Clinical Versus Laboratory for Estimating of Dehydration Severity
Total Page:16
File Type:pdf, Size:1020Kb
Clinical versus laboratory for estimating of dehydration severity Majid Malaki Pediatric Health Research Center, Tabriz Medical University, Tabriz, Iran ABSTRACT Background: Acute gastroenteritis is a common cause of dehydration and precise estimation of dehydration Materials and Methods: D is a vital matter for clinical decisions. We try to find how much clinically diagnosed scales are compatible with ORIGINAL ARTICLE laboratory tests measures. uring 2 years 95 infants and children aged between 2 and 108 months entered to emergency room with acute gastroenteritis. They were categorized as mild, moderate and severe dehydration, their recorded laboratory tests include blood urea nitrogen (BUN), creatinine, venous blood gases values were expressedP by means ±95% of confidence intervalResult and compared by mann-whitney test in each groups with SPSS 16, sensitivity, specificity and likelihood ratio measured for defined cut off values in severe dehydration group, value less than 0.05 was significant. : Severe dehydration includes 3% Conclusionof all hospitalization: R due to dehydration. Laboratory tests cannot differentiate mild to moderate dehydration definietly but this difference is significant between severe to mild and severe to moderate dehydration. outine laboratory test are not generally helpful for dehydration severity estimation but they can be discriminate severe from mild or moderate dehydration exclusively. Creatinine higher than 0.9 mg/dl and BaseKey words deficit: beyond-16A are specific (90%) for severe dehydration estimation in infant and children. cute gastroenteritis, children, dehydration, laboratory tests Introduction more yield for prediction of dehydration severity[4-5,8,11] Yilmaz et al.[12] find serum Urea and bicarbonate can be Acute gastroenteritis is a common reason of predictive for dehydration severity in small children and hospitalization.[1] It seems that precise diagnosis of infants as high risk groups to dehydration and We try to dehydration severity may be dependent on laboratory find that what laboratory tests can be more compatible for tests,[2] many tried to define dehydration severity with dehydration estimation compared to clinical estimation clinically guideline.[3-6] Capillary refill time, abnormal that can be helpful for physicians in perplexing cases. skin turgor, and respiratory pattern were suggested as the best signs.[7] All these valuable signs may be neither Methods and Materials sensitive nor specific for dehydration severity.[8-9] Clinical dehydration scale (CDS) considers general appearance, 95 children aged between 2 and 108 months diagnosed eyes changes, mucous membrane moisture and tears during 48 month from April 2009 to April 2011 as acute [10] existence hospital staying duration can be related to gastroenteritis categorized as mild, moderate and severe CDS not to laboratory tests (PH and serum bicarbonate) with standard criteria[13] by a pediatrician at emergency disturbances. With respect to this truth that clinical room. After preparation of laboratory results (blood findings are not sensitive for estimating of dehydration urea nitrogen, creatinine, venous blood gas, electrolytes severity solely and importance of laboratory tests results include sodium, potassium) these documents evaluated besides to clinical estimation of dehydration grades give and matched with clinical definition for dehydration severity. Information about age, sex, duration of Access this article online disease and complaints were asked by a pediatrician Quick Response Code: Website: residence children height and weight recorded beside to www.atmph.org vital signs (blood pressure, pulse rate, respiratory rate and temperature), no one of cases received hypnotic and cholinergic drugs. None of patients had previous DOI: 10.4103/1755-6783.109262 diseases like as metabolic disease, renal dysfunction, malnutrition or surgical operation, they affected to dehydration due to acute gastroenteritis in recent days. Correspondence: Dr. Majid Malaki, Pediatric Health Research Center, Tabriz Medical University, Tabriz, Iran. E-mail: [email protected] of Tropical Medicine and Public Health | Nov-Dec 2012 | Vol 5 | Issue 6 559 Malaki: Laboratory tests in dehydration Their tests include urea creatinine electrolytes (sodium (3) The mean ± SD (Minimum, Maximum) of BUN and potassium) venous blood gas (base excess, base, to creatinine ratio in mild dehydration was 47±16 Co2 pressure) weight, height, sex and age of children (28,100) albeit it was not differed significantly were also recorded. Serum sodium 135-145 mEq/dl related to moderate dehydration 49 ±19, (30,130) and potassium 3/5-5/5 mEq/dl considered as normal. (p 0.56). BUN/Cr was higher significantly in severe All results expressed as percent, ratio, means, 95% dehydration 64±36, (34,170) compare to both of confidence interval, minimum and maximum moderate and mild dehydration (p 0.04) and relation of mentioned laboratory tests in each