Doi:10.1016/j.jpeds.2004.05.035

DEVELOPMENT OF A CLINICAL DEHYDRATION SCALE FOR USE IN JEREMY N. FRIEDMAN, MBBCH, FRCPC, RAN D. GOLDMAN, MD, RAJENDU SRIVASTAVA, MD, FRCPC, MPH, AND To develop a clinical dehydration scale for use in children <3 years of age.
Prospective cohort study of children between 1 and 36 months of age who presented to a tertiary pediatric emergency department (ED) with gastroenteritis. Children were weighed and scored for 12 clinical signs, were rehydrated, andthen were reweighed and rescored when rehydration was completed. Weight change from pre- to post-rehydration was used toassess criterion validity with independent global assessments of dehydration severity by attending physicians and nurses asmeasures of construct validity. Formal approaches to item selection and reduction, reliability, discriminatory power, validity,and responsiveness were used.
137 children (median age: 18 months) with gastroenteritis were studied. The final dehydration scale consisted of four clinical characteristics: general appearance, eyes, mucous membranes, and tears. The measurement properties were asfollows: validity as assessed by Pearson’s correlation coefficient was 0.36 to 0.57; reliability as assessed by the intra-classcorrelation coefficient was 0.77; discriminatory power as assessed by Ferguson’s d was 0.83; and responsiveness to change asassessed by Wilcoxon signed rank test was significant at P < .01.
Clinicians and researchers may consider this four-item, 8-point rating scale, developed using formal measure- ment methodology, as an alternative to scales developed ad hoc. (J Pediatr 2004;145:201-7) From the Divisions of Pediatric Medicineand Emergency Medicine, Departmentof Pediatrics, University of TorontoFaculty of Medicine and the Hospital Dehydrationsecondarytogastroenteritisremainsamajorcauseofmorbidityand for Sick Children, Toronto, Ontario, mortality. In the United States, 9% of all hospitalizations of children <5 years of age are because of diarrhea and dehydration.One method to measure the magnitude of dehydration is the calculation of percent loss of body weight during the conception and design, acquisition ofdata, interpretation of data, drafting of illness.However, accurate baseline predehydration weights are not usually available to the clinician. Thus, historically, the measurement of dehydration has been based on several supervision of the research team. Dr RanGoldman and Dr Rajendu Srivastava clinical variables scaled into three categories: mild, moderate, and severe dehydrati These categories, and the clinical variables defining each category, have, however, been revision of the manuscript. Dr Patricia C.
Parkin participated in design, data anal- developed in an ad hoc manner rather than by using formal measurement methodology.
ysis, interpretation of data, revision of Further, several investigators have examined signs and symptoms of dehydration without placing these variables into a rating scale.
for Sick Children Research Institute Seed Established methodology for the development of outcome measures (for both acute Grant, and the Physician Services In-corporated Foundation (number 00-15).
and chronic health states) in the clinical setting are relatively recent.Measurement methodology formally considers the following issues relating to scale development: is supported by a grant from the Hospitalfor Sick Children Foundation, Toronto, selection of clinical items to be evaluated, assessment of item and scale validity, inter-rater reliability, discriminatory power, and responsiveness to change. Scales developed in this Submitted for publication July 29, 2003; last way may be used in clinical practice or as outcome variables in clinical research.
revision received Apr 5, 2004; acceptedMay 11, 2004.
The current use of clinical signs and symptoms of dehydration or categories of Reprint requests: Patricia C. Parkin, MD, dehydration (mild, moderate, severe) for which the measurement properties remain FRCPC, Division of Pediatric Medicine,The Hospital for Sick Children, 555 unknown formed the conceptual basis for the a priori development of a clinical scale. In addition, the current use of numerical rating scales in pediatric practice (for example, Apgar Copyright ª 2004 Elsevier Inc. Al rightsreserved.
scale, acute pain scales, Glasgow coma scale) supported the clinician (eg, the Apgar score) and its compatibility with development of a similar numerical rating scale for the clinical discriminating definitions for each characteristic.
measurement of severity of dehydration. Therefore, the Items with a low frequency of endorsement were objective of this study was to develop a clinical dehydration excluded (ie, if $95% of subjects scored 0 on that item).
scale for use in children with gastroenteritis between 1 and 36 Item homogeneity was assessed using the item-total correla- months of age using established measurement methodology.
tion method, with exclusion of those items with a Pearson’scorrelation coefficient of <0.20.
