic bilateral ground-glass opacities are consistent with diffuse alveolar damage. Although such nonspecific findings may be seen in response to several conditions that result in respiratory failure, its demonstration in the setting of COVID-19 helps to inform the clinical course of disease.

iratory failure and radiographic bilateral ground-glass opacities are consistent with diffuse alveolar damage. Although such nonspecific findings may be seen in response to several conditions that result in respiratory failure, its demonstration in the setting of COVID-19 helps to inform the clinical course of disease. Our study is limited by our inability to obtain larger tissue specimens. The present findings warrant further study with larger tissue samples, obtained by open or thoracoscopic lung biopsy, or autopsy, for example. This article was published at Annals.org on 12 March 2020. Note: Authors indicated with an asterisk (Drs. H. Zhang, P. Zhou, Y. Wei, H. Yue, Y. Wang, and M. Hu) contributed equally to this article. Disclosures: Authors have disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M20-0533. Financial Support: By the National Natural Science Foundation of China (grants 81974456 and 91749207); the Clinical Research Physician Program of Tongji Medical College, Huazhong University of Science and Technology (grant 5001540075); and the SARS-CoV-2 Pneumonia Emergency Technology Public Relations Project (grants 2020FCA009 and 2020FCA026).

The international community has witnessed the emergence of novel coronavirus–associated respiratory diseases, including severe acute respiratory syndrome (SARS) in 2002 to 2003 and Middle East respiratory syndrome (MERS) in 2012 to 2013. In 2014, Ebola emerged in western Africa, where it had not previously been seen. Now, 18 years after SARS, we are in the midst of an epidemic known as coronavirus disease 2019 (COVID-19), caused by the novel SARS coronavirus 2 (SARS-CoV-2). With these infections come significant morbidity and mortality, tremendous health care disruptions and resource use, and collateral economic and societal costs.

Now, 18 years after SARS, we are in the midst of an epidemic known as coronavirus disease 2019 (COVID-19), caused by the novel SARS coronavirus 2 (SARS-CoV-2). With these infections come significant morbidity and mortality, tremendous health care disruptions and resource use, and collateral economic and societal costs. In the first 6 weeks of the current epidemic, the number of cases of COVID-19 has surpassed those of SARS and MERS during the course of those epidemics, raising questions about strategies to control the spread of infection. A major strategy has focused on “macro” public health responses, such as travel restrictions, public gathering and school closures, and city quarantines. However, experience with other respiratory viruses suggests that travel restrictions have a limited effect. Mateus and colleagues (1) found that such restrictions decreased new cases of influenza by only 3% and delayed but did not prevent influenza epidemics. Similarly, Errett and colleagues (2) identified minimal evidence to support the effectiveness of travel bans as a control measure for emerging infectious diseases. Read and colleagues (3) suggest that, because only 5% of infections have been identified, even a travel reduction that is 99% effective may reduce the epidemic outside Wuhan province by no more than 24.9%. Other investigators (4) estimate that almost 59 000 cases occurred in Wuhan and 3500 in other regions of China before the travel ban was implemented. Hence, the ban may simply reduce the progression of the outbreak by only 3 to 5 days within China. Finally, a recent report (5) suggests that 46% of cases would be missed by airport-based screening because of COVID-19's incubation period, the spectrum of symptoms, and the time during the incubation period in which persons may fly. Available data specific to COVID-19 suggest that screening and restricting travelers may have a limited effect on containment.

at 46% of cases would be missed by airport-based screening because of COVID-19's incubation period, the spectrum of symptoms, and the time during the incubation period in which persons may fly. Available data specific to COVID-19 suggest that screening and restricting travelers may have a limited effect on containment. Because travel interventions will not prevent transmission to new regions, vigilant infection control measures are critical: aggressive patient screening, active contact tracing, and isolation. Ebola, SARS, MERS, and COVID-19 all have nonspecific clinical presentations, but each emerged in a specific geographic area, and the epidemiologic links to these regions were key in guiding clinicians to implement proper barrier protections and patient evaluation. This led public health agencies, including the World Health Organization and U.S. Centers for Disease Control and Prevention, to recommend a systematic approach to patients presenting with a relevant exposure and symptoms of an acute respiratory viral infection, such as SARS-CoV or MERS-CoV. Early recognition of potential cases was critical in limiting transmission by enabling enhanced prevention and control of infections and preemptive care. Mathematical models developed during the SARS and MERS outbreaks support the effectiveness of such strategies. Identifying patients with potential exposure or symptoms facilitated prompt isolation and, in health care settings, led to additional prevention and case-finding measures. Of note, it triggered health care personnel to use personal protective equipment, patient isolation, and hand hygiene. In the SARS outbreak, these measures prevented transmission of SARS-CoV even without the availability of effective vaccines and therapy. Indeed, these interventions have demonstrated superior efficacy over travel restrictions: Respiratory virus infections were reduced by 46% through hand hygiene, 77% through masks or respirators, and 32% to 33% through gowns and gloves (6).

ion of SARS-CoV even without the availability of effective vaccines and therapy. Indeed, these interventions have demonstrated superior efficacy over travel restrictions: Respiratory virus infections were reduced by 46% through hand hygiene, 77% through masks or respirators, and 32% to 33% through gowns and gloves (6). Climate change, increasing global travel, and an evolving human–animal interface are likely to increase the frequency of novel infectious diseases. Although early identification of acute respiratory viral illness is key to trigger actions to interrupt the chain of transmission, it is often delayed. Surveillance systems using artificial intelligence are promising, as is more effective personal protective equipment, but patient vital signs are available now as powerful indicators of how quickly we need to intervene and what path to take.

o trigger actions to interrupt the chain of transmission, it is often delayed. Surveillance systems using artificial intelligence are promising, as is more effective personal protective equipment, but patient vital signs are available now as powerful indicators of how quickly we need to intervene and what path to take. Vital signs—temperature, heart rate, respiratory rate, and blood pressure—help us assess a patient's health status, triage the patient to appropriate care, determine potential diagnoses, and predict recovery. Given the increasing frequency of emerging infectious diseases that are geographically linked, is it time to add a “fifth vital sign”? A simple, targeted travel history can help us put symptoms of infection in context and trigger us to take a more detailed history, do appropriate testing, and rapidly implement protective measures. An expanded set of vital signs may signal a lurking communicable infection and flag potential risks to health care personnel and other patients. Furthermore, electronic health records can integrate travel history with computerized decision support to suggest specific diagnoses in febrile returning travelers (7, 8).

