The Spectrum of Severe Acute Respiratory SyndromeAssociated Coronavirus Infection

Context Severe acute respiratory syndrome (SARS) is a rapidly progressive pneumonia, but do some infected patients have only mild or no symptoms? Contribution A total 1100 patients had serologic tests when evaluated at a SARS clinic in Hong Kong at the time of the 2003 epidemic. Of 910 patients who were managed without hospitalization, 6 patients had serologic evidence of SARS. Five of the 6 patients had normal chest radiographs, and 4 patients had symptoms such as myalgia, chills, coughing, and feeling feverish. Cautions Although this study suggests few cases of subclinical SARS during the 2003 outbreak, patients were self-referred rather than identified through community surveillance. The Editors Severe acute respiratory syndrome (SARS) is a rapidly progressive, atypical pneumonia (1-3) that is caused by a novel coronavirus (4-8) and has spread throughout the globe (9). The epidemiology (10), clinical features (11-15), and progression (16) of patients with confirmed SARS have recently been described and have shown a spectrum of clinical presentation and severity of disease (16). However, published data are insufficient to clarify whether patients have subclinical and milder forms of SARS that do not result in pneumonia. Strategies were devised for screening patients who attended our emergency department (15, 17) and for clinically confirming SARS in the Hong Kong outbreak (13, 18), but these strategies have never been evaluated by using a gold standard diagnostic test for SARS coronavirus infection. Virus isolation, molecular detection for viral RNA, and serologic tests have recently been developed for detecting SARS coronavirus infection (4-8). Among these, serologic testing on paired serum samples is regarded as both sensitive and specific (19). We studied the spectrum of clinical presentation in patients screened at a SARS screening clinic with paired serologic test results, investigated whether a subset of mildly symptomatic or asymptomatic patients exists, and evaluated the accuracy of clinical assessment of SARS. Methods Study Design and Setting This study was conducted in the newly opened SARS clinic located in the emergency department of the Prince of Wales Hospital, Hong Kong. This teaching hospital and tertiary referral center for the New Territories of the Hong Kong Special Administrative Region serves a mainly urban catchment area of 1.5 million people. Because of the severity of the crisis, pursuing normal channels of approval and consent was not possible. However, all patients received a questionnaire to complete, and they understood that the data would be used for surveillance, analysis, and research. The chief executives and associated institutional review boards of the Hospital Authority of Hong Kong and Prince of Wales Hospital approved the collection of clinical data and use of blood tests for surveillance, analysis, research, and reporting. Many of these patients have been described in previous reports (13, 15). Patients were included if they attended the clinic between 12 March 2003 and 12 May 2003. People would attend the emergency department of their own volition as per normal practice and were triaged either to fever or nonfever areas depending on whether they reported fever or had a temperature greater than 37 C. The SARS clinic was operated by emergency physicians and nurses. The clinic was initially located in the emergency department but was later moved outside the main building. All staff or patients from the community catchment area with suspected SARS were referred to the clinic. There are 14 hospitals with emergency departments in Hong Kong, and patients with suspected SARS would present to and be hospitalized in all of these during the outbreak. The Prince of Wales Hospital SARS clinic received a high proportion of hospital staff; thus, this study may not fully represent the spectrum in the community. We previously described in detail the process of care through the clinic (15). Patients were followed up daily after first attendance if they had a contact history, were symptomatic, had a body temperature greater than 38 C, had a normal or indeterminate chest radiograph, or had abnormal results on laboratory investigations (for example, leukopenia, lymphopenia, monocytosis, or thrombocytosis). Patients were given hygiene advice, advised to isolate themselves, and followed up the next day. Patients were hospitalized to a clearing ward if they met the case definition for probable SARS. Patients were discharged from the clinic after first attendance if they were asymptomatic, did not have a fever, and had a normal chest radiograph and normal results on laboratory investigations. They were given hygiene advice and advised to return if they experienced a fever. Patients who were followed up daily were discharged after 48 hours of remaining asymptomatic and afebrile and with normal chest radiographs and laboratory test results. Patient