Metabolic acidosis in preterm infants has been associated with the development of periventricular haemorrhage (Moriette 1977; Levene 1982; Cooke 1997), periventricular leucomalacia (Low 1990), and poor longer term neurodevelopmental outcomes in very low birth weight infants (Goldstein 1995; Aylward 1993). An inverse relationship between the degree of acidosis at birth and cognitive abilities assessed between four and seven years of age has been demonstrated in term and preterm infants (Stevens 1999). The development of intraventricular haemorrhage has been associated with fluctuations in cerebral arterial blood flow velocity (Perlman 1983). Cerebral vascular resistance is decreased in term infants with metabolic acidosis in the first week of life (Morrison 1995) and low arterial pH has been associated with increased cerebral artery blood flow velocity in very low birth weight infants (Weir 1999).
The definition of metabolic acidosis in preterm infants has not been clearly established. A normal arterial blood pH in the term infant has been defined as pH 7.27- 7.43 at up to 24 hours post birth and 7.32 - 7.42 at seven days of age (Koch 1968). In normal pregnancies fetal pH does not change with gestational age (Soothill 1986). Normal arterial base excess in the first 28 days of life has been defined as - 5 to + 5 mmol/litre in the preterm infant (Rennie 1999). The Joint Working Group of the British Association of Perinatal Medicine recommended maintaining an arterial pH above 7.25 as below this pH various physiological and cellular functions are compromised (BAPM 1992).
A postal survey of British neonatologists found that most used fluid boluses (4.5% human albumin, fresh frozen plasma, blood, normal saline) or base (sodium bicarbonate and Tris-(Hydroxymethyl)Amino Methane (THAM)) infusions to correct metabolic acidosis in preterm infants (Simpson 1994). However, the use of any type of additional fluid in the preterm infant must be considered carefully. Systematic review of the use of albumin in resuscitation of critically ill patients of all age groups demonstrated statistically significantly increased mortality rates for those receiving albumin compared to controls (Alderson 2001). A systematic review of restricted versus liberal water intake in preterm infants showed restricted water intake reduced the risk of patent ductus arteriosus, necrotizing enterocolitis and death and a trend toward reduced risk of bronchopulmonary dysplasia (Bell 2001).
Both sodium bicarbonate (4.2% or 2.1%) and THAM are hyperosmolar. Sodium bicarbonate may cause hypernatraemia and has been associated with intra-ventricular haemorrhage when given rapidly and in large quantities (Papile 1978; Simmons 1974). There are also concerns that intravenous infusion of base can cause a transient paradoxical worsening of intra-cellular acidosis, loss of cerebral vascular autoregulation, decreased cerebral blood flow and acute changes in cerebro-spinal fluid pH (Lou 1978; Lou 1979). Observational data have suggested that treatment with sodium bicarbonate on the first day of life is associated with an higher incidence of intra-ventricular haemorrhage in very preterm infants (Synnes 2001).
Pre-specified subgroup analyses:
1. Trials where all participating infants were treated in the first seven
days of post-natal life compared with infants treated at a later stage.
2. Trials where all participating infants were born at less than 32
weeks' gestational age compared with more mature preterm infants.
3. Trials where all participating infants were born at less than 28
weeks' gestational age compared with more mature preterm infants.
4. Trials where all participating infants had severe acidaemia (pH less than 7.15) compared with less severe acidaemia.
5. Trials where all participating infants had clinical indicators of
poor perfusion (low blood pressure, poor cutaneous perfusion, receiving inotropic
support or volume support for hypotension) compared with trial where infants
did not have these indicators.
Secondary:
1. Failure to improve arterial pH to more than 7.25 or base excess to better than -6 mmol/litre within four hours of treatment.
CJL and WM used the criteria and standard methods of the Cochrane Neonatal Review Group to assess independently the methodological quality of the included trials in terms of allocation concealment, blinding of parents or carers and assessors to intervention, and completeness of assessment in all randomised individuals. Where necessary, we requested additional information from trial authors to clarify methodology and results. WM and CJL extracted relevant information and data from each included study. Each reviewer extracted the data separately, compared data, and resolved differences by discussion until consensus was achieved. We contacted the trial investigators for further information as required.