es. An expanded set of vital signs may signal a lurking communicable infection and flag potential risks to health care personnel and other patients. Furthermore, electronic health records can integrate travel history with computerized decision support to suggest specific diagnoses in febrile returning travelers (7, 8). The lessons from SARS, MERS, and Ebola tell us that early case identification is critical to protect both patients and those caring for them. In 2014, a patient presented to a Dallas emergency department after returning from Liberia with low-grade fever, abdominal pain, dizziness, nausea, and headache (9). The patient had Ebola. Because clinicians did not obtain the 1 potentially distinguishing clinical clue—a travel history—patient and caregiver well-being was compromised.

nt presented to a Dallas emergency department after returning from Liberia with low-grade fever, abdominal pain, dizziness, nausea, and headache (9). The patient had Ebola. Because clinicians did not obtain the 1 potentially distinguishing clinical clue—a travel history—patient and caregiver well-being was compromised. All members of the health care team need training on how to integrate key epidemiologic information, such as travel history, into their risk assessments, in the same way they are trained to ask about tobacco exposure to assess risks for cancer and heart disease. They need a simple script to elicit clues for emerging infectious diseases and must be informed about current emerging pathogenic threats, such as COVID-19. Travel history could serve as a warning sign that prompts protective measures. Of course, we must implement such a change thoughtfully, with attention to unintended consequences—as shown by the inclusion of pain scores as a vital sign, which may have contributed to the opioid misuse crisis. However, we believe that the urgent threat of communicable diseases makes collection of travel history necessary. The current novel coronavirus is a troublesome reminder—on the heels of SARS, MERS, and Ebola—that national, regional, and institutional planning must learn from the past and remain vigilant and focused on vital measures to protect us all. This article was published at Annals.org on 3 March 2020. Acknowledgment: The authors thank James “Brad” Cutrell, MD, for his comments and suggestions.

All members of the health care team need training on how to integrate key epidemiologic information, such as travel history, into their risk assessments, in the same way they are trained to ask about tobacco exposure to assess risks for cancer and heart disease. They need a simple script to elicit clues for emerging infectious diseases and must be informed about current emerging pathogenic threats, such as COVID-19. Travel history could serve as a warning sign that prompts protective measures. Of course, we must implement such a change thoughtfully, with attention to unintended consequences—as shown by the inclusion of pain scores as a vital sign, which may have contributed to the opioid misuse crisis. However, we believe that the urgent threat of communicable diseases makes collection of travel history necessary. The current novel coronavirus is a troublesome reminder—on the heels of SARS, MERS, and Ebola—that national, regional, and institutional planning must learn from the past and remain vigilant and focused on vital measures to protect us all. This article was published at Annals.org on 3 March 2020. Acknowledgment: The authors thank James “Brad” Cutrell, MD, for his comments and suggestions. Disclosures: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M20-0643. Corresponding Author: Trish M. Perl, MD, MSc, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard Y7.312, Dallas, TX 75390; e-mail, Trish.Perl@UTSouthwestern.edu.

Acknowledgment: The authors thank James “Brad” Cutrell, MD, for his comments and suggestions. Disclosures: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M20-0643. Corresponding Author: Trish M. Perl, MD, MSc, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard Y7.312, Dallas, TX 75390; e-mail, Trish.Perl@UTSouthwestern.edu. Current Author Addresses: Dr. Perl: University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard Y7.312, Dallas, TX 75390. Dr. Price: Denver Health and Hospital, 777 Bannock Street MC-2600, Denver, CO 80204. Author Contributions: Conception and design: T.M. Perl, C.S. Price. Drafting of the article: T.M. Perl, C.S. Price. Critical revision of the article for important intellectual content: T.M. Perl, C.S. Price. Final approval of the article: T.M. Perl, C.S. Price. Administrative, technical, or logistic support: T.M. Perl, C.S. Price. Collection and assembly of data: T.M. Perl, C.S. Price.

It is increasingly apparent that severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is optimized to spread widely. It causes mild but prolonged disease, infected persons are contagious even when minimally symptomatic or asymptomatic, the incubation period can extend beyond 14 days, and some patients seem susceptible to reinfection (1–3). These factors make it inevitable that patients with respiratory viral syndromes that are mild or nonspecific will introduce the virus into hospitals, leading to clusters of nosocomial infections. The signs and symptoms of coronavirus disease 2019 (COVID-19) are largely indistinguishable from those of other respiratory virus infections. Less than one half of patients with confirmed disease have fever on initial presentation (4). The sensitivity of a single nasopharyngeal swab early in the course of disease is only 70% (5). Multiple reports already exist of delayed diagnoses leading to nosocomial transmissions.

se of other respiratory virus infections. Less than one half of patients with confirmed disease have fever on initial presentation (4). The sensitivity of a single nasopharyngeal swab early in the course of disease is only 70% (5). Multiple reports already exist of delayed diagnoses leading to nosocomial transmissions. How bad will it be? Characterizing the morbidity rate of COVID-19 is challenging because case detection in the early stages of an outbreak is biased toward severe disease. An initial series reported a mortality rate of 15% (6). A subsequent analysis that included patients who were less sick reported a mortality rate of 2.3% (7), but this is still likely an overestimate. Mortality rates are substantially lower outside than inside Hubei province, where the outbreak began (114 deaths among 13 152 patients [0.9%] vs. 2986 deaths among 67 707 patients [4.4%] as of 8 March 2020). This is presumably because of Hubei's initial focus on patients with severe disease, constraints on the province's testing and care capacity, and the passage of more time since the outbreak began in Hubei versus other provinces allowing more time for patients to declare themselves (8). More to the point, current mortality estimates minimally account for patients with mild or asymptomatic infections, an important aspect of this epidemic (9). Case detection is still primarily focused on identifying patients with fever, cough, or shortness of breath; this focus leads to underestimation of the number of infected persons, overestimation of the mortality rate, and ongoing spread of disease.

d or asymptomatic infections, an important aspect of this epidemic (9). Case detection is still primarily focused on identifying patients with fever, cough, or shortness of breath; this focus leads to underestimation of the number of infected persons, overestimation of the mortality rate, and ongoing spread of disease. What can we do to prevent further spread of infection? We have to be more aggressive about case detection. Current screening is still focused on identifying patients with foreign travel or contacts with known cases. Both of these foci no longer reflect the current status of this epidemic given increasing evidence of community spread. We need to be able to test patients with milder syndromes regardless of travel or contact history. The U.S. Centers for Disease Control and Prevention has updated its “person under investigation” criteria to permit this, but there is still a severe shortage of readily available tests.

