A predictive model basing on a four-item clinical score of cough before or concomitant with fever, myalgia, diarrhoea and rhinorrhoea or sore throat had a 100% sensitivity and 75.9% specificity of early detection of probable SARS. The addition of lymphopaenia and thrombocytopaenia increased the specificity to 86.2%. 38 In another model, a scoring system of attributing 11, 10, 3, 3 and 3 points to the presence of independent risk factors of epidemiological link, radiographic deterioration, myalgia, lymphopaenia and elevated ALT respectively, generated high- (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) and low-(0-10) risk scores for SARS. The sensitivity and specificity of this prediction rule in positively identifying a SARS patient were 97.7% and 81.3% respectively. 39 The prediction rule could be useful at the bedside. However, these studies were conducted in adult patients and would need to be validated in paediatric patients.