As a methodologic framework, Kirshner and Guyatt’s Weight change from pre- to post-rehydration was used as general approach to the development of a health status index the ‘ gold standard’ to assess criterion validity along with Streiner and Norman’s textbook on health Construct validity was determined using a hypothetical con- measurement scales were used.Validity is defined as the struct developed a priori, ie, there would be a correlation between extent to which a scale measures what it is intended to the clinical dehydration scale and proxy measures of dehydration measuReliability is defined as the extent to which a scale is severity. The proxy measure was a global assessment of reproducible or consistent.Discriminatory power is defined dehydration severity, as rated on a 5-point Likert scale, scored as the ability of a scale to discriminate amongst subjects, ie, independently by the attending emergency nurse and the scores should be spread along the entire possible rang attending emergency physician. Criterion and construct validity Responsiveness or sensitivity to change is defined as the ability were quantified using Pearson’s correlation coefficient. Validity of a scale to detect small but clinically important changes in was assessed for individual items and for aggregate scales.
The reliability of the individual items, as well as of the The intended use of our clinical dehydration scale is as aggregate scale, was assessed by having two observers a discriminative (ie, to assess severity) and evaluative (ie, to independently score children with acute dehydration. An assess response to therapy) tool for use in children between 1 intra-class correlation coefficient was used to quantify re- and 36 months of age, in the emergency department (ED) liability. For the final aggregate scale, a coefficient of $0.60 setting, for whom the attending physician has established the was considered an acceptable level of agree diagnosis of gastroenteritis with dehydration. Subjects meet-ing these criteria were recruited from the ED at the Hospital for Sick Children in Toronto. The Hospital for Sick Children The discriminating ability of the aggregate scale with is a pediatric tertiary care hospital affiliated with the University the highest reliability and validity was assessed using of Toronto, with approximately 50,000 patient visits to the Ferguson’s dThis coefficient is expressed as a value from ED annually. Exclusion criteria were: any cause of dehydration 0 to 1. A minimum value of 0 occurs when all subjects get the other than presumed gastroenteritis and any other chronic same score. A maximum value of 1 occurs when subjects are disease, eg, renal, gastrointestinal, cystic fibrosis, including equally divided among all possible scores (ie, the distribution co-existing malnutrition and treatment with intravenous rehy- dration within the previous 24 hours. Electrolyte measure-ments were conducted at the discretion of the treating physician, and children found to have hyponatremia (serum The responsiveness of the aggregate scale with the sodium <130 mmol/L) or hypernatremia (serum sodium >150 highest reliability and validity was determined by measuring mmol/L) were excluded from the study, as was done by previ- the change in score in children following a treatment of known ous investigators,given the possibility that these patients dis- effiThat is, children receiving rehydration therapy were play unique clinical signs and symptoms.
scored before and after therapy. Because the changes in theordinal scale were likely to be non-Gaussian, responsiveness was quantified using Wilcoxon signed rank test.
To generate items for the clinical scale, the published literature on the clinical and physiologic manifestations ofdehydration was reviewed. In addition, 10 experts in nephrology, The final scale consisted of the combination of items gastroenterology, emergency medicine, and general pediatrics yielding the most valid and reliable score. The final scale was were surveyed. The items were then operationalized. Precise clinical definitions were provided for each item.
All clinical characteristics were measured on a 3-point ordinal scale (0, 1, or 2) (Although it has been found Eligible children were identified, on a convenience basis, that 5- or 7-point scales minimize the loss of information,9 when a study nurse or physician was available. Four research a 3-point scale was chosen because of its familiarity to the nurses (all with a minimum of 5 years of ED experience) and Final four-item clinical dehydration scale is in bold.
*As reported by caregiver.
yMeasured at fingernail bed after applying just right amount of pressure to blanch the nail bed, in a warm room.
zMeasured by pinching the lateral side of the abdominal wall at the level of the umbilicus; mean of 3 measurements.
§Child in the upright position in a quiet state.
kBy history or examination.
{Count for 30 sec and multiply by 2.