sing evidence of community spread. We need to be able to test patients with milder syndromes regardless of travel or contact history. The U.S. Centers for Disease Control and Prevention has updated its “person under investigation” criteria to permit this, but there is still a severe shortage of readily available tests. More broadly, however, the best way to protect hospitals against COVID-19 is to bolster our approach to routine respiratory viruses (that is, influenza, respiratory syncytial virus, parainfluenza, adenovirus, human metapneumovirus, and “conventional” coronaviruses). This will simultaneously improve care for current patients, make work safer for clinicians, and help prevent the incursion of occult COVID-19 into hospitals. We underestimate the contagiousness and seriousness of routine respiratory viruses. We underappreciate that 30% to 50% of cases of community-acquired pneumonia are caused by viruses, that nosocomial transmission of respiratory viruses is common, and that “routine” respiratory viruses cause substantial morbidity and mortality that may not differ much from those caused by SARS-CoV-2 once minimally symptomatic COVID-19 is accounted for. Respiratory viruses infect millions of persons each year (about 10% of the population) and cause tens of thousands of deaths in the United States alone (10). They can cause severe pneumonia, predispose patients to bacterial superinfection, and exacerbate cardiac and pulmonary conditions up to and including death.

for. Respiratory viruses infect millions of persons each year (about 10% of the population) and cause tens of thousands of deaths in the United States alone (10). They can cause severe pneumonia, predispose patients to bacterial superinfection, and exacerbate cardiac and pulmonary conditions up to and including death. Most hospitals, however, manage respiratory viruses passively. We rely on signs alone to deter visitors with upper respiratory tract infections from visiting, we isolate patients in private rooms only if they test positive for influenza virus (even though many other viruses can cause influenza-like syndromes that are equally morbid), we discontinue precautions in patients with acute respiratory tract syndromes if they test negative for viruses (even though viral tests have variable and imperfect sensitivity), we consider masks alone to be adequate protection (even though viruses can be transmitted via fomites and eye contact as well as mouth and nose contact), and we tolerate health care workers coming to work with upper respiratory tract infections so long as they are not febrile.

have variable and imperfect sensitivity), we consider masks alone to be adequate protection (even though viruses can be transmitted via fomites and eye contact as well as mouth and nose contact), and we tolerate health care workers coming to work with upper respiratory tract infections so long as they are not febrile. Our halfhearted approach to endemic respiratory viruses is a source of harm to our patients and puts us at increased risk for COVID-19 infiltration. To cause a nosocomial outbreak, it will take just 1 patient with occult COVID-19 who is hospitalized, tests negative for influenza virus, and is taken off precautions despite persistent respiratory symptoms. Or just 1 visitor with COVID-19 and mild respiratory symptoms who is permitted free access to the hospital because it does not have an active screening and exclusion policy for visitors with respiratory tract symptoms. Or just 1 infected health care worker who decides to soldier through a shift despite a sore throat and runny nose.

itor with COVID-19 and mild respiratory symptoms who is permitted free access to the hospital because it does not have an active screening and exclusion policy for visitors with respiratory tract symptoms. Or just 1 infected health care worker who decides to soldier through a shift despite a sore throat and runny nose. We need to be more aggressive about respiratory hygiene and placing restrictions on patients, visitors, and health care workers with even mild symptoms of upper respiratory tract infection. Potential policies to consider include the following: 1) screening all visitors for any respiratory symptoms that may be related to a virus, including fever, myalgias, pharyngitis, rhinorrhea, and cough, and excluding them from visiting until they are better; 2) restricting health care workers from working if they have any upper respiratory tract symptoms, even in the absence of fever; and 3) screening all patients, testing for all respiratory viruses (including SARS-CoV-2) in those with positive screening results regardless of illness severity, and using precautions (single rooms, contact precautions, droplet precautions, and eye protection) for patients with respiratory syndromes for the duration of their symptoms regardless of viral test results. A collateral benefit is that if a patient is subsequently diagnosed with COVID-19, staff who used these precautions will be considered minimally exposed and will be able to continue working.

and eye protection) for patients with respiratory syndromes for the duration of their symptoms regardless of viral test results. A collateral benefit is that if a patient is subsequently diagnosed with COVID-19, staff who used these precautions will be considered minimally exposed and will be able to continue working. None of these measures will be easy. Restricting visitors will be psychologically difficult for patients and loved ones, maintaining respiratory precautions for the duration of patients' symptoms will strain supplies in all hospitals and bed capacity in hospitals that depend on shared rooms, and preventing health care providers with mild illness from working will compromise staffing. But if we are frank about the morbidity and mortality of all respiratory viruses, including SARS-CoV-2, this is the best thing we can do for our patients and colleagues regardless of COVID-19. This article was published at Annals.org on 11 March 2020. Disclosures: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M20-0751. Corresponding Author: Michael Klompas, MD, MPH, Department of Population Medicine, Harvard Medical School and Harvard Pilgrim Health Care Institute, 401 Park Drive, Suite 401 East, Boston, MA 02215; e-mail, mklompas@bwh.harvard.edu. Author Contributions: Conception and design: M. Klompas. Drafting of the article: M. Klompas. Critical revision of the article for important intellectual content: M. Klompas. Final approval of the article: M. Klompas.

Background: The coronavirus disease 2019 (COVID-19) epidemic began in Wuhan, China, in late 2019 and continues to spread globally (1), with exported cases confirmed in 109 countries at the time of writing (2). During the interval between 19 February and 23 February 2020, Iran reported its first 43 cases, with 8 deaths. Three exported cases originating in Iran were identified, suggesting an underlying burden of disease in that country greater than that indicated by reported cases. A large epidemic in Iran could further fuel global dissemination of COVID-19. Objective: To quantify the COVID-19 outbreak size in Iran on the basis of known exported case counts and air travel links between Iran and other countries, and to anticipate where infections originating in Iran may spread next.

