Study type: systematic review

EL 2+

(different population)
To systematically review the precision and accuracy of symptoms, signs, and basic laboratory tests for evaluating dehydration in infants and children.

They identified articles by direct searches of the MEDLINE database via the PubMed search engine. The first and most broad search strategy used dehydration and diagnosis, hypovolemia and diagnosis, or intravascular volume depletion and diagnosis. All were limited by age (all children: 0–18 years) and publication date (January 1966–April 2003). These searches produced 1537 articles. They supplemented this preliminary search with the standardized search technique used in the ‘Rational Clinical Examination’ series (available from the authors). This second search produced 24 additional articles.

Each of the authors reviewed the titles and available abstracts from the 1561 articles, selecting for further review those that appeared to address the evaluation of dehydration in children aged 1 month to 5 years. They did not exclude articles if the study enrolled some children outside of that age range. Through consensus, they identified 68 articles as potential sources of primary data or reviews with potential background information and thorough reference lists.

They performed a full review of the 110 retained articles to identify those with primary data comparing dehydration with a symptom, sign, or laboratory value in paediatric patients. Twenty-six articles met these criteria and underwent full quality assessment using an established methodological filter.

To ensure a comprehensive literature review, they used additional techniques to identify articles. One author (M.J.S.) searched for individual symptoms and signs associated with the diagnosis of dehydration in children. These terms included capillary refill, skin turgor, dry cry, tears, mucous membrane, sunken eyes, fontanelle and dehydration, urine specific gravity, urine and dehydration, haemoconcentration, BUN, urine, blood pressure, bioimpedance, orthostasis, respiration, parent and dehydration, pulse, and heart rate (all limit: aged 0–18 years, human, NOT dehydration and diagnosis). The Cochrane Library, reference lists of paediatric and physical examination textbooks, reference lists of all included articles, and articles from the collections of experts in the field were reviewed. Forty-two potential articles were identified from the supplemental searches. A second author then checked the initial quality review. The group always arrived at a consensus on the final evidence quality level assigned. Nine of the 110 articles that underwent a full text review were written in languages other than English. Medical school faculty, residents, or students at our institution who were primary speakers of the written language read each of these articles. Six of these 9 articles did not meet inclusion criteria and were excluded, while 3 were assigned an evidence quality level based on a translation of the article.

No studies on physical examination signs, symptoms, or laboratory results in childhood dehydration demonstrated evidence quality criteria for level 1 or 2. Four studies were assigned to level 3, but 1 of these was eventually excluded because the study population overlapped with that in another included study. Twelve studies were initially assigned to level 4, though 1 was excluded because of methodological flaws and another was excluded because of its retrospective design and restriction to children with pyloric stenosis.

They chose the difference between the rehydration weight and the acute weight divided by the rehydration weight as the best available gold standard of percentage of volume lost. Ten articles used gold standards based solely on examination signs or a general dehydration assessment. These were assigned an evidence quality level of 5 and were subsequently excluded.
Three studies evaluated the accuracy of history taking in assessing dehydration. All 3 of these studies evaluated history of low urine output as a test for dehydration. In the pooled analysis, low urine output did not increase the likelihood of 5% dehydration (LR, 1.3; 95% CI, 0.9–1.9). Porter et al showed that a history of vomiting, diarrhea, decreased oral intake, reported low urine output, a previous trial of clear liquids, and having seen another clinician during the illness prior to presenting to the ED yielded LRs that lacked utility in the assessment of dehydration. However, their data did suggest that children who had not been previously evaluated by a physician during the illness might be less likely to be dehydrated on presentation (LR, 0.09; 95% CI, 0.01–1.37). Similarly, parental report of a normal urine output decreases the likelihood of dehydration (Gorelick et al reported an LR of 0.27 [95% CI, 0.14–0.51] and Porter et al reported an LR of 0.16 [95% CI, 0.01–2.53]).