**Must be sample from before rehydration initiated.
three physicians were involved in recruiting and scoring the patients. Demographic data were gathered: age, sex, duration of A sample size of approximately 30 patients was illness, presence of diarrhea, vomiting, and fever. The child was calculated to be adequate under the following assumptions: weighed and scored for all clinical items prior to the initiation of the intra-class correlation coefficient for assessing inter-rater rehydration. Whenever a second study nurse or investigator was reliability would be $0.80, a = 0.05, b = 0.20, and 2 available, a second assessment was performed independently.
measurements per subject; and the Pearson’s correlation The ED attending nurse and ED attending physician were both coefficient would be = 0.70, a = 0.05, b = 0.20.An interim approached independently and asked to rate the degree of the analysis to test these assumptions indicated that additional child’s dehydration on a 5-point Likert scale. Rehydration patient enrollment was required to ensure the ability to detect therapy, oral or intravenous, was ordered by the ED staff without an intra-class correlation coefficient of $0.70 and a Pearson’s regard to the research study; patients received therapy as inpatients or in the observation unit in the ED.
The attending ED or inpatient pediatrician determined the time at which rehydration therapy was considered complete, and the child was then weighed and re-scored bythe study personnel. All weights were recorded without clothes or diapers, using the same designated research study A sample of 141 children were enrolled in the study scale (Scale-Tronix Pediatric Scale model 4800, accuracy 5 Of these, 94 children had electrolyte measurement gm, Global Medical Products, Inc, Mississauga, Ontario).
ordered by the attending physician. Two patients had hypo- The percentage weight change was calculated as (final weight- natremia and two had hypernatremia; they were excluded, initial weight)/final weight 3 100.
leaving 137 patients in the study. Of the 137 study patients, 94 The research was approved by the Hospital for Sick (69%) were rehydrated intravenously and 43 (31%) were Children Research Ethics Board and informed parental rehydrated orally; 14 (10%) were admitted to the hospital, and consent was obtained for all subjects.
123 (90%) were managed in the ED and the observation unit.
Development of a Clinical Dehydration Scale for Use in ChildrenBetween 1 and 36 Months of Age Table II. Characteristics of participants at baseline Pre-rehydration weight (kg), median (range) *P value for patients with pre- and post-rehydration weight vs patients with pre-rehydration weight only.
yNumbers do not add up to n due to some missing values.
Table III. Measurement properties of individual items *Pearson’s correlation coefficient.
yIntra-class correlation coefficient.
A dehydrated weight and rehydrated weight was recorded for 102 children (74%), allowing for a calculation A review of the published literaturshowed of percent dehydration for these children. The remaining that the clinical characteristics correlating with dehydration children were either transferred to their regional hospital EDs severity included: reported urine output, general appearance, or were discharged home at a time when a study nurse or capillary refill time, skin turgor, fontanelle, eyes, mucous physician was unavailable. There were no significant differ- membranes, tears, respiratory rate, heart rate, blood pressure, ences between those children with both a dehydrated and and urine specific gravity Of these items, fontanelle, rehydrated weight (n = 102), and those children with urine specific gravity, and blood pressure were excluded a dehydrated weight only (n = 35) on the baseline character- because of a low frequency of endorsement.
istics, including the baseline clinical score ( Of the 102 patients for whom percent dehydration was calculated, the distribution was as follows: 16 patients (16%) Measurement Properties of Individual Items had no dehydration; 50 patients (49%) were >0% to <3% Item-total correlation, quantified using Pearson’s cor- dehydrated (minimal dehydration); 26 patients (25%) were relation coefficient, was calculated for the remaining nine $3% to <6% dehydrated (mild dehydration); 9 patients (9%) items The P value for each of these correla- were $6% to <10% dehydrated (moderate dehydration); and 1 tions was < .01. Validity was measured for each of the nine patient (1%) was $10% dehydrated (severe deh individual items. The Pearson’s correlation coefficient for each The median time to rehydration was 46.5 hours (range, of the individual items as compared with the weight change 6.8-116.6 hours) for the patients admitted to the hospital and from pre- to post-rehydration (criterion validation) was 6.5 hours (range, 1.2-25 hours) for the patients managed in calculated (Inter-rater reliability for each of the nine individual items was calculated, and all items, except for Table IV. Measurement properties of the final four-item clinical dehydration scale *Median pre-rehydration score: 2 (range, 0-8). Median post-rehydration score: 0 (range, 0-2).