Assessments and Investigations All patients were seen by an emergency specialist; completed a health questionnaire that documented contact history and upper and lower respiratory tract, gastrointestinal, and systemic symptoms; and had pulse rate, systolic and diastolic blood pressure, respiratory rate, tympanic temperature, and oxygen saturation at room air recorded daily. All patients underwent daily frontal plain chest radiography until their symptoms subsided or pneumonic change appeared. If fever (temperature > 38 C) persisted for more than 2 days, standard and high-resolution computed tomography was requested for high-risk patients (those with contact history, persistent symptoms, and persistent fever) who had normal chest radiographs to confirm or exclude occult pneumonia. Chest radiographs were initially evaluated by a specialist emergency physician who had knowledge of clinical details and then by a radiologist without reference to clinical information. Case Definition and Viral Detection The World Health Organization (WHO) and Centers for Disease Control and Prevention (CDC) have described current case definitions for SARS (2, 3). In our study, patients with a probable case of SARS during the outbreak that would result in hospitalization included those with 1) a contact history; 2) systemic, respiratory, or gastrointestinal symptoms; and 3) pneumonic change on chest radiograph or computed tomography scan (Table 1). Table 1. Case Definitions for Severe Acute Respiratory Syndrome Before SARS coronavirus serologic tests were available, SARS was diagnosed clinically if the patient followed a typical 3-phase clinical course during the outbreak (Table 1). After SARS coronavirus serologic tests were available, SARS was confirmed by serologic evidence of SARS coronavirus (18). Phase 1 of the clinical course is a viral replication phase and involves an initial presentation of high fever and myalgia that last a few days and then resolve (symptoms resolve and temperature < 37 C). Phase 2 involves presumably immune-mediated lung injury and includes a recurrence of fever and progression of pneumonia. This phase begins about 8 days after fever onset. Most patients will improve after pulse steroid therapy and enter a third phase of rehabilitation. However, 20% may enter phase 3: severe lung injury characterized by the acute respiratory distress syndrome and need for ventilatory support (13, 16, 18). The level of anticoronavirus IgG antibody was measured by an immunofluorescence assay on the basis of Vero cells infected with coronavirus isolated from a patient with SARS (20). The full genome sequence of this SARS coronavirus strain is available at GenBank (GenBank accession number AY278554) (7, 8). According to the CDC, laboratory criteria for SARS may be used to classify cases as confirmed, negative, or undetermined status (3, 19). Severe acute respiratory syndrome is confirmed if any 1 of 3 criteria is met: 4-fold increase in SARS coronavirus antibody titer between acute-phase serum (measured within 7 days after the onset of fever) and convalescent-phase serum (measured >21 days after the onset of fever), detection of antibody to SARS coronavirus in specimens during acute illness, or detection of antibody to SARS coronavirus in specimens obtained from convalescent serum. Cases were classified as negative if there was no serologic evidence of SARS coronavirus, defined by the absence of antibody to SARS coronavirus in convalescent serum obtained more than 21 days after symptom onset (3, 18). Undetermined status includes patients in whom laboratory testing was either not performed or not complete (that is, discharged patients who declined to return to provide convalescent serum or patients who did not have serologic evidence of SARS coronavirus but had convalescent samples taken 21 days after onset of symptoms). For this study, convalescent or acute SARS coronavirus serologic status was considered the reference standard for SARS. Two experienced technicians cross-checked all results. The primary clinical outcome measure was a case with serologic evidence of SARS coronavirus. Controls were patients with no serologic evidence of SARS coronavirus. Cases of undetermined status are reported but are excluded in sensitivity analyses. Statistical Analysis Data are presented as means (SD) or absolute numbers (percentages), unless otherwise specified. Continuous variables were analyzed by using the unpaired 2-tailed t-test, and categorical data were analyzed by using the chi-square test or Fisher exact test. Data were analyzed by using Statview for Windows, version 5.0 (SAS Institute, Inc., Cary, North Carolina), and MedCalc, version 7.0 (MedCalc Software, Mariakerke, Belgium). Both SARS coronavirus serologic findings and clinical determination of SARS were binary (yes or no), categorical data. Sensitivity is the proportion of patients with the disease (serologic evidence of SARS coronavirus) and positive test results (that is, were hospitalized), while specificity is