The Postoperat ive Response of the Term and Preterm Newborn Infant to Sodium Administrat ion

By Thomas M. Krummel, David A. Lloyd, and Marc I, Rowe Pittsburgh, Pennsylvania

�9 Twenty surgical newborn infants aged less than 5 days were selected for study to determine the quantity of sodium administered during operation and the subsequent 48 hours, and to determine their response to this sodium load. There were 6 preterm infants with gestational age 35 weeks or less and 14 full-term infants aged more than 35 weeks. Measurements calculated at 12 hourly intervals for 48 hours after operation included sodium intake, sodium excretion, fractional excretion of sodium, and serum sodi- um. No special guidelines were utilized for fluid manage- ment. The mean sodium intake during the 48-hour study period for the term infants was 15.7 mEq/kg, of which 46% was given during the first 12 hours, and for the premature infants was 17.1 mEq]kg, of which 56% was given during the first 12 hours. In the two groups of patients, the amount of sodium given during the first 12 hours was 470% and 480% of their estimated maintenance requirements, respectively. The mean sodium output during the first 12 hours was low in the term group (1.2 mEq/L) and the premature group (1.3 mEq/L), and subsequently increased reaching maximum levels of 2.3 and 2.1 mEq/L, respec- tively, by 36 hours. The fractional excretion of sodium exceeded 1.0% in 53% of the term and 94% of the preterm infants. During the study period, the mean serum sodium levels exceeded 145 mEq/L (hypernatremia) in 64% of the term and 67% of the preterm infants. We conclude that the observed hypernatremia was primarily the result of exces- sive sodium administration during and immediately after operation, compounded by a slightly reduced ability to excrete sodium and a short period of postoperative sodium retention. We recommend reevaluation of the current approach to sodium management to avoid hypernatremia and its potential morbidity. �9 1985 by Grune & Stratton. Inc.

INDEX WORDS: Postoperative sodium excretion; neonatal sodium excretion.

E FFECTIVE management of sodium balance in the newborn surgical patient is difficult because

of the unique characteristics of neonatal renal function and the physiologic and pathophysiologie changes induced by anesthesia and operative trauma. The premature newborn infant under 35 weeks gestational age has a high sodium excretion irrespective of sodium load and tends to be a "salt-waster. ''~'2 Further, the preterm and term infant have a limited capacity to excrete a sodium load? '4 In the adult, there is evidence that the metabolic response to trauma includes sodium retention. 5'6 Wilkinson's 7 data suggest that this also occurs in the newborn infant, though this is controver- sial. s'9 To further complicate matters, the work of Moyer 5 and Shires ~~ suggest that operative trauma causes the sequestration of large quantities of water-

and sodium-containing fluids in the extracellular fluid space, the "third space."

In spite of the conflicting data, fluids containing high concentrations of sodium are freely administered intraoperatively and during the immediate postopera- tive periodJ M3 The purpose of this study was to determine the quantity of sodium actually adminis- tered during operation and the early postoperative period to surgical newborn infants and to determine their response to this sodium load. From this data, we hope to place sodium administration in the newborn surgical infant on a more phySiologic basis.

MATERIALS AND METHODS

Patients

Infants were selected according to the following criteria: (1) age less than 5 days; (2) normal renal function; (3) no radioopaque contrast materials had been administered either intravenously or via the gastrointestinal tract; (4) no diuretic therapy had been used; (5) infants had not received prolonged ventilatory care associated with episodes of hypoxia; (6) the surgical lesions should not have resulted in large losses of "third space" fluids and electrolytes. Twenty patients were selected and divided into two groups on the basis of gestational age. A number of studies have shown that urinary sodium excretion changes at 35 weeks gestational age. ~4-~ There- fore, we assigned infants aged 35 weeks and under into the "preterm" group and those over 35 weeks gestational age into the "term" group. The diagnosis for these infants is indicated in Table 1. The study period spanned the first 48 hours after operation.

Measurements

Sodium Intake. Sodium intake was calculated after operation at 12 hourly intervals as mEq/kg/12 hr. Adjustments were made for nasogastric losses, chest drainage, and blood losses.

Sodium Excretion. Urine sodium levels and urine volume were measured at 12 hourly intervals. Urine sodium excretion was calcu- lated in mEq/kg/12 hr.

