Hypernatremic Dehydration in Breast Fed Infants: Lessons from a Baby-Friendly Hospital

The impact of the baby friendly hospital initiative on healthcare utilization among newborns insured by Medicaid in Delaware

BACKGROUND

The Baby Friendly Hospital Initiative was created to enhance breastfeeding, although its impact on infant healthcare utilization is unclear. Breast feeding infants are vulnerable to readmission soon after birth secondary to dehydration and hyperbilirubinemia. Breastfeeding can also protect infants from unnecessary health care utilization later in life by preventing infection. The objective of this study was to examine the impact of the Baby Friendly Hospital Initiative on readmissions and emergency department utilization among Medicaid births in Delaware.

METHODS

The study was a quasi-experimental design. Medicaid claims files were used to study births at five hospitals in Delaware born between January 1, 2014, and December 31, 2018, and covered under Medicaid at time of birth. Three hospitals were designated Baby Friendly, two were not and served as controls. Outcomes included Emergency Department (ED) utilization and readmissions within 30 days and one-year of birth hospitalization. Exposure to the Baby Friendly Hospital Initiative was determined by year and hospital of birth. Logistic regression and interrupted time series segmented regression analysis with controls were used to assess the effect of Baby Friendly Hospital Initiative on healthcare utilization.

RESULTS

In total, 19,695 infants were born at five hospitals with 80% (15,939) born at hospitals that were designated Baby Friendly. ED utilization and readmissions over the 1st year of life for breastfeeding related diagnosis at the Baby Friendly hospitals occurred in 240 (1.5%) and 226 (1.4%) of infants, respectively. Exposure to the Baby Friendly Hospital Initiative was associated with increased odds of all cause 30-day readmission (AOR: 1.15; 95% CI: 1.03-1.28) but not readmissions over the 1st year of life. While 30-day ED visits did not change after BFHI, one-year ED visits were reduced (0.91, 95% CI 0.86-0.97). A significant negative trend was seen over time for ED utilization post BFHI compared to controls (B: -5.90, p < 0.01).

CONCLUSION

There was a small observed increase in the odds of all cause 30-day readmissions with no change in one-year readmissions after BFHI in Delaware. Although there were no observed changes in 30-day ED utilization, there was a reduction in one-year ED utilization following the implementation of the Baby Friendly Hospital Initiative in Delaware birth hospitals. Our data help to inform policy and decision making for statewide systems of care that may be used to support breast feeding.

Increasing Sodium Variability in the First 96 Hours after Birth is Associated with Adverse In-Hospital Outcomes of Preterm Newborns

Background

Neonatal intraventricular hemorrhage prevention bundles for preterm infants commonly defer daily weighing for the first 72 h, with reweighing occurring on day 4. Clinicians rely on maintaining stable sodium values as a proxy of fluid status to inform fluid management decisions over the first 96 h after birth. Yet, there exists a paucity of research evaluating whether serum sodium or osmolality are appropriate proxies for weight loss and whether increasing variability in sodium or osmolality during this early transitional period is associated with adverse in-hospital outcomes.

Objectives

To evaluate whether serum sodium or osmolality change in the first 96 h after birth was associated with percent weight change from birth weight, and to assess potential associations between serum sodium and osmolality variability with in-hospital outcomes.

Methods

This retrospective, cross-sectional study included neonates born at ≤30 gestational weeks or ≤1250 g. We evaluated associations between serum sodium coefficient of variation (CoV), osmolality CoV, and maximal weight loss percentage in the first 96 h after birth with in-hospital neonatal outcomes.

Results

Among 205 infants, serum sodium and osmolality were poorly correlated with percent weight change in individual 24-h increments (R2 = 0.01-0.14). For every 1% increase in sodium CoV, there was an associated 2-fold increased odds of surgical necrotizing enterocolitis and 2-fold increased odds of in-hospital mortality (odds ratio, 2.07; 95% CI: 1.02, 4.54; odds ratio, 1.95; 95% CI: 1.10, 3.64, respectively). Sodium CoV was more strongly associated with outcomes than absolute sodium maximal change.

Conclusions

In the first 96 h, serum sodium and osmolality are poor proxies for assessing percent weight change. Increasing variability of serum sodium is associated with later development of surgical necrotizing enterocolitis and all-cause in-hospital mortality. Prospective research is needed to evaluate whether reducing sodium variability in the first 96 h after birth, as assessed by CoV, improves newborn health outcomes.

Neuroimaging Spectrum of Severe Hypernatremia in Infants with Neurological Manifestations

BACKGROUND

Variable neurological manifestations and imaging findings have been described in children with severe hypernatremia. We aimed to describe the spectrum of neuroimaging changes in infants with severe hypernatremia.

METHODS

This retrospective study included infants with severe hypernatremia (serum sodium >160 mEq/L), abnormal neurological examination, and an abnormal magnetic resonance imaging (MRI) of the brain over a period of 2 years in a tertiary care hospital. Relevant clinical data, including the feeding practices, clinical features, complications, and biochemical and radiological parameters, were entered in a structured pro forma. MRI findings were classified as vascular (hemorrhages and cerebral sinus venous thrombosis), osmotic demyelination syndrome (pontine and extrapontine myelinolyses), and white matter changes.

RESULTS

The common clinical features in the neonates were poor feeding (n = 4) and decreased urine output (n = 4); the older infants presented with gastrointestinal losses (n = 5). All cases had dehydration with encephalopathy. The patterns of radiological injury were vascular (hemorrhages, n = 5 and venous thrombosis, n = 3), osmotic demyelination (n = 8), and white matter changes (n = 7). Coagulopathy was correlated with the vascular complications (r = 0.8, p < 0.0001); the degree of dehydration was correlated with the venous thrombosis (r = 0.7, p < 0.04) and acute kidney injury (r = 0.8, p < 0.001). Neurological sequelae were seen in four cases and correlated with hypernatremia (r = 0.6, p = 0.03) and hyperosmolarity (r = 0.6, p = 0.03).

CONCLUSION

Characteristic neuroimaging findings are vascular changes in the form of venous thrombosis and hemorrhages, osmotic demyelination and white matter tract injury, and/or mostly combinations of these findings. Severe hypernatremia and resulting hyperosmolarity frequently cause neurological sequelae in neonates and infants.