Does integrated management of childhood illness (IMCI) make a difference to the assessment of sick children in Papua New Guinea?

Abstract

Two provinces, one of which had introduced the Integrated Management of Childhood Illness (IMCI) policy to some degree and one in which there was no IMCI program, were selected to compare health workers' assessment of children attending provincial hospitals, district hospitals and health centres. 23 health workers were observed during 373 child assessments to determine their ability to detect the symptoms and signs detailed in the IMCI 10-step checklist. Health workers in the province that had introduced IMCI performed significantly better than their counterparts in 11 of the 24 criteria studied. These criteria included asking about 'too sick symptoms' (p < 0.001 for asking about vomiting and feeding and p < 0.012 for asking about convulsions), counting respiratory rate and checking for chest indrawing in children presenting with cough (p < 0.001), checking skin elasticity in children presenting with diarrhoea (p < 0.02), checking for neck stiffness in those presenting with fever (p < 0.001), checking for pallor (p < 0.001) and accurately plotting the child's weight on the weight graph (p < 0.001). Children in this province were more likely to be fully vaccinated (OR 1.96 [1.25-3.08]) than those in the province in which no attempt had been made to introduce IMCI. The facilities were ranked by the proportion of children correctly assessed. The best facility was the health centre which had been a pilot site for the introduction of IMCI in the province several years before the study. The results of the study, which clearly demonstrate that IMCI does make a difference, are in accordance with data from many parts of the resource-poor world and strongly support the Department of Health's decision to implement IMCI in the country. Every effort should be made to ensure that all provinces introduce the program and support its continuation as part of the Strategic Package for Child Survival.