Current World Health Organisation (WHO) surveillance guidelines for determining the proportion of suspect, probable and confirmed SARS cases may not be sufficiently sensitive when assessing patients in the community before admission to hospital, suggest researchers from The Chinese University of Hong Kong in this week's British Medical Journal.

　　After the recent outbreak in the Prince of Wales Hospital in Hong Kong, a temporary SARS screening clinic was opened. Of the first 556 hospital staff, patients, and relatives who had had contact with someone with SARS, 141 were admitted to hospital, and 97 had confirmed SARS.

　　Current WHO guidelines emphasise that a case is suspect SARS if they meet ALL of the criteria including contact history, fever defined as a temperature > 38°C, and respiratory tract symptoms (such as cough, shortness of breath, and breathing difficulty). Researchers found that these symptoms did not feature strongly in the early stages of the illness and if applied rigorously would miss 75% cases. Many patients do not have documentable temperature readings >38°C in the first 10 days of illness despite having obvious pneumonia on chest x-ray. Also, symptoms such as fever, chills, myalagia, malaise, loss of appetite, vomiting, and diarrhoea, were significantly more common in SARS patients than in other people who did not have SARS.

　　In screening patients for SARS, these symptoms may be better discriminators than the symptoms listed in the WHO guidelines, which were based on patients who were already in hospital, say the researchers.

　　The main conclusions were that if WHO criteria are used for screening subjects in the community to detect SARS then they should be refined to include documentation of systemic, gastrointestinal as well as respiratory symptoms, and should emphasise the importance of routine daily follow up and daily chest radiography until patients have passed at least 48 hours without symptoms.