Estimation of the Degree of Dehydration


Synopsis: Estimation of the degree of dehydration by the use of clinical scales combined with a measurement of serum bicarbonate level accurately identified all children with severe and most children with moderately severe dehydration as indicated by the actual percentage loss of body weight determined by weights during and after the acute episode.

Source: Vega RM, Avner JR. A prospective study of the usefulness of clinical and laboratory parameters for predicting percentage of dehydration in children. Pediatr Emerg Care 1997;13:179-182.

The physician’s estimation of a child’s hydration status is very important in determining proper fluid administration for a child with acute dehydration. The most accurate indicator of the magnitude of dehydration is the percent loss of body weight (PLBW) during the course of the illness. However, determination of PLBW requires accurate information about the child’s baseline weight before the illness, and this is rarely available. Vega and Avner, emergency service pediatricians at the Jacobi Hospital in the Bronx, studied 97 children who required intravenous therapy for acute dehydration. A clinical estimation of dehydration was made based upon a standard clinical scale. Serum electrolytes were obtained.

The children were classified according to PLBW into three groups: mild, PLBW less than 5% (n = 50); moderate, PLBW 6-10% (n = 30); and severe, PLBW greater than 10% (n = 17). (See table.) The physician’s clinical estimate of dehydration compared to PLBW had a sensitivity of 74% (95% confidence interval 60-85%) for mild dehydration; 33% (CI 17-53%) for moderate dehydration; and 70% (CI 44-89%) for severe dehydration. There was a significant difference in mean serum bicarbonate levels in the three groups. The combination of serum bicarbonate level was less than 17 mEq/L, and clinical assessment identified all 17 children with severe dehydration (PLBW > 10%) and 90% (27/30) of children with moderate dehydration (PLBW 6-10%).


Clinical Assessment of Severity of Dehydration

Mild Moderate Severe dehydratiodehydratiodehydration

Body weight loss 4-5% 6-9% 10% or more

General appearance Thirsty, Thirsty, drowsy, Drowsy, limp, restless, alert, postural cold, sweaty hypotensiohypotension

Radial pulse Normal Rapid, weak Rapid, thready or impalpable

Respirations Normal Deep, may be Deep, rapid rapid

Ant. fontanel Normal SunkeVery sunken

Systolic BP Normal Normal or low Low

Skin elasticity Pinch retracts Pinch retracts Pinch retracts

immediately slowly very slowly

Eyes Normal SunkeVery sunken

Tears Present Absent Absent

Mucous membranes Moist Dry Very dry


In this paper, the authors carried out a prospective study in which an attempt was made to develop clinical and laboratory parameters that will help in predicting PLBW in children who are dehydrated. The authors suggest that a combination of the traditional clinical parameters with a serum bicarbonate level less than 17 mEq/L will substantially improve the identification of children with severe and moderate dehydration. This study produces some interesting observations and raises some important issues relative to this fundamental component of fluid and electrolyte therapy. The authors recognize that the follow-up weights for their patients were not consistently done using the same scale on which the initial weight was obtained and that the duration from the initial rehydration to the final weight was highly variable (i.e., from 72 hours to a week or more). In addition, the level of serum bicarbonate that was chosen to help differentiate those patients with moderate or severe dehydration was chosen after the study had been completed rather than prospectively so that the accumulated results do not represent a validation of this observation but, instead, a retrospective analysis of the accumulated data. If one looks solely at the positive predictive value of the serum bicarbonate for children with severe dehydration, it is only 31% (16/51). Therefore, this parameter alone does not have a sufficiently high positive predictive value to be clinically relevant. Moreover, it is important to remember that in children with severe vomiting, serum bicarbonate is frequently elevated because of the association of chloride depletion from the vomiting with development of metabolic alkalosis that is sustained because of the intravascular volume depletion.

Overall, the authors’ observations are of interest but will require a careful prospectively designed study for validation prior to widespread clinical implementation. (Dr. Siegel is Vice-Chairman of Pediatrics and Head of Pediatric Nephrology, Yale-New Haven Children’s Hospital.)