general appearance, had an intra-class correlation coefficient explicitly stated methodologic framework to evaluate mea- greater than 0.60, indicating an acceptable level of agreement surement properties of validity, reliability, discriminatory power, and responsiveness. The four items in this scale arepractical and would be easy to teach to a wide range of Measurement Properties of the Final Scale A scale that included all nine items was assessed for There were limitations to this study. Weight gain after validity and reliability. Combinations of items with the rehydration was used to measure percentage dehydration as strongest measurement properties on univariate analyses were this has been considered the gold standard.However, tested. The measurement properties were strongest for a scale the use of a single rehydration weight rather than serial that included the following four items: general appearance, weights differs from previous studies. Gorelick et have eyes, mucous membranes, and tears (). The final four- shown that the median time to achieve a stable weight was 24 item clinical scale had the following properties: validity as hours (range, 12-72 hours) for patients admitted to the assessed by the Pearson’s correlation coefficient was 0.36 hospital. Of the 10% of patients admitted to hospital in this to 0.57; reliability as assessed by the intra-class correlation study, the median time to rehydration exceeded this. The time coefficient was 0.77; discriminatory power as assessed by required to achieve rehydration for patients managed in the Ferguson’s d was 0.83; and responsiveness to change as outpatient setting is not known. Given the potential assessed by Wilcoxon’s signed rank test was statistically limitations of weight gain following rehydration as a gold significant at P < .01. At baseline, the median clinical score standard to assess validity, in this study we assessed validity was 2 (range, 0-8; n = 126), and the median clinical score using two additional constructs, ie, the assessment of decreased to 0 (range, 0-2; n = 33) following rehydration dehydration severity scored independently by the attending emergency nurse and the attending emergency physician.
Finally, not all patients had a post-rehydration weight recorded. However, there was no evidence that this group differed significantly with respect to baseline characteristics, In this group of 137 children with a median age of 18 including baseline clinical score (clinical score at baseline was months and a diagnosis of gastroenteritis, the following four items in aggregate had the most significant measurement We excluded children known to have hypo- or hyper- properties for dehydration: general appearance, eyes, mucous natermia (4 of 94, approximately 4%) given that these children membranes, and tears. Many clinicians currently utilize may manifest different clinical signs. Electrolytes were not a categorical rating scale (mild, moderate, severe) to measure measured in 48 children, raising the possibility that some the severity of dehydration and to guide decision making and children with abnormal electrolytes were included in the study.
therapy. The measurement properties of this informally However, it is likely that these children represented <2 % of developed scale remains unknown. Clinicians and researchers the entire study sample. In addition, some dehydrated children may now consider this four-item, 8-point numerical rating may have been excluded from the study if they did not scale as an alternative, given its known measurement manifest the clinical signs commonly used by clinicians to establish the diagnosis of dehydration. However, it is likely The strengths of this study include a restricted age range that these signs are uncommon and may have been eliminated (<3 years) in the population studied, with exclusion of patients as a result of low frequency of endorsement.
with dehydration secondary to causes other than acute The scale is intended to discriminate among and gastroenteritis as well as those with hyponatremic and evaluate the response to therapy in children for whom the hypernatremic dehydration. Precise definitions for each of diagnosis of gastroenteritis with dehydration has been the clinical items were developed a priori. A small group of established. Assessing the diagnostic accuracy (ie, the ability experienced nurses and physicians were responsible for the to establish the diagnosis of dehydration) was not the aim of clinical scoring. The clinical scale was developed using an this study. However, a post hoc analysis identified a sensitivity Development of a Clinical Dehydration Scale for Use in ChildrenBetween 1 and 36 Months of Age of 0.85 (95% confidence interval: 0.73, 0.97) for an ‘‘abnormal for moderate dehydration (6%-10%), and 70% for severe de- score’’ (ie, a score $1) as compared with the gold standard (ie, hydration (>10%). Apart from the wide age range and possible dehydration of $3%). Further evaluation of the scale, in a large selection bias, other issues included the use of different population may allow further understanding of its diagnostic weighing scales for measurement of pre- and post-rehydration weight as well as the timing of the ‘‘well’’ weight.