Fractional Excretion of Sodium. The fractional excretion is a measure of tubular sodium reabsorption and indicates the capacity of the kidney for excreting and absorbing sodium. In the normal

From the Department of Pediatric Surgery, University of Pitts- burgh School of Medicine, Children's Hospital of Pittsburgh, Pittsburgh.

Presented before the 16th Annual Meeting of the American Pediatric Surgical Association, Kohala Coast, Hawaii, May 14, 1985.

Address reprint requests to Marc L Rowe, MD, Department of Surgery, Children's Hospital of Pittsburgh, 125 DeSoto St, Pitts- burgh, PA 15213.

�9 1985 by Grune & Stratton, Inc. 0022-3468/85/2006-0046503.00/0

Journal of Pediatric Surgery, Vol 20, No 6, (December), 1985: pp 803-809 803

804 K R U M M E L ET AL

Table 1. Diagnosis

Term Preterm

Small bowel a t res ia/s tenosis 3 2

Omphalocele 2 - -

Gastroschisis 2 3

Cyst ic adenomato id mal format ion 1 - -

Sacrococcygeal te ra toma 1 - -

Esophageal atresia wi th f is tu la 3 1

Mal ro ta t ion 1 - -

Imperforate anus (high) 1 - -

adult with a glomerular filtration rate of 120 mL/min and daily urine sodium excretion of 120 mEq, the fractional sodium excretion is usually 0.5%. Sodium and creatinine levels were simultaneously measured in serum and urine every 12 hours. The fractional sodium excretion was then calculated using the following formula:

urine sodium/serum sodium FeNa = x 100.

urine creatinine/serum creatinine

Serum Sodium. The serum sodium was determined every 12 hours. There is disagreement about the normal range of serum sodium in term and preterm infants. For this study, we have defined the normal range for term and preterm infants as 135 to 144 mEq/L. Serum sodium levels of 145 mEq/L or greater were regarded as indicating hypernatremia.

Fluid and Electrolyte Management No special guidelines were developed for fluid management for

this study. Current approaches to operative and postoperative fluid and electrolyte management employed by the Departments of Anesthesiology and Surgery were used.

Intraoperative Fluid and Electrolyte Management Intraoperative therapy was directed by the anesthesiologist."

There were three components: maintenance requirements, preopera- tive fluid deficit replacement, and intraoperative losses; the opera- tive losses consisted of "third space" losses and blood loss. Mainte- nance fluid was given as 5% dextrose in 0.2 normal saline (D5 0.2 NS) containing 34 mEq sodium throughout the operative procedure at a rate of 4 mL/kg/hr (96 mL/kg/24 hr). Estimated fluid deficit, if present, was replaced with 5% 0.2 NS.

Intraoperative third space fluid and electrolyte losses for intraab- dominal operations were estimated to vary between 6 and 10 mL/kg/hr, and for intrathoracic operations between 4 and 7 mL/ kg/hr. These losses were replaced with Ringer's lactate solution. Based on clinical evaluation of the patient by the anesthesiologist, if hypovolemia was suspected, additional Ringer's lactate solution was administered. Blood loss was measured in all patients. Blood loss in excess of 20 mL/kg was replaced with equal volumes of packed red blood cells. Smaller losses were replaced with Ringer's lactate solution in volumes one to two times the estimated blood loss, depending upon the anesthesiologist. In this study, only one patient required blood transfusion.

Postoperative Fluid and Electrolyte Management An initial assessment of the infant's fluid and electrolyte status

was made at the end of the operation. This was based on the underlying disease process, operative findings, intraoperative fluid management, clinical evaluation of the patient, and pertinent labo- ratory studies. The conceptional age of the infant was also taken into consideration. The maintenance fluids used were 0.2 normal saline in

5% or 10% dextrose when the patient was normovolemic and normonatremic. When the patient was hyponatremic or when large "third space" fluid and electrolyte losses were anticipated, 0.,15 normal saline with dextrose was given. Ringer's lactate solution in equivalent concentrations was used when the serum chloride level was elevated. The usual volumes of fluid administered ranged from 80 to 180 mL/kg/24 hr. Adjustments were made every 4 hours based on serial clinical and laboratory evaluations of the patient. Nasogastric fluid losses were replaced with equal volumes of 0.45 normal saline and chest tube drainage was replaced with equal volumes of plasma.

