We studied all 16 patients with confirmed COVID-19 released from the treatment center of People’s Liberation Army General Hospital in Beijing, China, between January 28 and February 9, 2020. On alternate days, all patients had throat swabs collected, which were then analyzed. Patients were discharged after their recovery and confirmation of “virus-negative” status by at least two consecutive real-time PCRs (3). There was only one case of a false-negative result in our study: patient 6 had a negative test result followed by a positive detection and then two consecutive negative tests. Travel and possible exposure history were obtained from the patients and noted on their records. Epidemiologically, 10 patients visited Wuhan after the outbreak; 3 had exposure to a known infected patient; 2 came in contact with people from Wuhan; and 1 had no known exposure. The basic clinical characteristics are given in Table 1. The median age was 35.5 years (range, 3–68 yr), with 11 of 16 being male. The major symptoms in these patients were fever (14 of 16), cough (11 of 16), pharyngalgia (5 of 16), and dyspnea (2 of 16). The day of onset and resolution of these symptoms were noted. Details of symptoms are indicated in the online supplement.

Ground-glass opacities were observed by computed tomography of the chest in both sides of the lungs in six patients and only in the right lung in one patient. Concentrations of C-reactive protein and procalcitonin between the first sample obtained at the time of hospitalization and the last sample obtained before discharge were comparable (Table 1).

All the patients received various medical care to treat COVID-19. Fifteen patients were treated with IFN-α together with other antiviral drugs, including oseltamivir (1 of 16), lopinavir/ritonavir (11 of 16), acyclovir (1 of 16), moxifloxacin (5 of 16), methylprednisolone (2 of 16), γ-globulin (2 of 16), vancomycin (1 of 16), and meropenem (1 of 16), either alone or in combination. Only one patient required respiratory support involving mechanical ventilation.

The time kinetics of symptom onset, duration of symptoms, and viral clearance is described in Table 1. The viral detection test was performed upon clinical presentation and repeated every other day until the patient had a negative test result. The negative test result was confirmed again the next day. Upon confirmation of the negative test result, the patient was asked to quarantine at home for the next 2 weeks, with a follow-up visit to the hospital after 1 week to confirm virus-negative status. The incubation periods were estimated on the basis of the history of the patient’s travel or potential exposure. Our data show an incubation period of 5 days (interquartile range, 1–6 d) among the patients (except for patient 12, who had no specific exposure). The mean duration of symptoms was estimated to be 8 days (interquartile range, 6.25–11.5). Most important, half (8 of 16) of the patients remained virus positive (a surrogate marker of shedding) even after the resolution of symptoms (median, 2.5 d; range, 1–8 d). Some of our patients had other comorbidities, which included diabetes (2 of 16) and tuberculosis (1 of 16), both of which did not affect the time course of the disease. Similarly, the clinical course for the 3-year-old boy did not significantly differ from that of the rest of the patients.

The current COVID-19 pandemic is the third and most lethal outbreak of coronavirus in the 21st century (4), in which the number of infections and mortality have surpassed those of both Middle East respiratory syndrome and severe acute respiratory syndrome within a short period (1, 5). Although the infection appears to be milder, with the most lethality in the older male population with preexisting morbidities (3, 6), it is contagious. The ability to spread may arise from the ability of the virus to transmit from subclinical patients. Cases have been reported in which patients infected their close contacts even after “apparent recovery” from the infection (2). This warrants investigation of the “shedding window” after the clinical recovery of the patient. In this study, we report that half of the patients continued to be virus positive even after the resolution of symptoms up to 8 days (Figure 1). The viral clearance kinetics were similar in another study by Young and colleagues (12 d), in which all the patients survived the infection (7). In contrast, the virus persisted for 20 days in another study, which had a significant high mortality of >40% (8). This information can provide a useful tool for clinicians and policy makers to ensure that recovered patients do not spread the virus. It is important to note that all our patients had milder infections and recovered from the disease. However, it is currently unclear if there is a delayed viral clearance in the more vulnerable population, such as those who are older, have immunodeficiencies, or are receiving immunosuppressive therapies.

Figure 1. Time kinetics of viral presence in relationship with resolution of symptoms. Day 0 (D0) is the first day of symptoms, whereas the blue dots indicate the resolution of symptoms. The first orange box represents the day of first positive viral detection, and the last orange box indicates the day of first negative viral quantitative PCR. F = first time point at which laboratory blood tests were conducted during the hospitalization; L = last time point at which laboratory blood tests were conducted during the hospitalization.

Download Figure | Download Powerpoint

[More] [Minimize]

The current data are derived from mostly young and male subjects, which is consistent with our previous report in Beijing (9). Similarly to a recent study, we report the case of a child (3-yr-old boy) with COVID-19, indicating the ability of this virus to infect young children (10).

Our study is limited by the number of patients because there have been limited cases outside the epicenter of the coronavirus outbreak that have been successfully treated so far in which the patients were able to be released from the hospital. Our study provides initial insight into the viral clearance kinetics and the ability of the virus to persist even after the resolution for as long as 8 days, which may pose a significant challenge in controlling the spread of the disease. However, further studies are needed to investigate whether the real-time PCR–detected virus is capable of transmission at the later stage of the disease.