Background: The coronavirus disease 2019 (COVID-19) epidemic began in Wuhan, China, in late 2019 and continues to spread globally (1), with exported cases confirmed in 109 countries at the time of writing (2). During the interval between 19 February and 23 February 2020, Iran reported its first 43 cases, with 8 deaths. Three exported cases originating in Iran were identified, suggesting an underlying burden of disease in that country greater than that indicated by reported cases. A large epidemic in Iran could further fuel global dissemination of COVID-19. Objective: To quantify the COVID-19 outbreak size in Iran on the basis of known exported case counts and air travel links between Iran and other countries, and to anticipate where infections originating in Iran may spread next. Methods: We assessed interconnectivity between Iran and other countries by using direct and total traveler volumes and final destination cities of travelers originating in Iran in February 2019, according to data from the International Air Transport Association (accounting for 90% of global air travel, with the other 10% modeled by using market intelligence). Because exported cases were identified in United Arab Emirates (UAE), Lebanon, and Canada, we used the methods of Fraser and colleagues (3) to estimate the size of the underlying epidemic in Iran that would be needed for these cases to be observed with a reasonable probability. To estimate the time at risk for COVID-19 exposure among travelers departing Iran, we obtained data from the United Nations World Tourism Organization for the proportion of international travelers who are residents of Iran (4) and the average length of stay of tourists to Iran (5), and assumed that the Iranian outbreak began in early January 2020. We evaluated the relationship between the strength of travel links with Iran and the ranking of destination countries on the Infectious Disease Vulnerability Index (IDVI), a validated metric that estimates the capacity of a country to respond to an infectious disease outbreak. Scores range from 0 to 1, with higher scores reflecting greater capacity to manage infectious outbreaks.

inks with Iran and the ranking of destination countries on the Infectious Disease Vulnerability Index (IDVI), a validated metric that estimates the capacity of a country to respond to an infectious disease outbreak. Scores range from 0 to 1, with higher scores reflecting greater capacity to manage infectious outbreaks. Findings: A total of 212 000 persons traveled from Iranian airports (Tehran, Rasht, and Arak) to international destinations in February 2019. Although Qom has reported COVID-19 cases, its international airport is still under construction. Global cities receiving the greatest number of total travelers from Iran during this period include Istanbul, Turkey (n = 46 550); Najaf, Iraq (n = 24 659); and Dubai, UAE (n = 16 340). Among the top 10 traveler-receiving cities, 4 (Najaf, Baghdad, Damascus, and Baku) are in countries with an IDVI score lower than 0.6, suggesting elevated vulnerability to infectious disease outbreaks as well as limited ability to detect cases (Figure 1). Figure 1. Top 20 international cities connected to Iran by commercial air travel and associated vulnerability to infectious disease outbreaks.

Findings: A total of 212 000 persons traveled from Iranian airports (Tehran, Rasht, and Arak) to international destinations in February 2019. Although Qom has reported COVID-19 cases, its international airport is still under construction. Global cities receiving the greatest number of total travelers from Iran during this period include Istanbul, Turkey (n = 46 550); Najaf, Iraq (n = 24 659); and Dubai, UAE (n = 16 340). Among the top 10 traveler-receiving cities, 4 (Najaf, Baghdad, Damascus, and Baku) are in countries with an IDVI score lower than 0.6, suggesting elevated vulnerability to infectious disease outbreaks as well as limited ability to detect cases (Figure 1). Figure 1. Top 20 international cities connected to Iran by commercial air travel and associated vulnerability to infectious disease outbreaks. Vulnerability is measured at the country level by using the IDVI score, with a lower value indicating reduced capacity to respond to outbreaks. Countries with the lowest IDVI scores are indicated in green. The top 20 cities accounted for 70% of international outbound traveler volumes from Iran in February 2019. The first and 20th ranked cities, Istanbul and Milan, had 46 550 and 2500 outbound passengers, respectively, during this period. IDVI = Infectious Disease Vulnerability Index.

t IDVI scores are indicated in green. The top 20 cities accounted for 70% of international outbound traveler volumes from Iran in February 2019. The first and 20th ranked cities, Istanbul and Milan, had 46 550 and 2500 outbound passengers, respectively, during this period. IDVI = Infectious Disease Vulnerability Index. United Arab Emirates, Lebanon, and Canada ranked third, 21st, and 31st, respectively, for outbound air travel volume from Iran in February 2019. We estimated that 18 300 COVID-19 cases (95% CI, 3770 to 53 470 cases) would have had to occur in Iran, assuming an outbreak duration of 1.5 months in the country, in order to observe these 3 internationally exported cases reported at the time of writing.

vely, for outbound air travel volume from Iran in February 2019. We estimated that 18 300 COVID-19 cases (95% CI, 3770 to 53 470 cases) would have had to occur in Iran, assuming an outbreak duration of 1.5 months in the country, in order to observe these 3 internationally exported cases reported at the time of writing. Given the low rankings for Lebanon and Canada for outbound air travel, it is unlikely that cases would be identified in these countries and not in Iraq, Syria, or Azerbaijan (countries with higher travel volumes but low IDVI scores). Considering traveler volume alone, the odds of a single case being imported into Iraq rather than Canada or Lebanon would be 33.6 to 1 and 15.4 to 1 respectively; for Azerbaijan, the odds would be 3.8 to 1 and 1.7 to 1, respectively; and for Syria, the odds would be 3.7 to 1 and 1.7 to 1, respectively. As such, we performed exploratory analyses in which we assumed that an unidentified exported case of COVID-19 was present in Iraq, Syria, Azerbaijan, or all 3 countries, in addition to Lebanon, Canada, and UAE, and estimated the outbreak size in Iran that would produce these results (Figure 2). We also evaluated a scenario in which we assumed perfect case detection in travelers from Iran, such that disease is truly absent in countries not reporting cases. Under this “best-case” scenario, the estimated outbreak size in Iran was smaller but still substantial (1820 cases [CI, 380 to 5320 cases]). Figure 2. Estimated outbreak size in Iran required to observe exported cases internationally.

