A cross-sectional study was conducted at the pediatrics department of a public tertiary care hospital from November, 2017 to April, 2019, after obtaining clearance from the institutional ethics committee. A written informed consent was obtained from the parents/guardians/caregivers.

Children aged 2 to 12 years either having an active breakthrough seizure or those who had a breakthrough seizure within the last 24 hours and receiving either one or two anti-seizure drug (phenytoin, carbamazepine, sodium valproate) were enrolled from the emergency room or the outpatient department. Patient’s demographic, clinical, and investigation details (serum sodium and potassium, blood glucose, and total leukocyte count) were recorded in a pre-designed case record form. Compliance with the drugs was ensured after taking a detailed history. For this study, a breakthrough seizure was defined as seizures in a child with epilepsy on one or two anti-seizure drugs, who did not have any seizure activity in the last one month. Generic form of anti-seizure drugs were available free of cost to the patients from the hospital formulary. The dose of antiseizure drug and treatment for epilepsy was given as per the guidelines followed in the hospital. Children who had received a loading dose of any antiseizure drug before presentation, those with deranged blood glucose/serum electrolyte levels, history of myoclonic or absence seizure, and known patients of chronic liver disease/chronic renal disease, meningitis, syndromic epilepsies, and known non-compliance (missed >3 doses of anti-seizure drug) were excluded.

A venous blood sample (3 mL) was collected aseptically in serum vial before the loading dose of the antiseizure drug. Serum was separated within 30 min and stored at 0-8 ºC. Drug level estimation was done using CEDIA II kits (Thermofisher Scientific/2018) based on recombinant DNA technology to create a homogenous immunoassay system. The minimum detectable concen-tration of CEDIA Phenytoin II, Carbamazepine II, and Valproic acid II assays was 0.6 mg/L, 0.5 mg/L and 3.0 mg/L. The reference range for serum levels of phenytoin, carbamazepine, and sodium valproate was considered between 10-20 mg/L, 4-12 mg/L, and 50 to 100 mg/L, respectively, as per laboratory and kit specifications.

According to a previous study [8], serum levels of first-line antiseizure drug were in the below reference range in 40% of cases with breakthrough seizures. Using the formula for hospital-based population proportion with confidence interval 95%, estimated prevalence of 40%, power of 80%, a error of 5%, and acceptable absolute precision of ±8% (acceptable relative precision of 20%), the sample size was calculated as 144. Thus, 145 children were planned to be enrolled.

Statistical analysis: Analysis was performed using SPSS version 20 software. Pearson Chi-square / Fisher exact test was used to assess the association between demographic and clinical variables with below and normal reference range of antiseizure drug; the above range subjects were not included in the comparison. The 95% confidence level of below reference range proportion was determined using the binomial Wald method. P value of <0.05 was taken as significant.

A total of 145 children (97 males) were enrolled with a mean (SD) age of 6.9 (3.0) years. Generalized seizures were seen in 93 (64.1%) children, and normal development was seen in 90 (62.1%) children. Microcephaly was found in 26 (17.9%) children. All children had baseline laboratory parameters within the normal range. A single drug was being used by 138 children (sodium valproate-104, phenytoin-26, and carbamazepine-8). Two drugs were prescribed in 7 children (5-phenytoin/sodium valproate and 2-carbamazepine/sodium valproate).

Drug levels in the below reference range were found in 64 (44.1%), within the reference range in 70 (48.3%), and above reference range in 11 (7.6%) children. Children with polytherapy had drug levels in below reference range in five children on phenytoin, one on carbamazepine and three on sodium valproate therapy. The proportion of children and the serum level of different drugs is shown in Table I. The associated clinical and demographic variables were comparable across the various drug levels of different anti-seizure drugs (Table II). Children with focal seizures had twice the number of children in the reference range group as compared with the below reference range group (P<0.05). Precipitating factors were fever in 19 (13.1%) and sleep deprivation in 4 (2.8%) children. Precipitating factors were not significantly associated with anti-seizure drug levels (Table II).