Abstracts

Rationale: Prior research suggests the use of sodium channel blocker antiepileptic drugs (AEDs) is associated with higher risk of hyponatremia among epilepsy patients (Asconapé, 2002; Dong et al., 2005; Lin et al., 2010). Patients with acute hyponatremia may experience symptoms such as headache, confusion, and in some cases seizures and coma may occur. The objective of this analysis was to estimate the prevalence of hyponatremia among epilepsy patients prior to the initiation of AED therapy and assess differences in treatment selection between patients with and without baseline hyponatremia. Methods: A retrospective, longitudinal analysis of patients (aged ≥ 12 years) diagnosed with epilepsy in a large electronic medical records database from 1996 - 2009 was conducted. Patients were required to have ≥1 prescription for AEDs following their initial epilepsy diagnosis. The date of first evidence of AED use defined each patient's index AED date. Selected patients were required to have at least 1 sodium measurement during the 6-month pre-index period. Using the most recent sodium measurement prior to the index AED date (among patients with multiple sodium measurements), patients were categorized as with (sodium measurement ≤ 135 mmol/L) or without (sodium measurement > 135 mmol/L) hyponatremia. Among the two patient groups, differences in pre-index period demographic (i.e., age, sex, race, payer, provider) and clinical (i.e., comorbidity burden using Charlson Comorbidty Index [CCI] score) characteristics and type of AED initiated (i.e., sodium channel blockers and non-sodium channel blockers) were assessed using univariate (Student's t-test and chi-square test) and multivariate logistic regression models. Results: A total of 3,850 patients with epilepsy who initiated AED therapy were identified. Among the selected patients, 13.5% (n=521) had hyponatremia prior to their index AED date. Patients with hyponatremia were older (mean [std dev]: 57.95 [17.59] vs. 52.40 [19.34] years; P<0.001) and had greater comorbidity burden (mean CCI [std dev]: 0.73 [1.38] vs. 0.55 [1.15]; P=0.001) compared with patients without hyponatremia. The presence of hyponatremia did not differ by other demographic characteristics, including sex (P=0.289) or race (P=0.071). Significant differences were observed in the distribution of provider (P=0.004) and payer (P<0.001) type among patients with and without hyponatremia. The proportion of patients initiating a sodium channel blocker AED did not differ significantly between patients with and without (60.1% vs. 59.1%; P=0.155) hyponatremia. Conclusions: In this retrospective, longitudinal analysis of an electronic records database, about 1 in 10 epilepsy patients had baseline sodium levels ≤ 135 mmol/L. Older age and greater comorbidity burden were associated with higher prevalence of baseline hyponatremia. Receiving initial treatment of sodium channel blocker AED is not influenced by the presence of baseline hyponatremia. Future analysis may assess differences in hyponatremia events by treatment post-baseline.