Given the low rankings for Lebanon and Canada for outbound air travel, it is unlikely that cases would be identified in these countries and not in Iraq, Syria, or Azerbaijan (countries with higher travel volumes but low IDVI scores). Considering traveler volume alone, the odds of a single case being imported into Iraq rather than Canada or Lebanon would be 33.6 to 1 and 15.4 to 1 respectively; for Azerbaijan, the odds would be 3.8 to 1 and 1.7 to 1, respectively; and for Syria, the odds would be 3.7 to 1 and 1.7 to 1, respectively. As such, we performed exploratory analyses in which we assumed that an unidentified exported case of COVID-19 was present in Iraq, Syria, Azerbaijan, or all 3 countries, in addition to Lebanon, Canada, and UAE, and estimated the outbreak size in Iran that would produce these results (Figure 2). We also evaluated a scenario in which we assumed perfect case detection in travelers from Iran, such that disease is truly absent in countries not reporting cases. Under this “best-case” scenario, the estimated outbreak size in Iran was smaller but still substantial (1820 cases [CI, 380 to 5320 cases]). Figure 2. Estimated outbreak size in Iran required to observe exported cases internationally. The estimated cumulative number of COVID-19 cases in Iran required to observe 3 cases exported to UAE, LBN, and CAN is shown in green. We also estimated the outbreak size required under alternate scenarios, including no additional exported cases to any other international destinations despite perfect case detection and 1 additional exported case to IRQ, AZE, or SYR (independently or to all 3 countries). Mean and 95% CIs are presented. The rate at which persons become infected while in Iran was assumed to be the same for residents and visitors. The rate of infection among air passengers (λ) was estimated as number of exported cases ÷ person-time at risk while in Iran. Person-time at risk was calculated as number of outbound air passengers × (average length of stay for visitors × proportion of air passengers who are visitors + outbreak duration × proportion of air passengers who are residents of Iran). Outbreak size in Iran was then estimated as λ × population size of Iran × outbreak duration. AZE = Azerbaijan; CAN = Canada; COVID-19 = coronavirus disease 2019; IRQ = Iraq; LBN = Lebanon; SYR = Syria; UAE = United Arab Emirates.

re visitors + outbreak duration × proportion of air passengers who are residents of Iran). Outbreak size in Iran was then estimated as λ × population size of Iran × outbreak duration. AZE = Azerbaijan; CAN = Canada; COVID-19 = coronavirus disease 2019; IRQ = Iraq; LBN = Lebanon; SYR = Syria; UAE = United Arab Emirates. Discussion: Given the low volumes of air travel to countries with identified cases of COVID-19 originating in Iran (such as Canada), Iran probably is currently experiencing a COVID-19 epidemic of substantial size for such exportations to be occurring. Our analysis would be modified by travel restrictions from Iran due to the recent political situation and by variations in the R0 value. Further, the lack of identified COVID-19 cases in countries with far closer travel ties to Iran suggests that cases in these countries are probably being missed rather than being truly absent. This is concerning, both for public health in Iran itself and because of the high likelihood for outward dissemination of the disease to neighboring countries with lower capacity to respond to infectious disease epidemics. Supporting capacity for public health initiatives in the region is urgently needed. This article was published at Annals.org on 16 March 2020. Note: Drs. Tuite and Bogoch contributed equally to this work. Grant Support: By grant 02179-000 from the Canadian Institutes of Health Research. Disclosures: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M20-0696.

Discussion: Given the low volumes of air travel to countries with identified cases of COVID-19 originating in Iran (such as Canada), Iran probably is currently experiencing a COVID-19 epidemic of substantial size for such exportations to be occurring. Our analysis would be modified by travel restrictions from Iran due to the recent political situation and by variations in the R0 value. Further, the lack of identified COVID-19 cases in countries with far closer travel ties to Iran suggests that cases in these countries are probably being missed rather than being truly absent. This is concerning, both for public health in Iran itself and because of the high likelihood for outward dissemination of the disease to neighboring countries with lower capacity to respond to infectious disease epidemics. Supporting capacity for public health initiatives in the region is urgently needed. This article was published at Annals.org on 16 March 2020. Note: Drs. Tuite and Bogoch contributed equally to this work. Grant Support: By grant 02179-000 from the Canadian Institutes of Health Research. Disclosures: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M20-0696. Reproducible Research Statement: Study protocol: Not applicable. Statistical code: Available from Dr. Tuite (ashleigh.tuite@utoronto.ca). Data set: Available from Dr. Khan (kamran.khan@unityhealth.to). Previous Posting: This manuscript was posted as a preprint on medRxiv on 25 February 2020. doi:10.1101/2020.02.24.20027375

“…make them believe, that offensive operations, often times, is the surest, if not the only (in some cases) means of defence.” —George Washington (1799)Coronavirus disease 2019 (COVID-19) is on the verge of being declared a pandemic. As of 7 March 2020, a total of 423 cases and 19 deaths, including several non–travel-related cases, areas of sustained community transmission, and a nursing home outbreak, have been reported (1). Best-case estimates suggest that COVID-19 will stress bed capacity, equipment, and health care personnel in U.S. hospitals in ways not previously experienced (2). How can health systems prepare to care for a large influx of patients with this disease? Develop a Strategy for Patient Volume and Complexity Approximately 95 000 critical care beds, including surgical and specialty unit beds, are available in U.S. hospitals today (3). Conservative estimates suggest that we may need almost twice this amount should the COVID-19 pandemic resemble the influenza pandemics of 1957 or 1968, especially if it is sustained (4). Because some patients will be critically ill and need scarce resources, such as extracorporeal membrane oxygenation and ventilators (5), hospitals must prepare now for how they will triage patients, allocate resources, and staff wards. The Table lists the essential elements of a hospital's planning process.

tained (4). Because some patients will be critically ill and need scarce resources, such as extracorporeal membrane oxygenation and ventilators (5), hospitals must prepare now for how they will triage patients, allocate resources, and staff wards. The Table lists the essential elements of a hospital's planning process. Table. Essential Components of a Hospital Preparedness Plan for COVID-19 Hospitals should attempt to geographically cohort patients with COVID-19 to limit the number of health care personnel exposed and conserve supplies. This type of geographic capacity generation is extremely difficult because many U.S. hospitals run at full capacity. Geographic cohorting options may also be challenged by locations of airborne isolation rooms, with negative pressure being scattered throughout the hospital. It may be necessary to use innovative approaches, such as converting single rooms to double occupancy; expediting discharges; slowing admission rates; and converting spaces like catheterization laboratories, lobbies, postoperative care units, or waiting rooms into patient care venues. For example, at Michigan Medicine, designated beds in critical care units and non–critical care settings for persons under investigation and patients who test positive for COVID-19 have been identified. A dedicated team of hospitalists and critical care providers has been established, with clinical schedules and roles for leadership, communication, and activation criteria. Contingency plans have been developed, including activation criteria for opening a respiratory intensive care floor where cohorting of both critically ill and noncritically ill patients can occur. Similarly, ensuring the ongoing care of vulnerable patients, such as those in the posttransplant and immunocompromised communities, remains imperative. Safe locations and staffing plans that separate vulnerable patients from COVID-19 activities have been carefully considered.

