Article Text

Download PDFPDF

Severe acute respiratory syndrome (SARS): epidemiology, diagnosis and management
  1. G W K Wong1,
  2. D S C Hui2
  1. 1Department of Paediatrics, Prince of Wales Hospital, Chinese University of Hong Kong, Hong Kong Special Administrative Region, China
  2. 2Department of Medicine and Therapeutics, Prince of Wales Hospital, Chinese University of Hong Kong
  1. Correspondence to:
    Dr G W K Wong, Department of Paediatrics, Prince of Wales Hospital, Shatin, New Territories, HKSAR, China;

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

SARS is a serious respiratory illness that frequently runs a rapidly progressive downhill course. In just 6 weeks it has spread to all continents of the globe. At the time of writing more than 7000 cases have been reported worldwide and over 500 have died. The primary mode of transmission appears to be by droplets. Good supportive care and the judicious use of ribavirin and steroids result in recovery in over 90% of patients, but randomised controlled trials are needed to define the roles of these treatments. Success in controlling the disease relies on early identification of suspect cases, proper isolation, and meticulous infection control measures. Development of sensitive and specific rapid diagnostic tests is underway.

Since February 2003 the world has been hit by a highly contagious respiratory infection which frequently results in rapidly progressive respiratory failure. In late 2002 and early 2003 there were reports of outbreaks of atypical pneumonia of unknown aetiology in Southern China. Initially the condition primarily affected close contacts of the patients and healthcare workers who looked after them. With increasing recognition of this unusual infection, the US Centers for Disease Control and Prevention termed the condition “severe acute respiratory syndrome” (SARS).1 Over the past few months the global medical community has worked together to achieve an unprecedented speed of progress in our understanding of SARS. Here we review the current knowledge of the epidemiology, clinical presentation, and treatment of this devastating condition.


The outbreak in Hong Kong began when an infected doctor from Southern China checked into a hotel on 21 February 2003. His symptoms of a respiratory tract infection had apparently started almost 1 week before his arrival in Hong Kong. He passed on the infection to eight key persons who had either stayed at the hotel as guests or who had visited friends at the hotel. These infected persons subsequently brought the infection back to their home countries and started the outbreaks in Canada, Singapore and Vietnam.2–5

A 26 year old man who had visited a friend in the hotel developed a febrile illness and was admitted to the Prince of Wales Hospital on 4 March 2003. Initially, he was found to have right upper lobe pneumonia which subsequently progressed to bilateral consolidation. In addition to intravenous antibiotics, he was also treated with nebulised salbutamol to improve his mucociliary clearance. He subsequently recovered without the use of any antiviral or steroid treatment. Six days after his admission 18 healthcare workers from the same ward reported having an acute febrile illness. Infection control investigations subsequently revealed that a total of 156 subjects were infected and they were admitted to hospital between 11 and 25 March with SARS.3 Included among them were 69 healthcare workers and 16 medical students who had examined the index case or the patients around him. The others were individuals who had visited other patients in the index ward. The sequence of events was much the same in Vietnam, Singapore, and Toronto.2,3 Unprotected healthcare workers are at the highest risk of infection; only about 5% of close family contacts of the initial cohort of patients developed SARS. This may be because healthcare workers were screened and admitted to hospital quickly once symptoms developed, and the infected individuals were probably not infectious during the incubation period.

The disaster that led to a major community outbreak in Hong Kong was caused by a 33 year old man with chronic renal disease on haemodialysis. He had also been admitted to the same index ward at the Prince of Wales Hospital during the same time period. Initially, his main symptom was diarrhoea. He had visited relatives at the Amoy Gardens Apartment complex a few times in late March where over 300 residents were infected. From the preliminary investigations the most likely route of spread in this outbreak was via leaky sewage pipes allowing an aerosol of infectious faecal material to escape into the narrow light well between the buildings and to spread in rising air currents. Furthermore, the floor drains of the bathrooms and kitchens are also connected to the sewage pipes, so backflow of contaminated aerosol into other apartments via these routes could also have played a role in the outbreak.

