November 1978 8 3 4 The Journal of P E D I A T R I C S Relationship of intravenous sodium bicarbonate infusions and cerebral intraventricular hemorrhage The incidence of cerebral intraventricular hemorrhage was determined by computed tomography in 100 infants with birth weights less than or equal to 1,500 gm. A comparison of lVH with serum sodium concentrations and the amount of intravenous sodium bicarbonate administered did not reveal a significant relationship. Analysis of the methodof infusion of sodium bicarbonate indicated that the rapid infusion of hyperosmolar (M to M/2) sodium bicarbonate is associated with a significantly increased incidence of IVH. LuAnn Papile, M.D.,* Jerome Burstein, M.D., Rochelle Burstein, M.D., Herbert Koflter, M.D., and Beverly Koops, M.D., A Ibuquerque, N. M. CEREBRAL INTRAVENTRICULAR HEMORRHAGE is one of the major causes of death in preterm infants. Although the etiology of IVH is unknown, the infusion of hyperosmolar solutions of NaHCO~ (M to M/2) for the treatment of neonatal acidosis has been implicated as a contributing factor to the development of IVH.', 2 Simmons et al '~ proposed that sodium bicarbonate infu- sions resulting in hypernatremia (i.e., a serum sodium concentration exceeding 150 mEq/1 or in a daily intake of sodium in excess of 8 mEq/kg) placed an infant at an increased risk for IVH. The findings of Wigglesworth et aP support this proposal. Conversely, reports by Roberton and Howat) by Anderson et at, 6 and by Corbet et aF do not support the association between the use of alkali therapy and IVH. The conflicting results concerning the association of alkali therapy with IVH reflect the problem inherent in studies based on autopsy data. As Leviton et al ~ pointed out, autopsy data may be extrapolated to the population of the living only if complete data from the whole population are available. In the present study, we have From the Department of Pediatrics and Radiology, University of New Mexico School of Medicine. Supported in part by National Institutes of Health Biomedical Research Support Grant (5S01RR05583) and the James Picker Foundation. *Reprint address: Department of Pediatrics, University of New Mexico School of Medicine, Albuquerque, NM 87131. reviewed the serum sodium concentrations, the amount of sodium administered, and the method of infusion of sodium bicarbonate in 100 consecutive infants with a birth weight less than or equal to 1,500 gm. Each of the infants had a computed tomography brain scan performed to document the presence or absence of IVH. Abbreviations used IVH: intraventricular hemorrhage CT: computed tomography ICH: intracranial hemorrhage MATERIALS AND METHODS All infants with a birth weight less than 1,500 gm who were admitted to the Newborn Intensive Care Unit at the University of New Mexico Medical Center from July, 1976, to July, 1977, were included in the study. The study population consisted of 45 infants who were born at the Medical Center and 55 infants who were transported to the Medical Center for the management of prematurity and its complications. Pertinent clinical information, including daily serum sodium concentrations, the amount of sodium intake, and the method of administration of alkali therapy was tabulated by one of the authors (L.A.P.). During the study period, infants in our unit received dilute (M/4) sodium bicarbonate only when a persistent Vol. 93, No. 5, pp. 834-836 0022-3476/78/110834+ 03500.30/0 © 1978 The C. V. Mosby Co. Volume 93 Sodium bicarbonate infusions and cerebral hemorrhage 8 3 5 Number 5 metabolic acidosis which could not be corrected with plasma expanders (10 ml/kg of whole blood or 1 gm/kg of salt poor albumin) was present. Thirty-five of the 55 transport infants received some sodium bicarbonate prior to transport; the method of administration varied with the referring physician. The diagnosis of cerebral intraventricular hemorrhage was determined by a computed tomography brain scan in each of the infants. The protocol for the use of CT in the diagnosis of IVH has been reported previously? RESULTS A comparison of the infants with and without IVH is contained in Table I. The differences between the two groups of infants with respect to gestational age, birth weight, and one-minute and five-minute Apgar scores are not statistically significant. The amount