Assessing the accuracy of three clinical dehydration scales in children with diarrhea in a resource-limited setting

Background: Diarrhea accounts for nearly one-fifth of child deaths worldwide. Since assessing dehydration status is critical to providing appropriate treatment and preventing death in children with diarrhea, we evaluate the accuracy of three popular clinical dehydration scales.

Methods: We prospectively enrolled children admitted with diarrhea to three rural Rwandan hospitals. A nurse or physician assessed each child using the World Health Organization (WHO), Gorelick, and Clinical Dehydration Scale (CDS). Children were rehydrated according to hospital protocols and weighed daily, with greater than 10% change between admission and stable weight considered as severe dehydration. Receiver operating characteristic (ROC) curves were created for each of the three scales compared to the composite outcome of severe dehydration or death.

Findings: 170 children under five were enrolled. The WHO, CDS, and Gorelick scales were all moderate predictors of our composite outcome, with area under the ROC curves (AUC) of 0.680 (p=.028), 0.733 (p=.004), and 0.691 (p=.019), respectively. When limited to patients evaluated by a nurse, only the CDS had a statistically significant AUC of 0.753 (p=.028). In patients evaluated by a physician, the WHO, CDS, and Gorelick scales had AUCs of 0.765 (p=.023), 0.761 (p=.025), and 0.794 (p=.011), respectively.

Interpretation: While all three scales were moderate predictors of dehydration or death, their accuracy differed by provider type. In particular, the WHO scale did not perform well when used by nurses, the front-line providers of pediatric care in most resource-limited settings. Further research is necessary to determine the best clinical dehydration scale in these settings.