Objective: To evaluate the healthcare and societal 
economic burden of adjunctively-treated patients with 
uncontrolled epilepsy. 
Methods: A retrospective claims database (representative 
of the commercially insured US population) was used to 
analyze direct and indirect costs (Jan-2006 to Dec-2011) for 
patients with epilepsy. Patients had ≥1 diagnosis code for 
epilepsy (ICD-9 345.xx), ≥18 years of age, and had 
received ≥1 add-on Anti Epileptic Drug (AED) within 60 
days of first-line treatment. Patients were grouped into: 
Well-controlled (no change in AED and no epilepsy-related 
Emergency Room (ER) or inpatient visits), Uncontrolled 
(≥2 AED changes in ≥30 days and ≥1 Epilepsy-Related ER 
or inpatient visits), and Intermediate-controlled (not 
grouped as Uncontrolled or Well-controlled). The groups 
were matched based on propensity scoring in a 1:1:1 ratio 
including age, sex, and Charlson comorbidities index score. 
A mixed-effects model was used to calculate the adjusted 
cost differences. 
Results: From 141,173 patients, 6,785 triads were included 
in the analysis. Mean age was 48±16.6 years (62% female). 
Uncontrolled patients had significantly higher direct costs: 
Uncontrolled vs. Well-controlled: $32,077; Uncontrolled 
vs. Intermediate-controlled: $14,855. Indirect costs were: 
Uncontrolled vs. Well-controlled: $5,016; Uncontrolled vs. 
Intermediate-controlled: $2,750. Differences were due to 
hospitalization costs (Uncontrolled/Well-controlled: $15,297; Uncontrolled/Intermediated-controlled: $9,123), 
outpatient visits (Uncontrolled/Well-controlled: $9,248; 
Uncontrolled/Intermediated-controlled: $4,534), AED costs 
(Uncontrolled/Well-controlled:
 $4,546; 
Uncontrolled/Intermediate-controlled: $3,079), sick leave 
costs
 (Uncontrolled/Well-controlled:
 $4,849; 
Uncontrolled/Intermediate-controlled: $2,577), and total 
work
 loss
 (Uncontrolled/Well-controlled:
Uncontrolled/Intermediate-controlled:
 $2,750)
 $5,016; 
 (all 
p<0.01).Adjusted costs were significantly higher for 
Intermediate-control vs. Well-controlled patients for all 
cost items (all p<0.01), with the exception of non-AED 
costs and disability, which were not significantly different. 
Conclusions: In this study, direct and indirect costs were 
highest for uncontrolled, and lowest for well-controlled 
epilepsy. These findings suggest the importance of 
improvements in control of seizure activity (either from 
uncontrolled
 to
 intermediate-controlled,
 or
 from 
intermediate-controlled to well-controlled), both from a 
societal cost and public health perspective.