• 
• 

Three Part Question

In [children attending the Accident and Emergency Department with suspected dehydration] [what is the degree of variation] between observers when [performing assessment of skin turgor]

Clinical Scenario

A 3 year old child attends the Accident and Emergency Department with a 2 day history of vomiting and diarrhoea. You wish to estimate the child's hydration status to determine whether rehydration therapy is needed but wonder how reliable is the clinical sign of skin turgor.

Search Strategy

Medline 1966 - 2006 November Week 3 and Embase 1980-07/04
Medline:
(turgor.af. OR (clinical sign$ OR examination).mp. OR exp Physical Examination/) AND (exp Dehydration/ OR dehydrat$.mp. OR exp Hypovolemia/ OR (hypovolaem$ OR hypovolaem$).mp.) AND (Best BETs Paediatric filter (maximally sensitive))
Embase:
(turgor.af. OR (clinical sign$ OR examination).mp. OR exp Physical Examination/) AND (exp Dehydration/ OR dehydrat$.mp. OR exp Hypovolemia/ OR (hypovolaem$ OR hypovolaem$).mp.) limit to Child, Unspecified Age

Search Outcome

478 papers were found using Medline and 107 using Embase , three of which were relevant to the topic of interest. One relevant systematic review summarised two of these papers but included no additional papers (Steiner et al, 2004).

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Gorelick et al, 1997, USA	84 patients, Age(1 month to 5 years)	Prospective cohort study Urban Paediatric Emergency Hospital 2 out of 17 independent assessors	Level of observer agreement	k=0.55	Small numbers Blinding Convenience sample
Duggan et al, 1997, Egypt	100 patients Age (2 months to 2 years)	Prospective cohort study Gastroenteritis Unit - 3 independent assessors	Level of observer agreement	k1=0.36 95% CI(0.525-0.203), k2=0.511 95% CI(0.66-0.362), k3=0.417 95% CI(0.568-0.266)	Convenience sample Standardised assessment of skin turgor
Otieno et al, 2004, Kenya	100 consecutive patients	Prospective cohort study Age(2 days to 10 years 11months) Rural district hospital 4 independent assessors	Level of observer agreement	k=0.55 95% CI(0.4-0.7)	Heterogenous group Nil exclusion criteria stated

Comment(s)

Dehydration is an important cause of childhood morbidity and mortality worldwide. Skin turgor is generally accepted as part of the clinical assessment of hydration status in children being a quick, non-invasive test that can be performed at the bedside. Unrecognised fluid deficit can lead to electrolyte and acid-base disturbances as well as end organ damage. Conversely over estimation of fluid deficit can result in inappropriate rehydration therapy. The value of any clinical test is a function of its ability to detect a particular condition and its reliability .The limited number of studies which have various sources of bias show only moderate agreement for inter-observer reproducibility when skin turgor is used to clinically assess hydration status in children.

Clinical Bottom Line

Skin turgor measurement whilst part of the initial assessment of children with suspected dehydration, is only moderately reliable and other clinical signs should be sought to confirm this diagnosis.

References

1. Gorelick M, Shaw K, O'Murphy K. Validity and reliability of clinical signs in the diagnosis of dehydration in children. Paediatrics May 1997, Vol 99, 1-6.
2. Duggan C, Refat M, Hashem M, Fayad I, Santosham M. Interrater agreement in the assessment of dehydration in infants. Journal of Tropical Paediatrics 1997, Vol 43, 119-121.
3. Otieno H, Were E, Ahmed I, Charo E, Brent A, Maitland K. Are bedside features of shock reproducible between different observers. Archives of Disease in Children 2004,Vol 89, 977-979.
4. Steiner MJ; De Walt DA; Byerley JS. Is this child dehydrated? JAMA 2004; 291: 2746-2754.