BACKGROUND

Treatment duration for acute symptomatic neonatal seizures is highly variable. Antiseizure medication (ASM) may be discontinued before or maintained at the time of discharge from the neonatal seizure admission.

OBJECTIVES

We assessed whether ASM treatment duration after acute symptomatic neonatal seizures alters the following: The risk of abnormal functional neurodevelopment or postneonatal epilepsy by 24 months corrected age (aim 1). The hospital length of stay (LOS) for the neonatal seizure admission (aim 2). Parent well-being (aim 3).

METHODS

We conducted a prospective, observational comparative effectiveness study of ASM duration for 303 survivors of acute symptomatic neonatal seizures (seizure onset <44 weeks postmenstrual age, excluding transient, genetic, and congenital infectious causes of seizures) who were born between July 2015 and March 2018 and were enrolled at 9 Neonatal Seizure Registry sites. At 24 months, functional development was assessed using the Warner Initial Developmental Evaluation of Adaptive and Functional Skills (WIDEA-FS) instrument (primary outcome). Clinical neonatal data, including LOS, were extracted from the neonatal records. Parent-reported well-being measures were collected at hospital discharge and at 12, 18, and 24 months. To enhance causal inference, we adjusted the risk of each outcome for the propensity to receive ASM at discharge. We hypothesized that WIDEA-FS scores of children whose ASMs were discontinued before hospital discharge would be noninferior to scores of children whose ASMs were maintained at discharge. The study was powered to demonstrate noninferiority if the lower bound on the CI for the WIDEA-FS was <0.5 SD of the WIDEA-FS score (7%, or 12 points).

RESULTS

ASMs were prescribed at discharge for 64% (higher propensity for those with high seizure burden, complex clinical course, and abnormal discharge examination) of infants. Among infants whose ASMs were discontinued before discharge, ASMs were prescribed for a median of 6 (IQR, 3-11) hospital , compared with a median of 4 (IQR, 3-8) for infants whose ASMs were maintained. Unadjusted 24-month WIDEA-FS scores (measured in 270/300 [90%] of survivors) were 4 points (2%) higher for children whose ASMs were discontinued before hospital discharge than for children maintained on ASMs (median, 165 [IQR, 150-175], compared with median, 161 [IQR, 129-174], respectively; = .09). In propensity-adjusted analysis, WIDEA-FS scores were noninferior for children whose ASMs were discontinued before hospital discharge compared with children whose ASMs were maintained (average difference, +3 points; 90% lower CI of −4 was within the a priori noninferiority margin of −12 points). Epilepsy (recurrent unprovoked seizures) developed in 13% (37/282) of infants before age 24 months; 5% had infantile spasms (13/282). The risk of epilepsy did not differ by ASM treatment duration (11% for infants whose ASMs were discontinued, compared with 14% for those maintained on ASM; odds ratio [OR], 0.8; 95% CI, 0.4-1.6; = .49; propensity-adjusted OR, 1.5; 95% CI, 0.7-3.4; = .3). The median LOS was 15 (IQR, 9-29) days. The point estimate of the LOS was shorter for children whose ASMs were discontinued (median, 13 [IQR, 8-25] days) than for those maintained on ASMs upon discharge from the neonatal seizure admission (median, 17 [IQR, 10-32] days; = .07]. After propensity adjustment, LOS was similar between groups. In a stratified analysis, the effect was most pronounced for newborns with seizures due to ischemic stroke (average estimated effect, 36% shorter hospital stay; 95% CI, 59% shorter to 2% longer; = .06). Parents reported high rates of symptoms of depression (54%) and anxiety (32%) at the time of hospital discharge. Several characteristics of parents (eg, higher maternal education and greater impact on family) and infants (eg, hypoxic-ischemic encephalopathy) were associated, with medium-to-large standardized effect sizes (ranging from 0.52 to 0.78), and with poorer parent and family quality of life and well-being. At 24 months, in both the unadjusted and adjusted analyses, there were no consistent differences in well-being measures between parents of children whose ASMs were discontinued before discharge and those whose children were maintained on ASM. In a comparison of parent well-being over time for children whose ASMs were discontinued with parents of children whose ASMs were maintained on discharge from the neonatal seizure admission, a multivariate analysis showed that the impact of being discharged on ASMs differed based upon the specific dimension of parent well-being. Answers to open-ended questions provided rich information for families and clinicians.

CONCLUSIONS

Discontinuing ASMs before discharge from the neonatal seizure admission for infants with acute symptomatic neonatal seizures is safe, as this practice is not adversely associated with risk of functional neurodevelopmental impairment, postneonatal epilepsy, or LOS, and it variably affected parent well-being.

LIMITATIONS

Although the study was well powered to determine noninferiority for the primary outcome of functional development at age 24 months, postneonatal epilepsy was a relatively rare outcome: only 13% of the cohort developed epilepsy. Thus, although there was no significant difference in epilepsy development between the treatment duration groups, the analysis suggests the possibility of up to 3.4 times the odds of developing epilepsy among infants whose ASMs were discontinued before hospital discharge, compared with those whose ASMs were maintained. Parent-measured outcomes were limited by self-report at each time point, and responses were mostly from the mother, with a minority of responses from the father.