Comparing the accuracy of the three popular clinical dehydration scales in children with diarrhea

BACKGROUND Dehydration due to acute gastroenteritis is one of the leading causes of mortality in children worldwide. The World Health Organization (WHO) scale, the Gorelick scale, and the Clinical Dehydration Scale (CDS) were created to estimate percentage dehydration in children with gastroenteritis based on clinical signs. Of these, only the CDS has been prospectively validated against a valid gold standard, though never in low- and middle-income countries. The purpose of this study is to determine whether these clinical scales can accurately assess dehydration status in children when performed by nurses or general physicians in a low-income country. METHODS We prospectively enrolled a non-consecutive sample of children presenting to three Rwandan hospitals with diarrhea and/or vomiting. A health care provider documented clinical signs on arrival and weighed the patient using a standard scale. Once admitted, the patient received rehydration according to standard hospital protocol and was weighed again at hospital discharge. Receiver operating characteristic (ROC) curves were created for each of the three scales compared to the gold standard, percent weight change with rehydration. Sensitivity, specificity, and likelihood ratios were calculated based on the best cutoff points of the ROC curves. RESULTS We enrolled 73 children, and 49 children met eligibility criteria. Based on our gold standard, the children had a mean percent dehydration of 5% on arrival. The WHO scale, Gorelick scale, and CDS did not have an area under the ROC curve statistically different from the reference line. The WHO scale had sensitivities of 79% and 50% and specificities of 43% and 61% for severe and moderate dehydration, respectively; the 4- and 10-point Gorelick scale had sensitivities of 64% and 21% and specificities of 69% and 89%, respectively, for severe dehydration, while the same scales had sensitivities of 68% and 82% and specificities of 41% and 35% for moderate dehydration; the CDS had a sensitivity of 68% and specificity of 45% for moderate dehydration. CONCLUSION In this sample of children, the WHO scale, Gorelick scale, and CDS did not provide an accurate assessment of dehydration status when used by general physicians and nurses in a developing world setting.