There have been few published studies examining the The four clinical items with the strongest measurement symptoms and signs of dehydration using subsequent weight properties identified in the current study show significant recovery as the criteria by which to measure the degree of overlap with those identified by Gorelick et al and Duggan dehydration. Gorelick et alconducted a prospective cohort et These similarities provide additional evidence of the study to evaluate the performance of common clinical signs in dehydration. The measurement methodology used by these We are indebted to the research nurses for their help with patient investigators differed from that used in our study. The enrollment and data collection: Pauline Mathews, Audrey Bell-Peters, investigators found that each of the 10 individual clinical signs Jane Ciordes, and Deborah Cutler. We also thank the emergency studied had a low sensitivity and high specificity. The inter- department physicians and nurses who participated in the study.
observer reliability for individual signs was generally good toexcellent (j statistic range, 0.4-0.75). Using logistic regressionmodeling, the investigators found that 4 of the 10 clinical signs (capillary refill >2 seconds, absent tears, dry mucous mem- Liebelt EL. Clinical and laboratory evaluation and management of branes, and ill general appearance) were independently children with vomiting, diarrhea and dehydration. Curr Opin Pediatr 1998;10:461-9.
associated with dehydration and together predicted dehydra- Adelman RD, Solhaug MJ. Pathophysiology of body fluids and fluid tion as well as the entire set. The investigators were able to therapy. In: Behrman RE, Kliegman RM, Jenson HB, eds. Nelson textbook demonstrate that the presence of any two or more of these of pediatrics. 16th ed. Philadelphia: WB Saunders; 2000. p. 211-5.
clinical signs indicates a fluid deficit of at least 5%.
Travis LB. Disorders of water, electrolyte and acid-base physiology. In: MacKenzie et prospectively studied the clinical signs Rudolph AM, Hoffman JIE, Rudolph CD, eds. Rudolph’s pediatrics.
20th ed. Stanford (CT): Appleton and Lange; 1996. p. 1320-7.
and symptoms of 102 children <4 years of age with acute Gorelick MH, Shaw KN, Murphy KO. Validity and reliability of clinical gastroenteritis, who were assessed by admitting junior doctors signs in the diagnosis of dehydration. Pediatrics 1997;99:e6.
to be $5% dehydrated on the basis of history and physical MacKenzie A, Barnes G, Shann F. Clinical signs of dehydration in examination. The investigators found the main clinical indicators of mild to moderate dehydration ($4%) to be Duggan C, Refat M, Hashem M, Wolff M, Fayad I, Santosham M.
How valid are clinical signs of dehydration in infants? J Pediatr Gastroenterol decreased peripheral perfusion, deep breathing, and decreased skin turgor. In this study, the clinical variables were not clearly Vega RM, Avner JR. A prospective study of the usefulness of clinical and defined a priori, and no attempt was made to examine inter- laboratory parameters for predicting percentage of dehydration in children.
rater reliability or responsiveness of the clinical findings. Other issues to note are that physicians-in-training were used for the Kirshner B, Guyatt G. A methodological framework for assessing health clinical assessments and only hospitalized patients were Streiner DL, Norman GR. Health measurement scales: a practical guide to their development and use. 1st ed. Oxford (UK): Oxford University Press; Duggan et alperformed a prospective cohort study examining boys 3 to 18 months of age with acute diarrhea, to 10. McDowell I, Newell C. Measuring health: a guide to rating scales and determine the ability of several clinical signs to distinguish questionnaires. 2nd ed. New York: Oxford University Press; 1996.
11. Feinstein AR. Clinimetrics. New Haven (CT): Yale University Press; among degrees of dehydration. This was part of a clinical trial comparing two types of oral rehydration solution. This study 12. Guyatt GH, Deyo RA, Charlson M, Levine MN, Mitchell A.
used two different empirically based clinical dehydration Responsiveness and validity in health status measurement: a clarification.
scales; one developed by Fortin and Pareand the other by Santosham et They found that prolonged skinfold, 13. Teach SJ, Yates EW, Feld LG. Laboratory predictors of fluid deficit inacutely dehydrated children. Clin Pediatr 1997;36:395-400.
altered neurologic status, sunken eyes, and dry oral mucosa 14. Saavedra JM, Harris GD, Li S, Finberg L. Capillary refilling (skin were the clinical signs that correlated best with the percent turgor) in the assessment of dehydration. AJDC 1991;145:296-8.
dehydration. Inter-rater reliability was not evaluated, and the 15. Laron Z, Crawford JD. Skin turgor as a quantitative index of definition of clinical findings was subjective. The amount of dehydration in children. Pediatrics 1957;19:810-5.
rehydration fluid given to the subjects in this trial was based on 16. Kramer MS, Feinstein AR. Clinical biostatistics LIV. The biostatisticsof concordance. Clin Pharmacol Ther 1981;29:111-23.
the investigators initial clinical assessment and, therefore, may 17. Ferguson GA. On the theory of test discrimination. Psychometrika have had an effect on their ultimate rehydration weight.