We presented outcomes for categorical data as relative risk and risk difference with respective 95% confidence intervals. For continuous data, we planned to use the weighted mean difference with 95% confidence interval. We planned to estimate the treatment effects of individual trials and examine heterogeneity between trial results by inspecting the forest plots and quantifying the impact of heterogeneity in any meta-analysis using a measure of the degree of inconsistency in the studies' results (I- squared statistic). If we detected statistical heterogeneity, we planned to explore the possible causes (for example, differences in study quality, participants, intervention regimens, or outcome assessments) using post hoc sub group analyses. We planned to use a fixed effects model for meta-analyses.
We included two trials (Corbet 1977; Dixon 1999). Details are presented in the table: Characteristics of included studies.
1. Corbet 1977 randomly allocated newborn preterm infants within the first few hours after birth to receive either "liberal" sodium bicarbonate infusions (N = 30) or "conservative" bicarbonate infusions (N = 32) given with their standard maintenance intravascular infusions of 10% glucose. Those infants in the "liberal" group were allocated to receive sodium bicarbonate at doses of 5 to 15 mmol /decilitre of 10% glucose, titrated on the degree of acidosis. The bicarbonate infusion was continued until the infant's arterial blood pH reached 7.30. These infants received an average of 3.6 mmol/kilogram of sodium bicarbonate in first 24 hours of post-natal life. The infants in the "conservative" group did not have any sodium bicarbonate added to their maintenance glucose infusions.
This trial did not strictly fulfil our a priori inclusion criteria since biochemically-confirmed acidosis was not per se an eligibility criterion for participation in the trial. The investigators recruited infants thought to be at "high-risk" of acidosis (birth weight less than 1500 grams, or birth weight less than 2000 grams and infant needing mechanical ventilation, or Apgar score less than four at one minute, or infant with a clinical diagnosis of hyaline membrane disease). However, we made a consensus decision to include the trial because there was post hoc evidence that most participating infants fulfilled our review criterion for acidosis: All of the participating infants were acidotic (pH less than 7.30) at enrolment and most had arterial blood pH values less than 7.25. The mean arterial blood pH of participating infants was 7.16 at trial entry. Participating infants were followed up until the end of the first week of post-natal life. The primary outcomes assessed were time to correction of acidosis, mortality in the first week of post-natal life, and incidence of intra-ventricular haemorrhage.
2. Dixon 1999 randomly allocated 36 normotensive newborn infants with metabolic acidosis (defined as a simultaneous arterial blood pH less than 7.25 and base excess worse than -6 mmol/litre) to receive an intravenous infusion over 30 minutes of either 10 millilitres per kilogram of 4.5% albumin infusion (N = 20) or 4.2% sodium bicarbonate (N = 16): dose in millimoles calculated as one-sixth of product of weight (kilograms) and base excess (mmol/litre). Most of the participating infants were preterm. Thirty-three of the 36 infants were born before 36 weeks' gestation. The primary outcome assessed was arterial blood pH and base excess two hours after the intervention. Other clinical outcomes were not reported.
Two trials were excluded (Bland 1976; Sinclair 1968). Details are presented in the table: Characteristics of excluded studies.
1. Sinclair 1968 randomly allocated 20 low birth weight infants with evidence of hypoxia and/or acidaemia to receive either a "slow" or a "rapid" infusion of sodium bicarbonate. This trial was not eligible for inclusion in this review as there was not a comparison group who did not receive any base or who received fluid bolus.
2. Bland 1976 randomly allocated 51 hypoproteinaemic preterm infants "at risk of developing acidaemia" to receive either an infusion of sodium bicarbonate over 5 to 10 minutes or an infusion of glucose or albumin within the first two hours after birth. This study was excluded because most of the participating infants were not acidotic at trial entry.
