Comparison of Clinical Associations and Laboratory Abnormalities in Children With Moderate and Severe Dehydration

Abstract

To search for possible early clinical associations and laboratory abnormalities in children with severe dehydration in northern Jordan. We prospectively evaluated 251 children with acute gastroenteritis. Dehydration assessment was done following a known clinical scheme. Probable clinical associations and laboratory abnormalities were examined against the preassigned dehydration status. Children with severe dehydration had significantly more hypernatremia and hyperkalemia, less isonatremia, and higher mean levels of urea, creatinine, and glucose (P < 0.005). Receiver operating characteristic curves showed statistically significant area under the curve values for laboratory variables. These area under the curve values were 0.991 (95% confidence interval [CI] 0.980-1.001) for serum urea, 0.862 (95% CI 0.746-0.978) for sodium, 0.850 (95% CI 0.751-0.949) for creatinine, 0.69 (95% CI 0.555-0.824) for potassium, and 0.684 (95% CI 0.574-0.795) for glucose (P < 0.05 for all). Certain independent serum cutoff levels of urea, creatinine, sodium, glucose, and potassium had high negative predictive value (100%), whereas other cutoff values for each, except potassium, had high positive predictive value (100%) for severe dehydration. Historic clinical characteristics of patients did not correlate to dehydration degree. Serum urea, creatinine, sodium, potassium, and glucose were useful independently in augmenting clinical examination to diagnose the degree of dehydration status among children presenting with gastroenteritis. Serum urea performed the best among all. On the contrary, none of the examined historical clinical patterns could be correlated to the dehydration status. Larger and multicenter studies are needed to validate our results and to examine their impact on final outcomes.