Although the primary mode of transmission is by close contact with contaminated droplets, preliminary studies suggest that the responsible viral agent can also be found in large quantities in urine and faeces from infected individuals. In the absence of a reliable and rapid laboratory test in many hospitals, the diagnosis of SARS is still based on clinical features. The case definition provided by the World Health Organization is updated periodically; the latest update was on 1 May 2003 when SARS was categorised into “suspect” and “probable” cases.4


Most patients present with an acute febrile illness after an incubation period of 5–8 days. The commonest symptoms at presentation are fever, myalgia, dry cough, headache and dizziness.3 The dizziness can be so severe that patients are unable to walk even a few steps. Productive cough and coryza are uncommon. Diarrhoea was found to be a more common symptom (over 50%) in patients from the Amoy Gardens. In a group of 20 patients with SARS from Singapore, dry cough has been reported to be very common (75%) while chills and rigours are relatively rare (15%).5 Some patients, especially the elderly, may not present with high fever and other cardinal symptoms. In many patients physical examination reveals a high swinging fever. Auscultation of the chest shows inspiratory crackles predominating in the lung base. For patients presenting with pneumonia who either live in a SARS affected area or have travelled to such an area, physicians should consider the possibility of SARS. A detailed contact history should be taken of any person with a severe respiratory infection. Patients with SARS should be treated under proper isolation and infection control measures to prevent a major outbreak in the hospital ward.

At the onset of fever about 20% of subjects will have normal chest radiographs, so a normal chest radiograph at presentation does not rule out the diagnosis. Careful follow up is necessary. Unilateral and bilateral or multifocal air space consolidations are frequently seen. These findings are similar to other types of bronchopneumonia. Pleural effusions and hilar lymphadenopathy are not usually present. In patients with normal chest radiographs, thoracic computed tomographic (CT) scans frequently show ill defined ground glass opacification, especially in the periphery of the lungs. Some of the CT findings are similar to those seen in patients with bronchiolitis obliterans organising pneumonia (BOOP).6

Although the laboratory findings of leukopenia (34%), lymphopenia (70%), and mild thrombocytopenia (45%) are non-specific, they provide clues to the possible diagnosis of SARS.3 Increased d-dimers and mild prolongation of activated partial thromboplastin time (45%) are found in about half the patients, in keeping with a mild picture of disseminated intravascular coagulation. Other biochemical abnormalities include increased levels of lactate dehydrogenase (LDH) and creatinine kinase of skeletal muscle origin.3

While some patients may have a mild course of pneumonia, many develop progressive dyspnoea and increasing oxygen dependency 6–8 days after the onset of the illness (table 1). In the initial stage many patients may appear to respond to oral steroids with resolution of the fever. However, in week 2 (second stage of the illness) the disease will frequently flare up again and they may respond to high dose pulse steroid therapy. It is thought that this deterioration (immune response phase) is caused by immunopathological dysregulation and uncontrolled activation of the cytokine system resulting in the observed lung damage. A proportion of patients will continue to deteriorate and develop acute respiratory distress syndrome (ARDS, third stage of the illness) requiring mechanical ventilatory support. Patients should therefore be carefully followed for 2–3 weeks before being discharged from hospital. About 25% of adult patients require intensive care, and 15% need mechanical ventilation. Our experience with younger children with SARS suggests that they are less severely affected, while adolescent patients behave very similarly to adults.7 At the time of writing there have not been any fatalities among approximately 80 paediatric cases in Hong Kong, and only one adolescent patient has required mechanical ventilation.

Table 1

Clinical course, progression, and laboratory features of SARS


With the establishment of the WHO laboratory network around the world and subsequent rapid progress in the isolation of the possible agent, it is thought most likely that SARS is caused by a novel strain of coronavirus.8–10 Coronaviruses are classified as members of the order Nidovirales, a group of enveloped positive sense RNA viruses which synthesise a 3‘ co-terminal set of subgenomic mRNAs in infected cells.11 Coronaviruses are known to cause common respiratory and enteric diseases of humans and domestic animals.12,13

Although there has been significant progress in the development of a rapid diagnostic test, the current rapid reverse transcription (RT) polymerase chain reaction (PCR) test for detection of this new coronavirus is not yet widely available for early diagnosis.14 The other diagnostic tests currently being used include viral isolation and serum antibody tests. However, these are only useful for epidemiological surveys or retrospective confirmation of the diagnosis and cannot be used to confirm the diagnosis early in the course of the illness. The early management of patients with suspected SARS is therefore still based on the clinical presentation and possible contact with known SARS patients. Among the initial cohort of patients admitted to the Prince of Wales Hospital with SARS,3 over 90% were subsequently confirmed to have evidence of infection with the SARS associated coronavirus either by serum antibody or RT-PCR testing.


The medical treatment of SARS remains controversial. None of the currently used medications has been tested in randomised controlled trials. Ribavirin, a broad spectrum antiviral agent, has been widely used. Necroscopic examination of fatal cases has revealed diffuse alveolar damage, hyaline membrane formation, desquamation of pneumocytes in alveolar spaces, and scanty interstitial inflammatory cell infiltrates in the lungs.3,10 Furthermore, a proportion of patients may have radiological features similar to those of BOOP, which is sensitive to steroid therapy. We have therefore used a combination of ribavirin and a steroid in treating the cases in Hong Kong, and most patients have stabilised and improved with such treatment.

