Travellers' diarrhoea: Impact of TD definition and control group design on study results

Background: Travellers' diarrhoea (TD) is a common health problem among visitors to the (sub)tropics. Much research deals with aetiology, prevention, and post-infection sequalae, yet the data may not allow comparisons due to incompatible definitions of TD and No TD control groups. Method: The impact of defining TD and No TD control groups was explored by revisiting our recent data. We set up two TD groups: classical TD i.e. >= 3 loose or liquid stools/day and WHO TD (diarrhoea as defined by the WHO) i.e. any diarrhoea, and four No TD groups by TD definition and timing (no classical/WHO TD during travel, no ongoing classical/WHO TD). Results: TD was recorded for 37% versus 65% of subjects when using classical versus WHO definitions, respectively; the proportions of the various pathogens proved similar. The strictest criterion for the No TD control group (no WHO TD during travel) yielded pathogens among 61% and the least strict (no ongoing classical TD) among 73% of the travellers; the differences were greatest for enteroaggregative Escherichia coll. and Campylobacter. Conclusions: Definition of TD and control group design substantially impact on TD study results. The WHO definition yields more cases, but the pathogen selection is similar by both definitions. Design of the No TD control group was found critical: only those remaining asymptomatic throughout the journey should be included.