l and noncritically ill patients can occur. Similarly, ensuring the ongoing care of vulnerable patients, such as those in the posttransplant and immunocompromised communities, remains imperative. Safe locations and staffing plans that separate vulnerable patients from COVID-19 activities have been carefully considered. Protect and Support Health Care Workers on the Front Lines The best evidence currently available suggests that COVID-19 spreads primarily via droplet transmission and direct contact. With the appropriate precautions, nosocomial transmission can be mitigated. Health care personnel should receive training on proper donning and doffing of personal protective equipment, including fit testing of N95 masks and use of powered air-purifying respirators, as well as basic infection prevention tenets, such as hand hygiene. Hospitals should monitor rates of equipment use to ensure an adequate supply of personal protective equipment for those on the front lines and may need to engage hospital security to avoid theft or hoarding of such equipment. Extended use or limited reuse of N95 respirators may become necessary, and communication about preservation is important.

tes of equipment use to ensure an adequate supply of personal protective equipment for those on the front lines and may need to engage hospital security to avoid theft or hoarding of such equipment. Extended use or limited reuse of N95 respirators may become necessary, and communication about preservation is important. To limit the total number of personnel engaged in patient care, hospitals should institute overtime and extended hours with appropriate compensation strategies. Clear exposure criteria with detailed plans outlining management of personnel in regard to work restrictions or other quarantine requirements must be developed. Hospitals must also safeguard their own by keeping logs of staff who care for patients and monitoring them for signs or symptoms of infection. Finally, even if care of patients with COVID-19 will be provided by a subset of providers, it is important not to lose sight of the needs of their family members and other staff. Support is important to the morale and well-being of the workforce.

care for patients and monitoring them for signs or symptoms of infection. Finally, even if care of patients with COVID-19 will be provided by a subset of providers, it is important not to lose sight of the needs of their family members and other staff. Support is important to the morale and well-being of the workforce. Define a Strategy to Allocate Health Care Resources During crises, health care resources should be allocated in an ethical, rational, and structured way to do the greatest good for the greatest number of patients. Hospitals and health systems must set aside a “business as usual” mentality and focus on how best to accommodate the patients likely to benefit the most from care. Specifically, a plan that outlines what services and types of procedures will be provided (for example, extracorporeal membrane oxygenation) and what will not (for example, elective cases) must be developed. Accordingly, clinical guidelines for use (or denial) of scarce services, such as mechanical ventilation and critical care, should be outlined, in consultation with ethics and medical staff. A protocol that defines how patients will be triaged for admission, observation, early discharge, and quarantine is important. Hospitals should anticipate that normal staffing ratios and some standards of care are unlikely to be maintained; plans for contingency and crisis standards that lay out a legal and ethical framework for care decisions, including who will make decisions, how, and under what circumstances, must be readied. At Michigan Medicine, scarce resource guidelines have not only been developed, but portions have been revised and circulated to staff to ensure agreement and buy-in for execution.

out a legal and ethical framework for care decisions, including who will make decisions, how, and under what circumstances, must be readied. At Michigan Medicine, scarce resource guidelines have not only been developed, but portions have been revised and circulated to staff to ensure agreement and buy-in for execution. Develop a Robust, Transparent, and Open Communication Policy Hospitals and health systems must develop agile ways to transmit timely and critical information in times of crises. A designated communication team that is integrated into the work so they have a strong understanding of the clinical care being provided and the communication needs of the workforce, patients, and public is recommended. Crisis communications should ideally occur via several media, such as a telephone hotline, the hospital Web page, social media platforms, or text-based messages. Important metrics, including the number of cases being triaged, investigated, or managed; bed capacity and availability; and new or emerging data on treatments or care strategies, should be provided. Similarly, timely communication of national updates on travel restrictions, policies for self-monitoring and quarantine, and trends in infection rates must occur. To this end, health care leaders must remember that patients and their families are as much in need of actionable information as hospital personnel. Comprehensive communication strategies for both internal and external stakeholders are key.

s for self-monitoring and quarantine, and trends in infection rates must occur. To this end, health care leaders must remember that patients and their families are as much in need of actionable information as hospital personnel. Comprehensive communication strategies for both internal and external stakeholders are key. The COVID-19 outbreak will test the resilience of our health care system. Planning for managing patients and our workforce must begin in full force. This article was published at Annals.org on 11 March 2020. Disclosures: Authors have disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M20-0907. Corresponding Author: Vineet Chopra MD, MSc, University of Michigan, 2800 Plymouth Road, Building 16 #432W, Ann Arbor, MI 48109; e-mail, vineetc@umich.edu. Current Author Addresses: Dr. Chopra: University of Michigan, 2800 Plymouth Road, Building 16 #432W, Ann Arbor, MI 48109. Dr. Toner: Center for Health Security, Bloomberg School of Public Health, Johns Hopkins University, 621 East Pratt Street, Baltimore, MD 21202. Dr. Waldhorn: Georgetown University, 3800 Reservoir Road, Washington, DC 20007. Dr. Washer: University of Michigan, F4151 University Hospital South, Ann Arbor, MI 48109. Author Contributions: Conception and design: V. Chopra, E. Toner, R. Waldhorn, L. Washer. Analysis and interpretation of the data: E. Toner. Drafting of the article: V. Chopra, E. Toner, R. Waldhorn, L. Washer. Critical revision of the article for important intellectual content: V. Chopra, E. Toner, R. Waldhorn, L. Washer.

Dr. Washer: University of Michigan, F4151 University Hospital South, Ann Arbor, MI 48109. Author Contributions: Conception and design: V. Chopra, E. Toner, R. Waldhorn, L. Washer. Analysis and interpretation of the data: E. Toner. Drafting of the article: V. Chopra, E. Toner, R. Waldhorn, L. Washer. Critical revision of the article for important intellectual content: V. Chopra, E. Toner, R. Waldhorn, L. Washer. Final approval of the article: V. Chopra, E. Toner, R. Waldhorn, L. Washer. Administrative, technical, or logistic support: V. Chopra. Collection and assembly of data: V. Chopra, E. Toner.