The treatment regime used in Hong Kong is oral ribavirin (loading dose of 2.4 g followed by 1.2 g three times a day) and a “low dose” corticosteroid (prednisolone 0.5–1 mg/kg/day). Those with progressive dyspnoea and hypoxia are treated with intravenous ribavirin (400 mg every 8 hours) combined with hydrocortisone (100 mg every 6 hours). Pulses of high dose methyprednisolone (0.5 g daily for 3 days) are given to patients who continue to have fever and progressive clinical and radiological deterioration. As a last resort, those who continue to deteriorate despite the use of pulsed methylprednisolone have been treated with convalescent serum obtained from patients who have recovered from SARS. Convalescent serum is obtained by apharesis using a cell separator (Baxter CS 300) operating on plasma exchange mode. Only a small number of patients have been treated with convalescent serum and it is too early to judge the effectiveness of this treatment. Other centres might not be as liberal in the use of steroid and ribavirin as we have been in Hong Kong, but they have achieved similar outcomes.2,3,5 Randomised controlled trials are needed to determine the effectiveness of steroid and convalescent serum in the management of SARS. The reported mortality worldwide is about 8–15%.2,3,5,15,16 The three independent predictive factors for a poor outcome (ICU admission or death) are advanced age, a high peak level of LDH, and a high absolute neutrophil count at presentation.3

During follow up of more than 200 adults and children with SARS who have been discharged after 21 days in hospital we have not seen any case of relapse, although side effects of ribavirin (haemolytic anaemia) and steroid (myopathy) have been reported commonly in adults. Many patients who have recovered clinically have expressed concern about the possibility of being infectious to others as their excreta may still contain coronavirus. More research is needed to determine the potential infectivity of these patients.


Strict infection control in the hospital setting is essential for the management of SARS. The infection is highly contagious and appears to spread by close contact droplet transmission.17 Given the right environmental factors, it may also spread by the faecal-oral route as in the outbreak at the Amoy Gardens in Hong Kong. In addition to respiratory secretions, urine and faecal material should be considered and handled as infectious materials. Although it is still not clear how long this new coronavirus can survive in the environment, preliminary studies by the WHO laboratory network have shown that the virus is stable in the urine and faeces for at least 2 days.18 The virus has been found to be even more stable (up to 4 days) in stools from patients with diarrhoea because of the higher pH of the stools in these patients. It is of paramount importance that healthcare workers are fully trained in infection control, and that patients are managed in wards with proper isolation facilities to avoid cross infection between patients. Ideally, where there is an outbreak in the community, they should be treated in wards designated for SARS patients only. Strict adherence to the steps of infection control is mandatory to avoid an unacceptable rate of infection among healthcare workers. Details of the infection control measures can be obtained from the WHO and related websites.1

Patients with SARS should be transferred to hospitals with specially trained staff and proper isolation facilities to avoid spread of the infection. Because of the highly infectious nature of this condition, we do not allow visitors into the wards for SARS patients. Furthermore, nebulisers should not be used as they may generate more infective droplets from the patients leading to enhanced transmission to healthcare workers. Similarly, we have reservations about the use of non-invasive positive pressure ventilation (NPPV) for patients with SARS, although some patients can be treated with NPPV to avoid intubation. It is unclear how long patients continue to shed the virus in their respiratory secretions, urine or faeces after recovery, but a preliminary study of 75 patients from Hong Kong suggested that over 50% of patients continued to excrete the virus 3 weeks after the onset of illness.19 Further studies are needed to define the period of infectivity of these patients. Finally, public health and quarantine measures are extremely important in controlling the spread of the infection in the community.1


This paper is dedicated to all those frontline healthcare workers who risked their lives to care for patients with SARS all over the world.

SARS is a serious respiratory illness that frequently runs a rapidly progressive downhill course. In just 6 weeks it has spread to all continents of the globe. At the time of writing more than 7000 cases have been reported worldwide and over 500 have died. The primary mode of transmission appears to be by droplets. Good supportive care and the judicious use of ribavirin and steroids result in recovery in over 90% of patients, but randomised controlled trials are needed to define the roles of these treatments. Success in controlling the disease relies on early identification of suspect cases, proper isolation, and meticulous infection control measures. Development of sensitive and specific rapid diagnostic tests is underway.


View Abstract