RESULTS

Sodium Intake

Term infants received a mean of 7.1 mEq sodium/ kg body weight during operation and the first 12 hours postoperation (Fig 1). This was 470% of the estimated sodium maintenance requirement for this period (1.5 m E q / k g / 1 2 hr). 3 Between 12 and 48 hours after operation, the sodium intake declined sequentially to near maintenance levels. The mean total sodium intake for term infants was 15.7 m E q / k g during the 48-hour period, of which 46% was given during the first 12 hours.

The preterm infants received a mean of 9.6 mEq sodium/kg body weight during operation and the subsequent 12 hr. This was 480% of the estimated sodium maintenance requirement (2.0 m E q / k g / 1 2 hr). 2 Between 12 and 48 hours after operation the sodium intake decreased to maintenance levels. The mean total sodium intake for the preterm infants was 17.1 m E q / k g , of which 56% was administered during the first 12 hours.

There was no statistical difference between the sodium intake of the term and preterm infants during any of the 12-hour periods.

Sodium Output

In the term group, the mean sodium output was 1.2 m E q / k g during the first 12 hours following operation and increased to a maximum of 2.3 m E q / k g between 24 and 36 hours postoperation (Fig 2). This increase is statistically significant (P < 0.02).

In the preterm group, the mean sodium output was 1.3 m E q / k g during the first 12 hours after operation and increased to a maximum of 2.1 m E q / k g by 36 hours. This increase is not statistically significant (P < 0.27). Af te r 24 hours the sodium outputs exceeded the estimated sodium maintenance require- ments in both the full term and preterm infants.

Although the average patient in both groups had a low sodium excretion during the first 12 hours after operation, 2 of 14 (14%) of the term infants and 2 of 6 (33%) of the preterm infants had increased sodium excretion varying between 2.7 and 3.8 m E q / k g / 1 2 hr.

POSTOPERATIVE RESPONSE OF INFANTS TO SODIUM

T E R M P R E T E R M

2:

E

o § z

OP 12 24 36 48 OP 12 24 36 48

IN H O U R S IN H O U R S

Fig 1. Sodium intake.

805

TERM PRETERM

:3.0 - 3 .0 -

2 . 0 - 2 . 0

E

1.0 1 .0

% Z

d b I l , . " i ' ' u v

OP 12 24 36 48 OP 12 24 36 48

IN HOURS IN HOURS

IV MAINT. meq /kg /12hr . I

Fig 2. Sodium output�9

806 KRUMMEL ET AL

Fractional Sodium Excretion

The fractional sodium excretions of the term infants ranged from 0.2% to 10.4% (Fig 3). Thirty-seven percent of the determinations were below 1.0%. In contrast, the fractional sodium excretions of the pre- term infants had a narrower range, namely, 0.2% to 5.6%, and only 6.2% of the determinations were less than 1.0%.

Serum Sodium

Serum sodium levels in the term infants varied between 130 and 149 mEq/L . During the 48-hour study, 64% of the term infants had at least one serum sodium level of 145 m E q / L or greater, and only 14% had serum sodium levels below 135 mEq during this period.

In the preterm infants, serum sodium levels varied between 134 and 152 mEq/L . During the study period, 67% of the preterm infants had at least one serum sodium level of 145 m E q / L or more, whereas only 16% had sodium levels below 135 m E q / L during the same period.

DISCUSSION

This study demonstrates that our term and preterm newborn surgical patients received large sodium loads during operation and the immediate operative period. Additionally it demonstrates that, in the face of this

large load, most patients developed at least one episode of elevated serum sodium. There are three possible explanations for the hypernatremia observed: (1) sodium intake was appropriate but the ability of the kidney to excrete sodium was reduced; (2) sodium intake was appropriate but there was inhibition of sodium excretion as a result of anesthesia and surgical trauma; (3) sodium intake was excessive.

The normal adult, when faced with a sodium load, can excrete sodium in excess of 9 mEq/kg /24 hr. 18 Our study demonstrates that, beyond 12 hours after opera- tion, both term and preterm infants are capable of mounting a response that approaches that observed in normal adults. Levels of urine sodium as high as 7.6 mEq/kg/24 hr were observed in term newborn infants and 5.8 mEq/kg /24 hr in preterm infants. These data suggest that the principal reason for the high serum sodium levels is not that the newborn kidney is unable to respond to a sodium load.

