NEONATAL SOCIETY ABSTRACTS
Chloride and metabolic acidosis in preterm infants
Presented at the Neonatal Society 2006 Summer Meeting (programme).
Hasan F, Green J, Wardle SP
Neonatal Intensive Care Unit, Nottingham University Hospitals, Nottingham, UK
Background: Metabolic acidosis is frequent in preterm infants. It may be related to a high lactate concentration reflecting poor tissue oxygenation however it may also occur with a normal lactate concentration. High concentrations of chloride administered to preterm infants have been associated with hyperchloraemia and metabolic acidosis (1) and further studies demonstrated a reduction in hyperchloraemia and acidosis when acetate was used in total parenteral nutrition (TPN) in place of chloride (2). Despite this, metabolic acidosis remains a significant problem that may result in unnecessary interventions and longer periods of ventilation, with possible harmful effects.
Aim: Observational study to quantify the incidence and severity of hyperchloraemia and metabolic acidosis in preterm infants, in the first week after birth when TPN containing acetate was used. We aimed to classify acidosis according to cause, using both conventional approaches to acid-base and Stewart’s strong ion theory (3) and to determine the relationship between chloride input and metabolic acidosis.
Methods: Blood gases, electrolyte concentrations and chloride (Cl) intake were collected prospectively in infants ≤28 weeks gestation, <1000g, who were ventilated during the first week after birth. Data were recorded anonymously and only routinely collected data were recorded. Plasma Cl concentrations were corrected for the amount of free water (corrected Cl concentration = Cl concentration x sodium concentration / 140). For each of the study variables median values were obtained for each baby for each study day and a mean value was calculated for each baby over the study period.
Results: 25 infants were studied. The median pH was ≤7.25 in 15 (60%) infants on 31 (23%) of the patient days studied. On 9 (29%) of those days (affecting 6 babies) there was a high PCO2; on 15 (48%) of the days there was a high lactate concentration (affecting 12 babies), and on 9 (29%) days (affecting 5 babies) an increased Cl concentration. On 7 days (22%) acidosis was associated with a low base excess (BE), but without hyperchloraemia or hyperlactataemia.
The median Cl intake was 4.03 (range 1.82 to 15.35) mmols/kg/day. 12 infants (48%) had a Cl intake of ≤4 mmol/kg/day however 5 infants (20%) received a median Cl input >6 mmols / kg / day during their first week.
There was no correlation between mean Cl intake and mean plasma corrected Cl concentration (r = 0.18, p = 0.45), and neither mean Cl intake (r = -0.35, p = 0.09) nor mean corrected Cl concentration (r = -0.3, p = 0.19) correlated with the pH. The mean corrected Cl concentration inversely correlated with the mean BE (r = -0.48, p = 0.03). On multivariate analysis the factors which were independently associated with the BE were corrected Cl concentration and lactate concentration (r2 = 0.579, p < 0.001).
Conclusions: Metabolic acidosis is a frequent problem despite using acetate in TPN. Chloride intakes are lower than in previous reports but remain high in some infants. The relative concentrations of strong ions (chloride and sodium) in the fluids given may be more important than absolute Cl concentrations. It may be possible to reduce the incidence of acidosis by using fluids that have a higher strong ion difference.
1. Richards CE, Drayton M, Jenkins H, Peters TJ. Acta Paediatr 1993;82:678-682
2. Peters O, Ryan S, Matthew L, Cheng K, Lunn J. Archives of Disease in Childhood 1997;7: F12-F15
3. Stewart PA. Canadian Journal of Physiology & Pharmacology 1983; 61(12):1444-61