Influence of recall period on estimates of diarrhoea morbidity in infants in rural Tamil Nadu.

Data collected on 689 infants, in a study to assess the incidence of diarrhoea during infancy, is used here to quantify the extent of under-reporting in diarrhoea morbidity surveys. The study area consisted of two contiguous primary health centres in Villupuram health unit district in Tamil Nadu, south India. Each day of infancy was assigned a recall period and proportion of diarrhoeal days for various recall period computed. The proportion of diarrhoea was 11.3%, 12.0% and 11.2% for zero, one and two days of recall period, respectively, (mean 11.5%) after which the proportion decreased. The under-reporting of diarrhoea was approximately 15%, 26% and 45% with three, six and 7-13 days of recall period, respectively compared to mean. As there is considerable under-reporting of diarrhoea morbidity when recall period exceeds three days, it would be better to collect information on diarrhoea twice a week in diarrhoeal morbidity surveys.