The second explanation for the hypernatremia observed in our patients is that the newborn infant mounts a metabolic response to major surgical trauma similar to that described in adults, in which there is an initial catabolic phase manifest by marked sodium retention. 5 This is thought to be a neuroendocrine mediated response. 19-21 We noted that in postoperative newborn infants urine sodium excretion was lowest during the first 12 hours, averaging 1.2 mEq/kg in the term infants and 1.3 mEq/kg in the preterm infants,

37%

TERM

6.2%

[ ] % DETERMINATION _< 1.0 Fig 3. Fractional sodium excretion.

POSTOPERATIVE RESPONSE OF INFANTS TO SODIUM 807

and doubled over the next 12 hours. This response was observed in 86% of the term infants and in 67% of the preterm infants suggesting that the hypernatremia observed in these infants could in part have been the result of the metabolic response to trauma. However, this period of sodium retention did not extend beyond 12 hours after operation and therefore is unlikely to be the major reason for the hypernatremia.

The third suggestion, that sodium intake was exces- sive, appears to be the likely explanation for the observed hypernatremia. Aperia 3 has suggested that the normal maintenance requirement for sodium is 3 mEq/kg/24 hr for term infants. The mean total sodium intake during the 48-hour period of this study was 15.7 mEq/kg for the fullterm infants, which was 260% of the estimated maintenance requirement. Of particular importance is the fact that 46% of the sodium load was administered during the first 12 hours after operation, the period of maximum sodium reten- tion. Similarly, the preterm infants had a mean sodium intake of 17.1 mEq/kg during the 48-hour study. According to A1-Dahhan, 2 the maintenance require- ment for premature infants less than 35 weeks gesta- tional age is 4 mEq/kg/24 hr. Therefore, over the 48-hour study period, our preterm infants received 215% of their estimated maintenance requirement. Of this amount, 56% was administered during the first 12 hours, again the period of maximum sodium reten- tion.

We believe that with the present intraoperative and postoperative sodium regimen hypernatremia will develop in the majority of infants, principally due to an excessive load compounded by a slightly reduced abil- ity to excrete sodium and a brief period of postopera- tive sodium retention. In the light of these findings, we believe that the current approach to sodium manage- ment should be re-evaluated, because an increase in sodium may lead to water retention, which is particu- larly hazardous to preterm infants. An increase in extracellular fluid may predispose to patent ductus arteriosus with congestive heart failure and left ven-

tricular failure, 22'23 and may contribute to the develop- ment of necrotizing enterocolitis 24 and pulmonary dys- plasia.

We conclude that the current practice of giving large sodium loads intraoperatively to replace actual and anticipated losses should be modified. Solutions of half strength normal saline or Ringer's lactate are recommended for initial management. Since the neo- natal response is variable, depending on gestational age and the length of time after operation, the best method of estimating sodium requirements is a feed- back system. On the basis of laboratory data, sodium levels are estimated and a tentative management pro- gram is developed. This is instituted and serum and urine sodium levels are estimated every 4 to 8 hours. If the serum sodium is high, regardless of the urine sodium, excess sodium has been given. If the serum sodium is normal and the urine sodium increased, excess sodium has been administered and renal excre- tion has increased to compensate for this. If the serum sodium level is decreased or normal and the urine sodium level is low, this suggests that sodium intake has been inadequate.

The problem is how best to monitor urine sodium excretion. Random urine sodium levels do not take into consideration the infant's weight or urine volume, but are a useful guide to sodium requirements. In our study, the urine sodium levels ranged from 5 to 125 mEq/L. Fractional excretion of sodium is a measure of the capacity of the kidney to excrete sodium but does not provide the actual amount excreted because urine volume is not taken into consideration in this calcu- lation. The most reliable method is estimation of the renal sodium excretion over a fixed period of time. To be meaningful, a urine collection for at least 4 to 8 hours is recommended. From this, the 24-hour urine excretion of sodium can be estimated. In our study we found that in fullterm newborn infants the urine sodium excretion ranged from 0.2 to 7.6 mEq/kg /24 hr, and in the preterm infants the minimum and maximum values were 0.8 to 5.8 mEq/kg /24 hr.

