Geographic, symptomatic and laboratory predictors of parasitic and bacterial causes of diarrhoea in travellers.

An observational study of patients presenting with diarrhoea to a walk-in service for returning travellers was conducted with the aim of identifying features that would help predict whether pathogens were bacterial or parasitic. In total, 509 cases were included, of which a bacterial aetiology was found in 55/440 (12.5%) and a parasitic cause in 51/428 (11.9%). Patients with symptoms of ≤14 days were significantly more likely to have a bacterial diagnosis than those with longer symptoms (p<0.001), whereas parasitic causes of diarrhoea were not associated with length of symptoms and became proportionately more likely with time. Raised CRP, faecal white cells and fever were all predictive of positive bacterial culture (p<0.001, p=0.001 and p=0.001, respectively) but did not predict parasitic infection. Travellers to South and Southeast Asia were more likely to have parasites detected in their stool than travellers to other tropical areas (OR=1.96; p=0.041). Gender, ethnicity, reason for travel and length of stay abroad were not significantly associated with the faecal pathogen identified. These findings should help guide appropriate antimicrobials when empirical therapy is indicated.