of sodium bicarbonate administered to the study infants is outlined in Table II. The majority of the infants received alkali therapy only during the first 24 to 48 postnatal hours. No infant was given more than maintenance sodium replacement (3 to 4 mEq kg/day) after the second postnatal day. Seven of the 29 infants (two with IVH, five with no IVH) who received no alkali therapy died within three hours of birth. Analyses of the clinical courses of these seven infants indicate that each of the infants would have been a candidate for parenteral sodium bicarbonate therapy if they had survived. An intake of sodium in excess of 8 mEq/kg/day was not associated with an increased incidence of IVH (P > 0.20). Serum sodium concentrations are given in Table III. Twelve infants (seven with IVH and five with no IVH) died prior to 12 hours of age and did not have serum sodium determinations performed. Only one of these 12 infants received greater than 8 mEq-sodium bicarbon- ate/kg and six received no alkali therapy. Of the five infants with hypernatremia (serum sodium > 150 mEq/1) only one infant (IVH) was given greater than 8 mEq sodium bicarbonate/kg. Two infants (no IVH) received less than 8 mEq sodium bicarbonate/kg and two infants (no IVH) had no alkali infusions. Statistical analysis indicated that hypernatremia was not associated with IVH (P > 0.10). Because some of the infants who were transferred to our unit received hyperosmolar (M to M/2) sodium bicarbonate infusions prior to transport, an analysis of the relationship between IVH and the infusion of hyperosmo- lar alkali therapy at a rapid rate was done (Table IV). There was a significant relationship between the infusion of concentrated sodium bicarbonate solutions and IVH (P < 0.05). Table I. Clinical characteristics of infants with and with- out IVH IVH ] No IVH No. of infants 44 56 Birth Weight (gm) Mean 1,070 1,158 Range 590-1,500 624-1,500 Gestational age (wk) Mean 29.4 30.5 Range 26-35 26-37 Apgar score I min (mean) 3 4 5 rain (mean) 4 5 Assisted ventilation (<24 hr) 34 33 Table II. Sodium bicarbonate therapy and IVH >8mEq sodium bicarbonate/ kg/24 hr <8mEq sodium bicarbonate/ kg/24 hr IVH[NolVH ITotal 5 1 6 39 55 94 Table lIl. Serum sodium concentration and IVH Serum sodium > 150 mEq/1 1 4 5 Serum sodium < 150 mEq/l 36 47 83 *Twelve infants (7 with IVH and 5 with no IVH) who died prior to 12 hours of age did not have a serum sodium determination. Table IV. Method of sodium bicarbonate infusion and IVH Hypertonic sodium bicarbonate (M to M/2) No hypertonic sodium bicarbonate No. of infants 23 77 IVH 15 29 No IVH 8 48 DISCUSSION Intracranial hemorrhage has been produced experi- mentally with the rapid infusion of hypertonic sodium solutions.l 1-13 The mechanism for the production of I CH is attributed to the abrupt changes in serum osmolarity induced by such infusions, with rapid shifts of fluid from the intracellular to the extracellular compartment? ~ Although these studies define a relationship between the rapid sodium infusion and ICH, the type of ICH produced is not analogous to IVH found in preterm infants. 8 3 6 Papile et al. The Journal of Pediatrics November 1978 The clinical observation of a possible association between the vigorous parenteral use of sodium bicarbon- ate with or without resultant hypernatremia was made by Simmons et al? In a retrospective study they noted a higher incidence of ICH among nonsurviving infants when an infusion of concentrated sodium bicarbonate was administered rapidly for neonatal acidosis than when a slow infusion of dilute (M/3) alkali was used. Since Simmons et al considered the spectrum of ICH, it is difficult to delineate what relationship the use of alkali and hypernatremia may have to IVH in preterm infants. Several investigators have reviewed autopsy data in an attempt to define an association between the administra- tion of hypertonic alkali solutions and IVH in preterm infants? -~ Wigglesworth et aP concluded that there was a possible causal relationship, yet Roberton and Howat 5 and Anderson et aP found no association. Finbergy' in a recent editorial, has emphasized that the issue of IVH and alkali therapy relates more to the concentration and the rate of infusion of sodium bicar- bonate than to the total amount infused. Our findings lend support to Finberg's