Objectives: To determine the clinical and radiological features of severe acute respiratory syndrome (SARS) and to evaluate the accuracy of the World Health Organization's guidelines on defining cases of SARS.
Design: Prospective observational study.
Setting: A newly set up SARS screening clinic in the emergency department of a university hospital in Hong Kong's New Territories.
Participants: 556 hospital staff, patients, and relatives who attended the screening clinic and who had had contact with someone with SARS.
Main outcome measure: Number of confirmed cases of SARS.
Results: Of the 556 people, 141 were admitted to hospital, and 97 had confirmed SARS. Fever, chills, malaise, myalgia, rigor, loss of appetite, vomiting, diarrhoea, and neck pain but not respiratory tract symptoms were significantly more common among the 97 patients than among the other patients. The overall accuracy of the WHO guidelines for identifying suspected SARS was 83% and their negative predictive value was 86% (95% confidence interval 83% to 89%). They had a sensitivity of 26% (17% to 36%) and a specificity of 96% (93% to 97%).
Conclusions: Current WHO guidelines for diagnosing suspected SARS may not be sufficiently sensitive in assessing patients before admission to hospital. Daily follow up, evaluation of non-respiratory, systemic symptoms, and chest radiography would be better screening tools.