RESUMO
This paper describes how the methodology used for measuring diarrhoeal morbidity has changed over time and assesses how differences in episode definition have affected estimates of diarrhoeal morbidity among children aged less than five years. The episode definition used in 73 studies included in three previously-published literature reviews was identified. In earlier work, a method was developed that adjusts morbidity estimates to take account of differences in episode definition. This adjustment method was applied to the studies identified in these three literature reviews. Episode definitions were better documented and were more consistent in studies published after 1980. Adjusting morbidity estimates to account for definitional differences did not substantially alter the reviews' conclusions. Diarrhoeal surveillance has steadily improved since 1980, with methodology becoming more consistent between studies and better documented. Although episode definitions have changed over time, the morbidity estimates derived in the three reviews appear robust to these changes.