Abstract

Objectives: We sought to evaluate hospital resource usage patterns and determine risk factors for neonatal withdrawal syndrome (NWS) in the United States.

Methods: Using the 2016 Kids’ Inpatient Database (KID), we conducted a retrospective cross-sectional analysis of a nationally representative sample of neonates with NWS. The KID is the largest publicly available pediatric (20 years of age and younger) inpatient care database in the United States. We analyzed a stratified probability sampling of 3.1 million pediatric hospital discharges weighted to 6.3 million national discharges. Descriptive statistics for hospital and patient characteristics were identified and binary variables were analyzed using the Student t test. Multivariate regression was performed to assess the predictors of NWS. We excluded discharges if total cost or hospital length of stay (LOS) exceeded mean values by >3 standard deviations. Hospitalizations with NWS diagnosis were identified using the International Classification of Diseases, 10th Revision, Clinical Modification code P96.1 in any 1 of 30 discharge diagnostic fields.

Results: We estimated that 25,394 pediatric discharges were associated with an NWS diagnosis, totaling 403,127 inpatient days at a cost of $1.8 billion. Compared with non-NWS newborns, neonates with NWS had higher mean hospital charges ($71,540 vs $15,765), longer mean hospital stays (16 days vs 3 days), and a significantly higher proportion of low birth weight (7.2% vs 1.9%), feeding problems (19.0% vs 3.5%), respiratory diagnoses (5.6% vs 2.5%), and seizure (0.3% vs 0.1%). Among newborns with NWS, 53% were boys, 80.0% were white, 7.2% were black, 7.4% were Hispanic, and 5.3% were of other races. Hispanic neonates had the highest mean hospital charges and LOS of any other ethnic group ($123,749, 21 days). The largest proportion (83.0%) of NWS-related hospital stays were billed to Medicaid, followed by private insurance (10.3%) and self-pay (4.8%). More than one-third of NWS-related discharges (39.3%) occurred in areas with the lowest mean household annual income (≤$42,999) compared with 28.4% of neonates without NWS. Most NWS cases (53%) had ≥5 diagnoses, compared with 11% of non-NWS neonates. In the multivariate analysis, neonates with a birth weight <2500 g, feeding problems, respiratory diagnoses, seizure, >4 diagnoses, LOS >5 days, rural hospitals, Medicaid, and low-income households were significantly associated with NWS. There was a statistically significant mean hospital charge difference of $55,775 between NWS and non-NWS neonates.

Conclusions: Since 2000, the number of infants treated for NWS in the US neonatal intensive care units has increased fivefold, accounting for an estimated $1.5 billion in annual hospital expenditures. The high hospital resource usage among NWS neonates raises the possibility that care for expectant mothers who use opiates and their newborns may be able to be delivered in a more efficient and effective manner. Because the majority of the study population was covered by Medicaid programs, state policy makers should be mindful of the impact the opioid crises continue to have on expectant mothers and their infants.