Table : Summary characteristics for clinical findings to detect 5% dehydration.
LR summary, Value (95 °CI) or range
FindingTotal No.PresentAbsentSensitivity (95% CI)Specificity (95% CI)
Prolonged CRT4784.1 (1.7–9.8)0.57 (0.39–0.82)0.60 (0.29–0.91)0.85 (0.72–0.98)
Abnormal skin turgor6022.5 (1.5–4.2)0.66 (0.57–0.75)0.58 (0.40–0.75)0.76(0.59–0.93)
Abnormal respiratory pattern5812.0 (1.5–2.7)0.76 (0.62–0.88)0.43 (0.31–0.55)0.79(0.72–0.86)
Sunken eyes5331.7 (1.1–2.5)0.49 (0.38–0.63)0.75 (0.62–0.88)0.52 (0.22–0.81)
Dry mucus membranes5331.7 (1.1–2.6)0.41 (0.21–0.79)0.86 (0.80–0.92)0.44 (0.13–0.74)
Cool extremity2061.5,18.80.89–0.970.10, 0.110.93, 1.00
Week pulse3603.1, 7.20.66–0.960.04, 0.250.86, 1.00
Absent tears3982.3 (0.9–5.8)0.54 (0.26–1.13)0.63 (0.42–0.84)0.68 (0.43–0.94)
Increased heart rate4621.3 (0.8–2.0)0.82 (0.64–1.05)0.52 (0.44–0.60)0.58 (0.33–0.82)
Sunken fontanelle3080.9 (0.6–1.3)1.12 (0.82– 1.54)0.49 (0.37–0.60)0.54 (0.22–0.87)
Poor overall appearance3981.9 (0.97–3.8)0.46 (0.34–0.61)0.80 (0.57–1.04)0.45 (−0.1 to 1.02)
LR: likelihood ratio.
Three signs were evaluated in multiple studies, had a clinically helpful pooled LR in detecting 5% dehydration, and had 95% CIs wholly above 1.0. Capillary refill time was evaluated in 4 different studies, and the pooled sensitivity of prolonged capillary refill time was 0.60 (95% CI, 0.29–0.91), with a specificity of 0.85 (95% CI, 0.72–0.98), for detecting 5% dehydration. The LR for abnormal capillary refill time was 4.1 (95% CI, 1.7–9.8). This was the highest value among examination signs with pooled results. Abnormal skin turgor had a pooled LR of 2.5 (95% CI, 1.5–4.2) and abnormal respiratory pattern had a pooled LR of 2.0 (95% CI, 1.5– 2.7).

Presence of cool extremities or a weak pulse or absence of tears also may be helpful tests for dehydration. Absence of tears had a pooled LR of 2.3 (95% CI, 0.9–5.8), but the potential utility is limited by a wide 95% CI that crosses 1.0. Two studies examined a weak pulse quality as a test for dehydration. One study found a reasonably precise LR for weak pulse of 3.1 (95% CI, 1.8–5.4), but in the other study, the 95% CI was too wide to make a reasonable estimate (LR, 7.2; 95% CI, 0.4–150). The 2 studies that evaluated cool extremities as a test of dehydration found imprecise point estimates for the LR positive in detecting 5% dehydration (LR, 18.8; 95% CI, 1.1–330and LR, 1.5; 95% CI, 0.2–12).

Sunken eyes and dry mucous membranes offer little help clinically; both had narrow 95% CIs but pooled LRs of 1.7. An increased heart rate, a sunken fontanelle in young infants, and an overall poor appearance are frequently taught as good tests for dehydration. However, the objective evidence reveals that all have summary LRs of less than 2.0 and 95% CIs that cross 1.0.

Some tests may be clinically useful in decreasing the likelihood of dehydration. Absence of dry mucous membranes (LR, 0.41; 95% CI, 0.21–0.79), a normal overall appearance (LR, 0.46; 95% CI, 0.34–0.61), and absence of sunken eyes (LR, 0.49; 95% CI, 0.38–0.63) had pooled LRs of less than 0.5. Most clinical scenarios will necessitate lower LRs than these to rule out dehydration effectively.

From: Evidence tables

Cover of Feverish Illness in Children
Feverish Illness in Children: Assessment and Initial Management in Children Younger than 5 Years.
NICE Clinical Guidelines, No. 47.
National Collaborating Centre for Women’s and Children’s Health (UK).
London (UK): RCOG Press; 2007 May.
Copyright © 2007, National Collaborating Centre for Women’s and Children’s Health.

No part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK []. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.