We are aware of one on-going trial that may potentially be eligible for inclusion in a future update of this review (Lawn 2005). This randomised controlled trial compares the effect of continuous intra-arterial infusion of a weak solution of either sodium bicarbonate or sodium chloride. The primary outcome is the incidence of acidosis treated with bolus infusions of sodium bicarbonate or crystalloid solution. The trial aims to recruit 160 infants. 20 infants recruited to date.
1. Infusion of base versus no treatment
Mortality (01:01): Corbet 1977 did
not find evidence of an effect on mortality: Relative risk 1.39 (95% confidence
interval 0.72 to 2.67), risk difference 0.12 (95% confidence interval -0.12
to 0.36).
Intraventricular haemorrhage (01:02): Corbet 1977 did not find a statistically significant difference in the incidence of intra/peri-ventricular haemorrhage (only post-mortem diagnoses were available): Relative risk 1.24 (95% confidence interval 0.47 to 3.28), risk difference 0.05 (95% confidence interval -0.16 to 0.25).
This trial did not assess longer term neurodevelopmental outcomes.
Failure to improve pH to more than 7.25 and base excess to less than -6 mmol/litre within 4 hours of treatment: Corbet 1977 did not find any statistically significant difference in the rate at which pH was corrected, or in the mean arterial blood pH levels two hours after commencing the intervention. These data were presented graphically and could not be extracted for calculation of mean differences.
2. Infusion of fluid bolus versus no treatment
No trials identified.
3. Infusion of base versus fluid bolus
Dixon 1999: There were not any data on mortality, incidence of peri-ventricular haemorrhage, or longer term neurodevelopment.
Failure to improve pH to more than 7.25 and base excess to less than -6 mmol/litre within 4 hours of treatment (02:01): Dixon 1999 reported that two of 16 infants given base versus nine of 20 infants given a fluid bolus had persistent acidosis (arterial blood pH less than 7.25 and base excess worse than -6 mmol/litre) at two hours post-intervention. This difference was of borderline statistical significance: Relative risk 0.28 (95% confidence interval 0.07 to 1.11), risk difference -0.33 (95% confidence interval -0.60 to -0.05).
Sub-group analyses
1. All infants treated in the first 7 days of life: In both included
trials, all of the participating infants were enrolled on the first day of
post-natal life.
2. Infants less than 32 weeks' gestational age: Sub-group data were not available from either trial.
3. Infants less than 28 weeks' gestational age: Sub-group data were not available from either trial.
4. Infants with severe acidaemia (pH less than 7.15): Sub-group data were not available from either trial.
5. Infants who had clinical indicators of poor perfusion (low blood pressure,
poor cutaneous perfusion, receiving inotropic support or volume support for
hypotension): Sub-group data were not available from Corbet 1977. In Dixon 1999, infants with hypotension were not eligible to participate.
The rationale for the use of an infusion of base is that correction of metabolic acidosis will prevent the adverse outcomes that are associated with acidosis in preterm infants. In fact, the trial that compared treatment with sodium bicarbonate infusion versus no treatment did not find a statistically significant difference in the rate of resolution of metabolic acidosis (Corbet 1977). However, this may be due to the very slow rate of infusion of sodium bicarbonate in the intervention group. Infants received 3.6 millimoles per kilogram over the first 24 hours of post-natal life. The trial did not find a statistically significant effect on early neonatal mortality, but included only 62 infants and was underpowered to detect an small but clinically important effect. Given the lack of robust evidence of benefit (and theoretical potential for harm), it has been argued that treating preterm infants with metabolic acidosis with infusions of base should be regarded as experimental and should not be undertaken out with the context of a randomised controlled trial (Ammari 2002).