REFERENCES

1. Engelke SC, Shah BL, Vasan U, et al: Sodium balance in very low-birth-weight infants. J Pediatr 93:837-841, 1978

2. AI-Dahhan J, Haycock GB, Chantler C, et al: Sodium homeo- stasis in term and preterm neonates. Arch Dis Child 58:335-342, 1983

3. Aperia A, Broberger O, Thodenius K, et al: Renal control of sodium and fluid balance in newborn infants during intravenous maintenance therapy. Acta Paediatr Scand 64:725-731, 1975

4. Aperia A, Broberger O, Herin P, et al: Postnatal control of water and electrolyte homeostasis in pre-term and full-term infants. Acta Paediatr Scand 305:61-65, 1982 (suppl)

5. Moyer CA: Acute temporary changes in renal function asso- ciated with major surgical procedures. Surg 27:198-207, 1950

6. Moyer CA: Fluid Balance: A Clinical Manual. Chicago, Yearbook, 1952

7. Wilkinson AW: Body Fluids in Surgery (ed 3). Edinburgh, ES Livingstone, 1969

8. Colle E, Paulsen EP: Response of the newborn infant to major surgery. Pediatrics 23:1063, 1959

9. Knutrud O: The water and electrolyte metabolism in the newborn child after major surgery, in Norwegian Monographs on Medical Science. Oslo, Scand University Books, Universitets Forla- get, 1965

10. Shires T, Williams J, Brown F: Acute change in extracellular fluids associated with major surgical procedures. Ann Surg 154:803-810, 1961

808 KRUMMEL ET AL

11. Bennett E J: Fluid balance in the newborn. Anesthesio! 43:210-224, I975

12. Downes JJ, Raphaely RC: Anesthesia and intensive care in Ravitch, Welch, Benson et al (eds): Pediatric Surgery (ed 3). Chicago, Year Book, 1979

13. Cook DR, Rowe MI: Parenteral fluid therapy, in Cook DR, Marcy JH (eds): Neonatal Anesthesia. Pasedena, Appleton (in press)

14. Arant BS: Developmental patterns of renal function matura- tion compared in the human neonate. J Pediatr 92:705-712, 1978

15. Sulyok E, Varga F, GySry K, et al: On the mechanism of renal sodium handling in newborn infants. Biol Neonate 37:75-79, 1980

16. Aperia A, ZetterstrSm R: Renal control of fluid homeostasis in the newborn infant. Clin Perinatol 9:523-533, 1982

17. Espinol CH: The FeNa test. JAMA 236:579-58l, I976 18. West JB (ed): Best and Taylor's Physiological Basis of

Medical Practice (ed 11). Baltimore, Williams & Wilkins, 1985

19. Moore FD, Ball MR: Metabolic Response to Surgery. Springfield, Ill, Thomas, 1952

20. Moore FD: Metabolic Care of the Surgical Patient. Philadel- phia, Saunders, 1959

21. Kinney JM, Moore FD: Surgical metabolisms in metabolism of body fluids, in Bland JB (ed): Clinical Metabolism of Body Water and Electrolytes. Philadelphia, Saunders, 1963

22. Stevenson JG: Fluid administration in the association of patent ductus arteriosus complicating respiratory distress syndrome. J Pediatr 90:257-261, 1977

23. Bell EF, Warburton D, Stonestreet BS, et al: Effect of fluid administration on the development of symptomatic patent ductus arteriosus and congestive heart failure in premature infants. N Engl J Med 302:598-604, 1980

24. Bell EF, Warburton D, Stonestreet BS: High volume fluid intake predisposes premature infants to necrotising enterocolitis. Lancet 2:90, 1979

Discuss ion

Arnold Coran (Ann Arbor): In 1952, Dr Moore described the metabolic response of the adult to sur- gery, and he compiled all his data between those next 7 years and described them in his classical textbook, The Metabolic Care of the Surgical Patient. In that book, he clearly described the response as being one of sodium and water retention in the adult after surgery, or after any other major trauma, due to the increased secretion of aldosterone and ADH. Working in his lab at that time was Peter Rickham, who in 1959 also wrote a book, entitled The Metabolic Response of the Neonate to Surgery, based on nine neonates, in which he showed that a metabolic response did not occur in the newborn. Thus, the controversy arose. In 1963, Dr Ola Knutrud in Norway had the opportunity, over a 10-year period, to repeat the work that Rickham did, on a much larger group of babies, and clearly showed that the neonate did respond the same way as the adult did to a major metabolic stress--namely, a major operation--in addition to the stress that the baby experienced from being born. The results of all these studies led all of us in the late 1950s and 1960s to believe that our postoperative fluid orders should be limited to water and should almost completely elimi- nate sodium. So, the typical adult after a gastrectomy or gallbladder operation got 1,500 cc of D5W. Unfor- tunately, in the early 1960s Tom Shires did a series of studies in dogs, using radioactive sulfate, to demon- strate that there was marked sequestration of extracel- lular fluid during operation and immediately postoper- atively; and out of that came the major change that all of us experienced in the early 1960s, to administer tremendous amounts of water and sodium during anesthesia and after surgery, mainly by the anesthesi-

