Does the Administration of Free Water in Labor Increase Morbidity in Premature Newborns?

Alejandro Dinerstein, Marcelo Rojas, Ricardo Nieto, Claudio Solana, Soledad Scacchi, Gaston Pérez, Lydia Otheguy, Miguel Larguia.

Maternidad Sarda, Argentina.

BACKGROUND

Fluids administerd to women in labor may produce hyponatremia (HypoNa) and increase the risk of neonatal morbidity.

OBJECTIVE

To evaluate the effect of the administration of free water in labor (FWL) on the morbidity of very low birth weight infants (VLBW).

DESIGN/METHODS

Prospective, observational, cohort. All women in labor with gestational age (GA) < 32 weeks and newborn birth weight <1500g were included. Major malformations, congenital infections, asphyxia, and kidney failure were exclusion criteria. Two explicative models of multivariate logistical regression were constructed. Interest variable: FWL (0: < 94 ml/hr, 1: 94 ml/hr). Resulting variable of the first model was HypoNa (< 130 mEq/l) in the first 24 hr of life and at least one of the following morbidities in the second one: RDS, PDA, and BPD. The first model was checked with the covariates GA (0 : 30, 1: < 30 weeks) and cesarian section, and the second with the covariates GA, HypoNa, amount of fluids (0: <] 100, 1 100 ml./kg/day), fluids between the 3rd and the 7th day (0: < 130, 1: 130 ml/kg/day), administration of Na in the first 48 hs (0: 0.60, 1: > 0. 60 mEq/kg/day), administration of Na between the 3rd and 7th day. Statistical software: STATA v7.

RESULTS

From 12/2002 to 11/2004, 118 mother-child binomials were included. The univariate analysis showed that the risk of HypoNa in VLBW was greater when FWL was 94 ml/hr (OR: 2.45, 95% CI: 1.1-5.45). This relation continues to be significant adjusted for the variables of the model (OR: 2.97, 95% CI:1.23-7.13). Both covariates were confusing, only GA had a modifying effect and its interaction with FWL was not significant The model shows a good Hosmer-Lemeshow adjustment of chi2(6) = 3.52, p = 0.74; area under the ROC curve: 0.66. The second model showed no relation between FWL and neonatal morbidity, though greater morbidity was observed when there was an increase in the unintentional administration of Na (p = 0.02) and the administration of fluids was greater than 100m/kg/day (p = 0.002) in the first 48 hours.

CONCLUSIONS

Infusion of free water in labor of more than 94 ml/hr was associated with an increased risk of HypoNa in VLBW, but no association with neonatal morbidity was observed. Greater attention should be given to the unintentional administration of Na in the first 48 hours of life.