statement. Among our patients treated with conservative alkali therapy (M/4), the inci- dence of IVH in the non-surviving infants was lower than that in other reported series. '~, ,7 In the group that received a rapid infusion of hyper- tonic sodium bicarbonate (M to M/2), there was not only a greater incidence of IVH, but also a significant relation- ship between this method of alkali therapy and IVH. This group did not differ from the group treated conservatively in regarc o birth asphyxia or the need for assisted ventilation. Rather, the aggressive use of sodium bicar- bonate depended upon the practice of the physician caring for the infant. Since all of the infants in this group were treated with sodium bicarbonate at birth or shortly thereafter, it would not appear that alkali therapy was given after the onset of IVH. Our data support the theory that the use of hypertonic sodium bicarbonate as a rapid infusion in the neonate is associated with an increased incidence of IVH. In view of our findings, and those of othersy '" the use of sodium bicarbonate for the correction of a persistent metabolic acidosis in the neonate should be limited to a slow infusion of a dilute (M/4) solution. We thank Ms. Nancy Kanitchis for assistance in preparation of the manuscript. REFERENCES 1. Usher RH: Treatment of respiratory distress syndrome, in Winters RW, editor: The body fluids in pediatrics, Boston, 1973, Little, Brown & Company, pp 303-337. 2. Bland RD, Clarke TL, and Harden LB: Rapid infusion of sodium bicarbonate and albumin into high-risk premature infants soon after birth: A controlled prospective study, Am J Obstet Gynecol 124:263, 1976. 3. Simmons MA, Adcock EW, Bard H, and Battaglia FC: Hypernatremia and intracranial hemorrhage in neonates, N Engl J Med 291:6, 1974. 4. Wigglesworth JS, Keith IH, Girling DJ, and Slade SA: Hyaline membrane disease, alkali, and intraventricular hemorrhage, Arch Dis Child 51:755, 1976. 5. Roberton NRC, and Howat P: Hypernatraemia as a cause of intracranial hemorrhage, Arch Dis Child 50:938, 1975. 6. Anderson JM, Cockburn F, Bain AD, Forfar JO, Turner TL, Brown JK, and Maehin GA: Hyaline membrane disease, alkaline buffer treatment, and cerebral intraventric ular hemorrhage, Lancet 1:117, 1976. 7. Corbet A J, Adams JM, Kenny JD, Kennedy J, and Rudolph AJ: Controlled trial of bicarbonate therapy in high-risk premature newborn infants, J PEDIATR 91:771, 1977. 8. Leviton A, Epidem SM, and Gilles FH: Limitations of autopsy data in relating neonatal intracranial hemorrhage to buffer therapy, Hum Pathol 8:599, 1977. 9. Papile LA, Burstein J, Burstein R, and Kottler H: Incidence and evolution of subependymal and intraventricular hemorrhage: A study of infants with birthweights less than 1,500 grams, J PEDIATR 92:529, 1978. 10. Luttrell CN, Finberg L, and Drawdy LP: Hemorrhagic encephalopathy induced by hypernatremia. II. Experimen- tal observations on hyperosmolality in cats, Arch Neurol 1:153, 1959. 11. Finberg L, Luttrell C, and Redd H: Pathogenesis of lesions in the nervous system in hypernatremic cats II. Experimen- tal studies of gross anatomic changes and alterations of chemical composition of the tissues, Pediatrics 23:46, 1959. 12. Kravath RE, Aharon AS, Abal G, and Finberg L: Clinically significant physiologic changes from rapidly administered hypertonic solutions: Acute osmol poisoning, Pediatrics 46:267, 1970. 13. Turbeville DF, Bowen FW, and Killam AP: Intracranial hemorrhages in kittens: Hypernatremia versus hypoxia, J PEDIATR 89:294, 1976. 14. Finberg L: Dangers to infants caused by changes in osmolal concentration, Pediatrics 40:103 l, 1967. 15. Finberg L: The relationship of intravenous infusions and intracranial hemorrhage-a commentary, J PEDIATR 91:777, 1977. 16. Frederick J, and Butler NR: Certain causes of neonatal death. II. Intraventricular hemorrhage, Biol Neonate 15:257, 1970. 17. Pevsner PH, Garcia-Bunuel R, Leeds N, et al: Subependy- mal and intraventricular hemorrhage in neonates, Radiolo- gy 119:111, 1976. • 18. Steichen JJ, and Kleinman LI: Studies in acid-base balance. I. Effect of alkali therapy in newborn dogs with mechani- cally fixed ventilation, J PEDIATR 91:287, 1977. 19. Ostrea EM, and Odell GB: The influence of bicarbonate administration on blood pH in a "closed system": clinical implications, J PEDIATR 80:671, 1972.