ologist, in spite of the fact that a couple of years later all his data was completely refuted because the sulfate space turned out to be inaccurate when you used a 20-minute equilibration period. And here we have an example of that phenomenon--excessive administra- tion of sodium, in this case to a neonate, but the same phenomenon has occurred in adults. I think Dr Rowe and his colleagues have clearly demonstrated that anesthesiologists are still convinced that the neonate is exposed to a tremendous amount of sequestration during anesthesia and needs a lot of Ringer's lactate or normal saline; and I think this clearly has shown me and, I hope, the audience, that that is probably an inappropriate approach.

One of the most interesting pieces of information from the presentation is the very low urinary sodium that these preterm and term babies excreted. This means, to me, not only that the baby may be unable to completely excrete a sodium load, but in addition, that he or she is very capable of retaining sodium, which means, again, that she is capable of harboring a metabolic response to trauma with an increased secretion of aldosterone and ADH, and that he can end up, therefore, with minimal amounts of sodium in his urine, as the adult does, in the range of 1 to 2 mEq/L. In regard to that, I wonder whether Dr Rowe and the group measured ADH or aldosterone in the serum in any of these patients.

I have one question for the authors. They state that if the serum sodium is low and the urine sodium is low, sodium intake has been inadequate and should be given. Since the normal metabolic response to surgery in the neonate is dilutional hyponatremia, which is manifested by a low serum sodium, a low urinary

POSTOPERATIVE RESPONSE OF INFANTS TO SODIUM 809

sodium, and a high total body sodium, and, therefore, indicates that sodium should be withheld, how do they account for that in their recommendations?

Arthur Cooper (New York): This paper is an impor- tant contribution to our understanding of the postoper- ative metabolic response of the neonate. The authors have shown that the term infant is eminently capable of retaining sodium following operation, thus confirm- ing the earlier studies of Knutrud; they have also shown that the preterm infant has a respectable, if lesser, ability to retain sodium in the face of operative trauma. Moreover, despite massive perioperative salt loading, both the term infant, and the preterrn infant to a lesser extent, appear to overcome their presumed inabilities to handle large sodium loads, in that they can increase their sodium outputs within 48 hours of operation to levels which exceed any reasonable sodium intake. Unfortunately, the authors present no preoperative data with respect to the fractional excre- tion of sodium, so we cannot be certain if these really remarkable capabilities are the result of the neuroen- docrine changes which accompany operation, or whether they are intrinsic; regardless, surgical neo- nates will, fortunately, resist our efforts to overload them with salt. However, the fact that sodium output increases progressively following operation does not

necessarily imply that sodium is not retained, as the authors state; indeed, the adrenocortical response is known to persist for several days following operation, so sodium retention should be expected to persist for much longer than the 12 hours the authors suggest.

Thomas Krummel (closing): Dr Coran beautifully summed up the problem that faces pediatric surgeons and, in particular, anesthesiologists, in terms of how much sodium to administer; and we couldn't agree more that more than adequate sodium is administered. We don't have measurements of aldosterone or ADH at the present time, though that's an area of obvious interest. And with regards to the treatment of hypo- natremia, our thought is that severe hyponatremia is not particularly better for the patient than hyperna- tremia and, therefore, when serum sodium falls below 130, steps should be taken to correct it.

Dr Cooper, as you indicated and as we agree, this is a clinical study, not a balanced study; and therefore, we don't have as much information in terms of the actual sodium retention that occurs. Clearly, in all patients the urine outputs were satisfactory. Rather than measuring specific gravities, we found that urine osmolarity is a more helpful indication in terms of the state of hydration of the patient.