Similar concerns about the lack of evidence of benefit and potential for harm have been raised about the use of infusion of a fluid bolus in treating preterm infants with metabolic acidosis or other evidence of systemic underperfusion (Hope 1998). There is some biological plausibility that a fluid bolus may improve organ perfusion in infants with hypovolaemia. However, in the absence of a clear precipitating event such as haemorrhage, hypovolaemia is probably a rare cause of metabolic acidosis in preterm infants. Because systemic underperfusion and hypovolaemia are very difficult to assess clinically many infants may receive treatment with a fluid bolus inappropriately. A Cochrane review found insufficient evidence that routine early volume expansion with blood products or crystalloid solutions prevents morbidity and mortality in very preterm infants (Osborn 2004). In this review, we did not find any randomised controlled trials that compared the effect of treatment with fluid bolus versus no treatment in preterm infants with metabolic acidosis. Dixon 1999 compared base infusion with fluid bolus and found some evidence (of borderline clinical significance) that base infusion corrected metabolic acidosis more effectively than fluid bolus. This trial specifically excluded infants with evidence of hypotension from participating.
| Study | Methods | Participants | Interventions | Outcomes | Notes | Allocation concealment |
| Corbet 1977 | Blinding of randomisation: can't tell Blinding of intervention: can't tell Complete follow-up: yes Blinding of outcome measurement: can't tell | Newborn preterm infants within the first few hours after birth: birth weight less than 1500 grams, or birth weight less than 2000 grams and infant needing mechanical ventilation, or Apgar score less than four at one minute, or infant with a clinical diagnosis of hyaline membrane disease. | 1. "Liberal" sodium bicarbonate infusions (N= 30) versus 2. "conservative" sodium bicarbonate infusions (N= 32) given with the standard maintenance intravascular infusions of 10% glucose. | Time to correction of acidosis, mortality in the first week of post-natal life, and incidence of intra-ventricular haemorrhage. | Those
infants in the "liberal" group were allocated to receive sodium bicarbonate
at doses of 5 to 15 mmol /decilitre of 10% glucose, titrated on the degree
of acidosis. The bicarbonate infusion was continued until the infant's arterial
blood pH reached 7.30. The infants in the "conservative group did not have
any sodium bicarbonate added to their maintenance glucose infusions. | B |
| Dixon 1999 | Blinding of randomisation: yes Blinding of intervention: no Complete follow-up: yes Blinding of outcome measurement: can't tell | Neonates (mostly preterm) with metabolic acidosis (arterial blood pH <7.25 and base excess of more than -6 mmol/litre). Exclusion criterion: infants with mean blood pressure less than third percentile for birth weight (Versmold 1981). | Infusion over 30 minutes of : 1. 10 ml/kg 4.5% albumin (N= 20) versus 2. 4.2% sodium bicarbonate (N=16): dose in mmol calculated as one-sixth of product of weight (kg) and base excess (mmol/L). | Arterial blood pH and base excess measured two hours post-intervention. | A |
| Study | Reason for exclusion |
| Bland 1976 | Bland
1976 compared giving rapid infusions of sodium bicarbonate versus albumin
or dextrose-water infusions in preterm infants at risk of acidaemia within
the first two hours of post-natal life. However, most of the participating
infants were not acidotic at trial entry. |
| Sinclair 1968 | Sinclair 1968 randomised 20 infants with a birth weight of 1000-2500 g to receive one of four different treatment combinations which included a trial of rapid versus slow infusion of sodium bicarbonate, but without a comparison group who did not receive any base (or who received fluid bolus). |
| Study | Trial name or title | Participants | Interventions | Outcomes | Starting date | Contact information | Notes |
| Lawn 2005 | Effect of continuous intra-arterial weak bicarbonate infusion on the use of fluid and bicarbonate boluses to correct metabolic acidosis in babies weighing less than 1000 g and who are less than 32 weeks gestation at birth | http://controlled-trials.com |
Corbet AJ, Adams JM, Kenny JD, Kennedy J, Rudolph AJ. Controlled trial of bicarbonate therapy in high-risk premature newborn infants. Journal of Pediatrics 1977;91:771-6.
Dixon 1999 {published data only}
Dixon H , Hawkins K, Stephenson T. Comparison of albumin versus bicarbonate treatment for neonatal metabolic acidosis. European Journal of Pediatrics 1999;158:414-5.
