Neonatal seizures are a relatively common marker of brain dysfunction in the newborn. Population based studies from Western countries(1) suggest a relatively low incidence of 2.6/1000. The incidence in outborn babies admitted to Indian NICUs is close to 12%, and reflects the actual reality in our country where babies are born in small hospitals and nursing homes(2).

The recurring question uppermost in both parents and pediatrician’s mind is what will be the long term neurological outcome in a baby with neonatal seizures. For many decades, it has been clear that the etiology of neonatal seizures is one factor critical in determining outcome. Newborns with transient correctible metabolic abnormalities, focal ischemia and without clear etiology do well, while those with hypoxic-ischemic encephalopathy (HIE), CNS infections and cerebral dysgenesis regularly do poorly(1,3). In India, where perinatal care is uneven, transient metabolic disturbances like hypocalcemia and hypoglycemia still account for about a fifth of the neonatal morbidity(2). Also, the outcome of hypoglycemia is not necessarily favorable(4,5). It is difficult, if not impossible, to separate out the impact of the etiology from the impact of seizures themselves on the immature developing nervous system, though recent evidence seems to suggest an independent deleterious effect of both status epilepticus(6) and electrographic seizures(7). Other parameters studied to predict outcome include a neonatal neurological examination(3), EEG abnormalities especially abnormal background activity(1,3), USG and MRI, especially diffusion-weighted abnormalities(3). However, interpretation of the newer methods is technically demanding and can be often misleading.

In this issue of Indian Pediatrics, Iype and colleagues(8) have followed up a large cohort of newborns with neonatal seizures over 2-8 months to assess what happens to them in terms of mortality, neurological outcome and the developmental of post-neonatal epilepsy (PNE). HIE, hypoglycemia and meningitis accounted for the bulk of cases and often multiple risk factors were present. Clinical rather than electrographic recognition of seizures, as the authors have done may lead to overestimates as was shown in a landmark study of video-EEG in neonatal seizures in the 1980s(9). Clonic and focal tonic seizures had EEG correlates confirming their epileptic nature, while generalized tonic, some myoclonic and most subtle seizures had no EEG correlate suggesting that these were non-epileptic phenomena related to an injured neonatal brain. In the Indian setting, availability, affordability and interpretation issues make video EEG an unviable option in more than 90% of nurseries; hence clinical recognition of seizures is perfectly acceptable with the caveat that some non-epileptic phenomena will be misdiagnosed as seizures. Neurological outcome was overall good in the Iype study, with about 2/3 of patients having a normal outcome and only a minority having post-neonatal epilepsy (PNE). Birth weight and gestational age did not correlate with outcome. In a Canadian population-based outcome study of neonatal seizures followed up for >10 years(1), the prognosis was a lot more discouraging with only 35% having a normal outcome and 34% having PNE. The outlook was even worse in preterms, with only 12% considered normal at the end of a decade and 48% having PNE. Full term babies seem to do uniformly better(1,3). As the authors of this study(8) admit, the short follow-up of only a few months may explain this cohort’s better outcome. Milder motor/cognitive deficits may be not recognized in infants and PNE may have its onset later(1). The authors report a normal EEG in the majority of newborns and showed that interictal epileptic discharges (IEDs) correlated with a poor outcome. The prevalence of EEG abnormalities is much higher in Western literature and abnormal background activity rather than IEDs are more correlative of poor outcome(1,3). Unfortunately, the authors have not reported on the EEG background activity. One study, however, did find that temporal IEDs correlated with outcome and PNE(10). Another novel observation in this study was the association of hypocalcemia and neonatal mortality. However, numbers were small (only 6 newborns had hypocalcemia of which 2 died) and confidence intervals wide, making this finding difficult to interpret. In general, hypocalcemia of later-onset has a good long-term prognosis(3).

The major strength of this study is that it is the first prospective study looking at outcome in a cohort with neonatal seizures in the Indian setting. It reiterates the need to reduce correctible etiologies like hypoglycemia and kernicterus which have long disappeared from developed nations. Though the follow-up is short, it does suggest that many survivors of neonatal seizures do well and aggressive treatment is warranted. This study has not studied the newborn neurological examination in predicting outcome. It has been shown that a normal examination is highly predictive of outcome in at least term babies though the converse is not true(3). This tool does not need sophisticated technology and is ideally suited for outcome prediction in the Indian setting.

Funding: None.

Competing interests: None stated.