Background: The behavior of the general public will probably have an important bearing on the course of the coronavirus disease 2019 (COVID-19) epidemic. Human behavior is influenced by people's knowledge and perceptions (1). Objective: To assess knowledge and perceptions about COVID-19 among a convenience sample of the general public in the United States and United Kingdom. Methods and Findings: This study is a cross-sectional survey conducted on an online platform managed by Prolific Academic Ltd. The platform's pool of participants numbers approximately 80 000 individuals, of whom approximately 43% reside in the United Kingdom and 33% in the United States (2). For this study, Prolific selected a convenience sample of 3000 participants residing in the United States and 3000 participants residing in the United Kingdom who were chosen to have approximately the same distribution of age, sex, and ethnicity (and each combination thereof) as the U.S. and U.K. general population (by using numbers from the last census in each country). Specifically, Prolific established population strata (Table 1) with a predetermined number of open slots into which eligible participants in the online pool could enroll on a first-come, first-served basis.

ination thereof) as the U.S. and U.K. general population (by using numbers from the last census in each country). Specifically, Prolific established population strata (Table 1) with a predetermined number of open slots into which eligible participants in the online pool could enroll on a first-come, first-served basis. Table 1. Sample Characteristics Participants, who had to have indicated that they were fluent in English, received US$1.50 for completing the survey. They completed the online questionnaire between 23 February and 2 March 2020. The questionnaire (Supplement, available at Annals.org) consisted of 22 questions on knowledge and perceptions of COVID-19, including specific questions about “myths” or falsehoods listed on the World Health Organization's “myth busters” Web site (3). Supplement. Questionnaire Click here for additional data file. To summarize the survey findings, I dichotomized categorical variables and computed the median and interquartile range for continuous variables. For binomial proportions, I used a score interval (Wilson score interval without continuity correction [4]) to construct a 95% CI. No sampling weights were used given that this was not a probabilistic sample.

I dichotomized categorical variables and computed the median and interquartile range for continuous variables. For binomial proportions, I used a score interval (Wilson score interval without continuity correction [4]) to construct a 95% CI. No sampling weights were used given that this was not a probabilistic sample. In total, 2986 and 2988 adults residing in the United States and United Kingdom, respectively, completed the questionnaire. Participants' sociodemographic characteristics are shown in Table 1. Although participants generally had good knowledge of the main mode of disease transmission and common symptoms, the survey identified several important misconceptions on how to prevent acquisition of COVID-19, including beliefs in falsehoods that have circulated on social media (Table 2). A substantial proportion of participants also expressed an intent to discriminate against individuals of East Asian ethnicity for fear of acquiring COVID-19. A more detailed analysis and visualization of all survey responses are available (5).

luding beliefs in falsehoods that have circulated on social media (Table 2). A substantial proportion of participants also expressed an intent to discriminate against individuals of East Asian ethnicity for fear of acquiring COVID-19. A more detailed analysis and visualization of all survey responses are available (5). Table 2. Summary of Survey Findings Discussion: The findings of this study could be used to set priorities in information campaigns on COVID-19 by public health authorities and the media. Such information provision could, for instance, emphasize the comparatively low case-fatality rate, the recommended care-seeking behavior, the low risk posed by individuals of East Asian ethnicity living in the United States and United Kingdom, and that children appear to be at a lower risk for a fatal disease course than adults. In addition, to ensure that individuals focus their attention on those prevention measures that are most effective, this study suggests that it will be important to inform the public about the comparative effectiveness of common surgical masks versus frequent and thorough handwashing and avoiding close contact with people who are sick.

nsure that individuals focus their attention on those prevention measures that are most effective, this study suggests that it will be important to inform the public about the comparative effectiveness of common surgical masks versus frequent and thorough handwashing and avoiding close contact with people who are sick. This study has several limitations. First and foremost, given that participants had to have both chosen to register with Prolific and to take the survey at the time it was published, this convenience sample of adults is unlikely to be representative of the general U.S. and U.K. population. The generalizability of the findings is, therefore, limited. Second, it is possible that some participants may have randomly selected responses to spend the least amount of time to earn the $1.50 reward. I believe this is unlikely to be an important source of bias because only 2 participants (who were excluded from the analysis) completed the survey in under 2 minutes (while it was physically possible to complete it in well under 90 seconds), there was no bimodal distribution in the time taken to complete the survey, and $1.50 is a relatively small monetary incentive. Third, it is possible that participants looked up the answers to some of the questions online before answering. Participants were asked at the end of the survey (while being reassured that their payment is not influenced by their response) for which, if any, questions they searched for an answer online. These responses were set to missing in the analysis.

ts looked up the answers to some of the questions online before answering. Participants were asked at the end of the survey (while being reassured that their payment is not influenced by their response) for which, if any, questions they searched for an answer online. These responses were set to missing in the analysis. In conclusion, the general public in the United States and United Kingdom appears to have important misconceptions about COVID-19. Correcting these misconceptions should be targeted in information campaigns organized by government agencies, information provision by clinicians to their patients, and media coverage. This article was published at Annals.org on 20 March 2020. Disclosures: The author has disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M20-0912. Reproducible Research Statement: Study protocol: Available from Dr. Geldsetzer (e-mail, pgeldsetzer@stanford.edu). Statistical code and data set: Available from https://purl.stanford.edu/zx357dw0759. Previous Posting: This manuscript was posted as a preprint on medRxiv on 17 March 2020. doi:10.1101/2020.03.13.20035568

The coronavirus disease 2019 (COVID-19) pandemic has upended clinicians' sense of order and control. Such disruption may lead to substantial stress in the short term and higher risk for burnout over the long term. While natural disasters, such as Hurricane Katrina, demonstrated the effectiveness of short-term emergency planning (1), the COVID-19 pandemic poses unique long-term stressors and risks to clinicians' physical, mental, spiritual, and emotional well-being. Leaders and front-line clinicians need to proactively protect the well-being of themselves and their colleagues to avoid adverse outcomes for clinicians and adverse effects on quality of patient care (2). We provide practical suggestions to encourage a culture that will sustain the clinician workforce during the pandemic. Regardless of practice location or size, everyone must commit to supporting the well-being of those involved in patient care.

outcomes for clinicians and adverse effects on quality of patient care (2). We provide practical suggestions to encourage a culture that will sustain the clinician workforce during the pandemic. Regardless of practice location or size, everyone must commit to supporting the well-being of those involved in patient care. First and foremost, organizational leaders should provide clear messages that clinicians are valued and that managing the pandemic together is the goal. Front-line clinicians must individually and collectively identify concerns that arise while facing the reality of the pandemic. Leaders must communicate current best practices clearly and compassionately, manage expectations, clarify work hours, and provide sufficient resources and effective personal protective equipment. To better enable clinicians to maintain personal well-being and resilience throughout the pandemic, leaders should aim to monitor clinician wellness and proactively address concerns related to the safety of clinicians and their families.