dehydration severity groups compared by one way (4) The mean ± SD of venous HCO3 level (mEq/L) in ANOVA test and followed by Post-Hoc test of Tukeys mild dehydration was 16±3 it was not significantly test in parametric distribution data for non-parametric differed to moderate dehydration 14.5±4 (P>0.05) ones the mann-whitney test used to compare every two but the mean ± SD of HCO3 was different in groups individually (mild -moderate), (mild -severe) and severe dehydration 10±3 compare to both mild or (moderate-severe). Four laboratory tests include blood moderate dehydration forms (P<0.001) [Table 1]. urea nitrogen above 40 mg/dl, creatinine more than 0.9 Venous HCO3 measures below 17 mEq/L were mg/dl, venous HCO3 below 17 and base deficit beyond observed in 15 out of 29 with mild, 30 out of 43 to-16 considered as cut off points for severe dehydration with moderate and 22 out of 23 patients with and we try to find sensitivity, specificity and positive severe group (+ LR 1.5 sensitivity 98% specify 38% likelihood ratio(+LR) of these defined values for for severe dehydration form at cut off point of 17 estimating severe form of dehydration particularly. p meq/l) [Table 2]. was significant if less than 0.05. (5) The mean ± SD of Base excess (BE) decreases Results significantly (P<0.0001) in severe dehydration compared to mild (-17±5 mEq/L vs-8±4) and 95 children means age (Minimum, Maximum), moderate dehydration (-17±5 vs-10±5). BE 19 months (2, 108) include 29 children with mild average was not differed between mild and dehydration means age (month) ± SD equal to 27±1.5 moderate (-8±4 vs.-10±5)(p 0.19) [Table 1]. Base months, 43 children with moderate dehydration means excess beyond to -16 mEq/L has not observed with age ± SD equal to 15±1 and 23 children with severe mild dehydration, it happened in 7 out of 43 with form aged 14±5 months. moderate and 15 out of 23 patients with severe group. (+ LR 6.5, sensitivity 65%, specificity 90% (1) Average blood urea level ± SD (mg/dl) in mild and in severe dehydration form) [Table 2]. moderate were 27±12 and 29±13 these measures were not differed between these two groups (mild Discussion and moderate), BUN level in severe form was 49±27 it was higher significantly in severe dehydration Attempts to estimate dehydration severity has been compared to mild or moderate dehydration. done for a long time but using laboratory tests for (P<0.001) [Table 1]. two out of 29 patients with accurate dehydration severity was noticed recently.[8,11] mild, 8 out of 43 with moderate and 11 out of 23 Some signs considered as the best tools for dehydration patients with severe dehydration have BUN more severity.[7] For example losing weight[14] although than 40 mg/dl (positive LR 3.4, sensitivity %48 baseline weights may not be inaccessible and signs specificity 86% for severe form occurrence) [Table 2]. like as skin turgor, deep breathing and capillary filling time have some limitations to use for dehydration (2) Mean ± SD of creatinine (mg/dl) in mild severity assessment.[7] Urea and bicarbonate level dehydration was 0.6±0.4 while in moderate grade were valuable to predict dehydration severity, serum it was 0.6± 0.1 there is not difference between bicarbonate above 15 can be valuable for mild and mild and moderate dehydration for creatinine moderate dehydration while it occurs infrequently in levels but in severe dehydration creatinine level was severe dehydration they consider urea above 100 mg/ 0.9±0.6 it was significantly higher in relation to dl suggestive for severe dehydration although serum mild and moderate dehydration (p 0.01) [Table 1]. bicarbonate below 18 mEq/L is another index that Creaetinine levels above 0.9 mg /dl observed 1 out can be used for severe or moderate dehydration of 29 with mild, 5 out of 43 with moderate and 8 estimation.[15] Another study showed that bicarbonate out of 23 with severe dehydration group (+ LR 4.3, above 15 mEq/L can occur in 7% of severe dehydration sensitivity 34%, specificity 92% for severe form) episodes. They suggested urea and bicarbonate level [Table 2]. can be helpful beside to clinical findings to estimate 560 Annals of Tropical Medicine and Public Health | Nov-Dec 2012 | Vol 5 | Issue 6 Malaki: Laboratory tests in dehydration Table 1: Comparison of laboratory tests among three groups of dehydration there is not difference between mild and moderate dehydration but it was significant between severe and moderate /mild forms. (p significant if less than 0.05) Variables Severe Dehydration severity Mild P value Moderate P value Severe BUN (mg/dl) 27 ±12 0.59 29±13 0.0001* 49±27 Creatinie (mg/dl) 0.6±0.4 0.51 0.6±0.1 0.01* 0.9±0.6 Base deficit (mEq/l) -8±4 0.19 -10±5 0.0001* -17±5 Venous Hco3 (mEq/l) 14.5±4 0.06 16±3 0.0001* 10±3 BUN/Cr 47±16 0.56 49±19 0.047* 64±36 Table 2: Positive likelihood ratio (+LR), sensitivity and beyond-16 has an acceptable specificity (90%) and the specificity of four laboratory tests blood urea nitrogen highest likelihood values (+ LR=6.5) compare to other (BUN), creatinine, venous HCO3, base deficit for tests.