Bland RD, Clarke TL, Harden LB. Rapid infusion of sodium bicarbonate and albumin into high-risk premature infants soon after birth: A controlled, prospective trial. American Journal of Obstetrics and Gynecology 1976;124:263-7.
Sinclair 1968 {published data only}
Sinclair JC, Engel K, Silverman WA. Early correction of hypoxemia and acidemia in infants of low birth weight: a controlled trial of oxygen breathing, rapid alkali infusion, and assisted ventilation. Pediatrics 1968;42:565-89.
Lawn C, Weir F. Effect of continuous intra-arterial weak bicarbonate infusion on the use of fluid and bicarbonate boluses to correct metabolic acidosis in babies weighing less than 1000 g and who are less than 32 weeks gestation at birth. Personal communication: see http://controlled-trials.com.
* indicates the primary reference for the study
Alderson P, Bunn F, Lefebvre C, Li Wan Po A, Li L, Roberts I , Schierhout G. The Albumin Reviewers. Human albumin solution for resuscitation and volume expansion in critically ill patients (Cochrane Review). In: The Cochrane Database of Systematic Reviews, Issue 1, 2001.
Ammari AN, Schulze KF. Uses and abuses of sodium bicarbonate in the neonatal intensive care unit. Current Opinion in Pediatrics 2002;14:151-6.
Askie LM, Henderson-Smart DJ. Restricted versus liberal oxygen exposure for preventing morbidity and mortality in preterm or low birth weight infants (Cochrane review). In: The Cochrane Database of Systematic Reviews, Issue 4, 2001.
Askie LM, Henderson-Smart DJ, Irwig L, Simpson JM. Oxygen-saturation targets and outcomes in extremely preterm infants. New England Journal of Medicine 2003;349:959-67.
Aylward GP. Perinatal asphyxia: effects of biologic and environmental risks. Clinics in Perinatology 1993;20:433-49.
Development of audit measures and guidelines for good practice in the management of neonatal respiratory distress sydrome. Report of a Joint Working Group of the British Association of Perinatal Medicine and the Research Unit of the Royal College of Physicians. Archives of Disease in Childhood 1992;67:1221-7.
Bell EF, Acarregui MJ. Restricted versus liberal water intake for preventing morbidity and mortality in preterm infants. In: The Cochrane Database of Systematic Reviews, Issue 1, 2001.
Cooke RW. Factors associated with periventricular haemorrhage in very low birthweight infants. Archives of Disease in Childhood 1981;56:425-31.
Goldstein RF, Thompson RJ Jr, Oehler JM, Brazy JE. Influence of acidosis, hypoxemia and hypotension on neurodevelopmental outcome in very low birthweight infants. Pediatrics 1995;95:238-43.
Hope P. Pump up the volume? The routine early use of colloid in very preterm infants. Archives of Disease in Childhood 1998;78:F163-5.
Koch G, Wendel H. Adjustment of arterial blood gases and acid base balance in the normal newborn infant during the first week of life. Biology of the Neonate 1968;12:136-61.
Levene MI, Fawer CL, Lamont RF. Risk factors in the development of intraventricular haemorrhage in the preterm neonate. Archives of Disease in Childhood 1982;57:410-7.
Lou HC, Lassen NA, Friis-Hansen B. Decreased cerebral blood flow after administration of sodium bicarbonate in the distressed newborn infant. Acta Neurologica Scandinavica 1978;57:239-47.
Lou HC, Lassen NA, Friis-Hansen B. Impaired autoregulation of cerebral blood flow in the distressed newborn infant. J Pediatr 1979;94:118-21.
Low JA, Froese AF, Galbraith RS, Sauerbrei EE, McKinven JP, Karchmar EJ. The association of fetal and newborn metabolic acidosis with severe periventricular leukomalacia in the pre-term newborn. Am J Obstet Gynecol 1990;162:977-81.
Moriette G, Relier JP, Larroche JC. Intraventricular haemorrhages in hyaline membrane disease. Archives Francaises de Pediatrie 1977;34:492-504.
