Association of Hospital Resource Utilization With Neurodevelopmental Outcomes in Neonates With Hypoxic-Ischemic Encephalopathy

Abstract

Importance: Intercenter variation exists in the management of hypoxic-ischemic encephalopathy (HIE). It is unclear whether increased resource utilization translates into improved neurodevelopmental outcomes.

Objective: To determine if higher resource utilization during the first 4 days of age, quantified by hospital costs, is associated with survival without neurodevelopmental impairment (NDI) among infants with HIE.

Design, setting, and participants: Retrospective cohort analysis of neonates with HIE who underwent therapeutic hypothermia (TH) at US children's hospitals participating in the Children's Hospitals Neonatal Database between 2010 and 2016. Data were analyzed from December 2021 to December 2022.

Exposures: Infants who survived to 4 days of age and had neurodevelopmental outcomes assessed at greater than 11 months of age were divided into 2 groups: (1) death or NDI and (2) survived without NDI. Resource utilization was defined as costs of hospitalization including neonatal neurocritical care (NNCC). Data were linked with Pediatric Health Information Systems to quantify standardized costs by terciles.

Main outcomes and measures: The main outcome was death or NDI. Characteristics, outcomes, hospitalization, and NNCC costs were compared.

Results: Among the 381 patients who were included, median (IQR) gestational age was 39 (38-40) weeks; maternal race included 79 (20.7%) Black mothers, 237 (62.2%) White mothers, and 58 (15.2%) mothers with other race; 80 (21%) died, 64 (17%) survived with NDI (combined death or NDI group: 144 patients [38%]), and 237 (62%) survived without NDI. The combined death or NDI group had a higher rate of infants with Apgar score at 10 minutes less than or equal to 5 (65.3% [94 of 144] vs 39.7% [94 of 237]; P < .001) and a lower rate of infants with mild or moderate HIE (36.1% [52 of 144] vs 82.3% [195 of 237]; P < .001) compared with the survived without NDI group. Compared with low-cost centers, there was no association between high- or medium-hospitalization cost centers and death or NDI. High- and medium-EEG cost centers had lower odds of death or NDI compared with low-cost centers (high vs low: OR, 0.30 [95% CI, 0.16-0.57]; medium vs low: OR, 0.29 [95% CI, 0.13-0.62]). High- and medium-laboratory cost centers had higher odds of death or NDI compared with low-cost centers (high vs low: OR, 2.35 [95% CI, 1.19-4.66]; medium vs low: OR, 1.93 [95% CI, 1.07-3.47]). High-antiseizure medication cost centers had higher odds of death or NDI compared with low-cost centers (high vs. low: OR, 3.72 [95% CI, 1.51-9.18]; medium vs low: OR, 1.56 [95% CI, 0.71-3.42]).

Conclusions and relevance: Hospitalization costs during the first 4 days of age in neonates with HIE treated with TH were not associated with neurodevelopmental outcomes. Higher EEG costs were associated with lower odds of death or NDI yet higher laboratory and antiseizure medication costs were not. These findings serve as first steps toward identifying aspects of NNCC that are associated with outcomes.

Figure 1.. Study Population

CHND indicates Children’s Hospitals Neonatal Database; NDI, neurodevelopmental impairment; PHIS, Pediatric Health Information System.

Figure 2.. Adjusted Electroenchaphalography (EEG) and Antiseizure Medication (ASM) Costs by Center for Survivors Without Neurodevelopmental Impairment

Scatterplot where each bubble represents the volume of hypoxic-ischemic encephalopathy cases in each center and the bubble position represents the center median ASM and EEG costs (expressed on the logarithmic scale). The vertical axis indicates the 25th and 75th percentile of the median overall EEG center costs while the horizontal axis of the median overall ASM center costs. Centers with higher median EEG costs had lower median ASM costs.