Vega and Avnerincluded children <18 years of age who 18. Deyo RA, Diehr P, Patrick DL. Reproducibility and responsiveness of were receiving intravenous fluid for rehydration in their study health status measures. Statistics and strategies for evaluation. Control Clin on the usefulness of clinical parameters for predicting Trials 1991;12:142S-58S.
19. Donner A. Sample size requirements for reliability studies. Stat Med percentage dehydration. The physician’s clinical estimate of dehydration compared with percent loss of body weight had 20. Lachin JM. Introduction to sample size determination and power a sensitivity of 74% for mild dehydration (5% or less), 33% analysis for clinical trials. Control Clin Trials 1981;2:93-113.
21. Anonymous. Practice parameter: the management of acute gastroenter- 25. Anonymous. A manual for the treatment of diarrhoea. In: World Health itis in young children. American Academy of Pediatrics, Provisional Organisation, Programme for Control of Diarrhoeal Diseases. World Health Committee on Quality Improvement, Subcommittee on Acute Gastroenter- 26. Gorelick MH, Shaw KN, Baker MD. Effect of ambient temperature on 22. Fortin J, Parent MA. Dehydration scoring system for infants. Trop capillary refill in healthy children. Pediatrics 1993;92:699-702.
Pediatr Environ Child Health 1978;24:110-4.
27. Harvey E. Fluids and electrolytes. In: Shefler AG, editor. The HSC 23. Santosham M, Brown KH, Sack RB. Oral rehydration therapy and handbook of pediatrics. 8th ed. St Louis (MO): Mosby–Year Book; 1992. p.
dietary therapy for acute childhood diarrhea. Pediatr Rev 1987;8:273-8.
24. Fuchs SM, Jaffe DM. Evaluation of the ‘‘tilt test’’ in children. Ann 28. Lowrey GH. Growth and Development of children. 8th ed. Chicago (IL): The Yearbook Medical Publishers; 1986. p. 246–77.
50 Years Ago in The Journal of Pediatrics TETRACYCLINE: STUDIES ON ABSORPTION, DISTRIBUTION, EXCRETION AND CLINICAL TRIAL INCHILDREN Schwarzer S, Reeves R, Claps A, Anderson A. J Pediatr 1954;43:285-92 Chlortetracycline was isolated from Streptomyces aureofaciens in 1944, and in 1953 tetracycline was derived from chlortetracycline by chemists at Pfizer and Lederle pharmaceuticals. Chlortetracycline, which was not developed forhumans because of toxicity concerns, is still an additive to animal feed. Schwarzer and colleagues performeda pharmacokinetic study of tetracycline in infants and children seeking to derive dosage guidelines. Recognizing thatchildren are not ‘‘miniature adults,’’ studies of this type were and are needed. In 1994 only five of the 80 drugs mostcommonly administered to infants and children were labeled for pediatric use. In response to this need, the NationalInstitute of Child Health and Human Development, under the leadership of Dr Duane Alexander, the vision of DrSumner Yaffe, and the hard work of Drs George Giacoia and Gilman Grave, established a Network of PediatricPharmacology Research Units. These units obtain data that place drug administration to children on a scientific basis.
Recently the Food and Drug Administration added funds to the program recognizing the importance of performing suchstudies with child-friendly protocols with minimal risk to children.
Tetracyclines are deposited in calcifying areas of bones and teeth, and cause discoloration in a dose-dependent manner.
Deciduous tooth discoloration occurs if the infant receives tetracycline in utero after 14 weeks’ gestation, and up to the ageof 3 months. Mineralization of permanent teeth starts at about 6 months of age and concludes at about 6 years; repeatedadministration of tetracycline during this period produces life-long discoloration. Tetracycline is still one of the first lineantibiotics for plague and tularemia, and there is minimal drug toxicity associated with this use in children.
One interesting feature of the Schwarzer article is that most of the infectious diseases for which the children received tetracycline are those we now know as viral illnesses. Our current understanding of upper respiratory infections placeantimicrobic therapy on a more scientific footing. However, in another 50 years we may learn that our current approach toantimicrobic therapy is equally primitive.
Development of a Clinical Dehydration Scale for Use in ChildrenBetween 1 and 36 Months of Age

Source: http://pedneph.info/Fluids%20Talk/DW%20fluids/DW%20fluids%20PDF/Development%20of%20a%20clinical%20dehydration%20scale%20for%20use%20in%20children%20between%201%20and%2036%20months%20of%20age.pdf