urs, and provide sufficient resources and effective personal protective equipment. To better enable clinicians to maintain personal well-being and resilience throughout the pandemic, leaders should aim to monitor clinician wellness and proactively address concerns related to the safety of clinicians and their families. Leaders should aim for work schedules that promote physical resilience by enabling adequate sleep and providing access to call rooms for hospital-based clinicians working long or multiple shifts. Leaders should also take initiatives to provide basic provisions during work hours, such as easy access to water, healthy snacks, chargers for phones and other devices, and toiletries. Leaders must also designate times for clinicians to take breaks, eat, and take medications. It may also be helpful to advise clinicians working such shifts to bring at least 3 days of their own medications to work and designate a source for emergency refills. Clinicians should also continue using wellness activities that have worked for them in the past and make efforts to support each other during this challenging time.

be helpful to advise clinicians working such shifts to bring at least 3 days of their own medications to work and designate a source for emergency refills. Clinicians should also continue using wellness activities that have worked for them in the past and make efforts to support each other during this challenging time. Reduction of noncritical work activities may help to promote mental well-being. Examples include rescheduling preventive and routine patient follow-up visits and eliminating nonessential administrative tasks. Anxiety can be reduced by providing a central source for updated information and clear communication of well-defined protocols, expectations, and such resources as childcare via e-mails, tweets, and automated calls. When an individual clinician feels well but cannot be present in the clinical setting because of mandatory isolation or childcare, hospitals and practices should aim to redistribute work and have these clinicians participate in computer- and phone-based care while home. During the pandemic, clinicians should be encouraged to openly discuss vulnerability and the importance of protecting one's emotional strength. Health care organizations can provide information on managing stress, reducing burnout, and identifying mental health professionals available to support clinicians (3). Deploy designated wellness champions in health care systems and practices to field clinicians' concerns, advocate for clinicians, and distribute messages of gratitude and support.

n provide information on managing stress, reducing burnout, and identifying mental health professionals available to support clinicians (3). Deploy designated wellness champions in health care systems and practices to field clinicians' concerns, advocate for clinicians, and distribute messages of gratitude and support. We also suggest fostering spiritual resilience through distribution of positive messaging that emphasizes appreciation for clinicians' dedication and altruism. Disseminating strategies for connecting with colleagues to share stories of success, rather than focusing on failures and stresses, can help clinicians find joy amidst chaos (4). Helping clinicians recognize what they can and cannot control helps to balance expectations with realities. A supportive work culture is vital to maintaining the resilience of clinicians during a crisis such as COVID-19. We suggest developing an evidence-based menu of interventions, to be carefully selected from, and tailored to various workplace settings. For larger health systems, wellness committees and employee assistance programs are the logical resources to organize these interventions. In smaller settings, appointing a wellness champion could help to elucidate colleagues' needs and implement solutions. Surveys to assess stress points, fears, and concerns can inform leaders and provide insight into areas requiring attention. We also suggest developing plans to back up, cross-train, and rotate leadership to avoid leader burnout.

ting a wellness champion could help to elucidate colleagues' needs and implement solutions. Surveys to assess stress points, fears, and concerns can inform leaders and provide insight into areas requiring attention. We also suggest developing plans to back up, cross-train, and rotate leadership to avoid leader burnout. Sharing challenges and successes will help to meet urgent needs during the evolving pandemic. Examples of settings for such sharing include the American College of Physicians Physician Well-Being and Discussion Forum (5), the Society of General Internal Medicine GIMConnect (6), and the American Medical Association Physician Health (7) resources that members can access. Other professional organizations, or organizations with access to community discussion boards, could develop similar venues for highlighting best practices in wellness. Emphasizing clinician wellness during the COVID-19 pandemic (8) is necessary to enable them to provide high-quality care. We propose some preliminary, common sense steps toward this goal and encourage colleagues to share strategies they find successful. How we meet the wellness needs of our clinicians may determine how well we survive the COVID-19 pandemic and future public health crises. This article was published at Annals.org on 20 March 2020.

Emphasizing clinician wellness during the COVID-19 pandemic (8) is necessary to enable them to provide high-quality care. We propose some preliminary, common sense steps toward this goal and encourage colleagues to share strategies they find successful. How we meet the wellness needs of our clinicians may determine how well we survive the COVID-19 pandemic and future public health crises. This article was published at Annals.org on 20 March 2020. Disclosures: Dr. Linzer reports grants from the American Medical Association and the American College of Physicians outside the submitted work. Authors not named here have disclosed no conflicts of interest. Disclosures can also be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M20-1033. Corresponding Author: Charlene M. Dewey, MD, MEd, Center for Professional Health, Vanderbilt University School of Medicine, 1107 Oxford House, Nashville, TN 37232-4300; e-mail, Charlene.dewey@vumc.org. Current Author Addresses: Dr. Dewey: Center for Professional Health, Vanderbilt University School of Medicine, 1107 Oxford House, Nashville, TN 37232-4300. Dr. Hingle: Southern Illinois University School of Medicine, 913 North Rutledge cHOP, Mailcode 9623, Springfield, IL 62794-9623. Dr. Goelz: Hennepin Healthcare, 701 Park Avenue, (P5), Minneapolis, MN 55415. Dr. Linzer: Institute for Professional Worklife, Hennepin Healthcare, 701 Park Avenue, (G5), Minneapolis, MN 55415. Author Contributions: Conception and design: C. Dewey, S. Hingle, E. Goelz. Drafting of the article: C. Dewey, S. Hingle, M. Linzer.

Dr. Hingle: Southern Illinois University School of Medicine, 913 North Rutledge cHOP, Mailcode 9623, Springfield, IL 62794-9623. Dr. Goelz: Hennepin Healthcare, 701 Park Avenue, (P5), Minneapolis, MN 55415. Dr. Linzer: Institute for Professional Worklife, Hennepin Healthcare, 701 Park Avenue, (G5), Minneapolis, MN 55415. Author Contributions: Conception and design: C. Dewey, S. Hingle, E. Goelz. Drafting of the article: C. Dewey, S. Hingle, M. Linzer. Critical revision of the article for important intellectual content: C. Dewey, S. Hingle, E. Goelz. Final approval of the article: C. Dewey, S. Hingle, E. Goelz, M. Linzer. Administrative, technical, or logistic support: C. Dewey, S. Hingle.