Morrison FK, Patel NB, Howie GJ, Mires GJ, Herd RM. Neonatal cerebral arterial flow velocity waveforms in term infants with and without metabolic acidosis at delivery. Early Human Development 1995;42:155-68.
Osborn DA, Evans N. Early volume expansion for prevention of morbidity and mortality in very preterm infants. In: The Cochrane Database of Systematic Reviews, Issue 2, 2004.
Papile LA, Burstein J , Burstein R, Koffler H, Koops B. Relationship of intravenous sodium bicarbonate infusions and cerebral intraventricular hemorrhage. Journal of Pediatrics 1978;93:834-6.
Perlman JM, McMenamin JB, Volpe JJ. Fluctuating cerebral blood-flow velocity in respiratory distress syndrome. Relation to the development of intraventricular hemorrhage. New England Journal of Medicine 1983;309:204-9.
Rennie JM, Roberton NRC. In: Textbook of Neonatology. 3rd edition. Churchill Livingstone, 1999:Appendix 6, 1408.
Simmons MA, Adcock EW 3rd, Bard H, Battaglia FC. Hypernatremia and intracranial hemorrhage in neonates. New England Journal of Medicine 1974;291:6-10.
Simpson JM, Hawkins K, Gull N, Stephenson TJ. Treatment of metabolic acidosis in the newborn. An observational and questionnaire study. British Journal of Intensive Care 1994;4:80-6.
Soothill PW, Nicolaides KH, Rodeck CH, Gamsu H. Blood gases and acid-base status of the human second-trimester fetus. Obstetrics and Gynecology 1986;68:173-6.
Stevens CP, Raz S, Sander CJ. Peripartum hypoxic risk and cognitive outcome: a study of term and preterm birth children at early school age. Neuropsychology 1999;13:598-608.
Synnes AR, Chien LY, Peliowski A, Baboolal R, Lee SK; Canadian NICU Network. Variations in intraventricular hemorrhage incidence rates among Canadian neonatal intensive care units. Journal of Pediatrics 2001;138:525-31.
Tin W, Milligan DW, Pennefather P, Hey E. Pulse oximetry, severe retinopathy, and outcome at one year in babies of less than 28 weeks gestation. Archives of Disease of Childhood. Fetal and neonatal edition 2001;84:F106-10.
Versmold HT, Kitterman JA, Phibbs RH, Gregory GA, Tooley WH. Aortic blood pressure during the first 12 hours of life in infants with birth weight 610 to 4,220 grams. Pediatrics 1981;67:607-13.
Weir FJ, Ohlsson A, Myhr TL, Fong K, Ryan ML. A patent ductus arteriosus is associated with reduced middle cerebral artery blood flow velocity. European Journal of Pediatrics 1999;158:484-7.
| Comparison or outcome | Studies | Participants | Statistical method | Effect size |
|---|---|---|---|---|
| 01 Base (sodium bicarbonate or THAM) versus no treatment | ||||
| 01 Mortality at 7 days | 1 | 62 | RR (fixed), 95% CI | 1.39 [0.72, 2.67] |
| 02 Intra (and/or peri) ventricular haemorrhage (post-mortem diagnoses only) | 1 | 62 | RR (fixed), 95% CI | 1.24 [0.47, 3.28] |
| 02 Base (sodium bicarbonate or THAM) versus fluid bolus | ||||
| 01 Failure to improve pH to more than 7.25 or base excess above -6 mmol/litre at 2 hours post treatment. | 1 | 36 | RR (fixed), 95% CI | 0.28 [0.07, 1.11] |
Dr Fiona J Weir
Consultant
Neonatal Medicine
Royal Sussex County Hospital
E-mail: fiona.weir@brighton-healthcare.nhs.uk
| The review is published as a Cochrane review in The
Cochrane Library, Issue 2, 2005 (see http:www.thecochranelibrary.com for
information). Cochrane reviews are regularly updated as new evidence emerges
and in response to comments and criticisms, and The Cochrane Library should
be consulted for the most recent version of the Review. |