Abstracts

Economic evaluations in epilepsy have been limited by the lack of valid and generalizable information on changes in health care use in medically refractory patients. We report changes in patterns of use and their relationship to seizure control from 2 years before to 2 years after evaluation for surgery in 68 patients treated at 5 centers that used a common evaluation protocol. Subjects were refractory temporal lobe epilepsy (TLE) patients in the Multicenter Study of Epilepsy Surgery cohort. Health care use in the 4 consecutive 6-month periods before evaluation, during evaluation, and in the 4 consecutive 6-month periods after evaluation was abstracted from records of community- and center-based providers. Use was summarized as Year 2000 dollars by Medicare cost-accounting methods. Subjects were either seizure-free after surgery (SF group), had persisting seizures (1 or more) after surgery (PS group) or were evaluated, but did not have surgery (NS group). 68 of 147 subjects (28 in the SF group, 30 in the PS group and 10 in the NS group) had sufficient records to include in the analysis, as determined by a priori criteria that ultimately limited imputed use from missing records to [lt] 3% of total use. Included subjects were more likely to have surgery and characteristics associated with lower surgical risk or better outcome, but otherwise did not differ from excluded subjects. Seizure-related use increased in all groups from $1187 in the first 6 month period to $3166 in the 6 months before video-EEG evaluation (p[lt].0001). Newer, but not older, AED use increased in all groups across these periods (p [lt].01). By the last 6 month period after evaluation, seizure-related use had declined in the SF group (p[lt].01, mean change in 6 month use = $921), due to declines in AED and inpatient use (p[lt].01, p[lt].05). Seizure-related use did not change in the PS and NS groups (both p [gt] .40). SF and PS groups decreased use of older AEDs (p[lt].01, p[lt].05), but only the SF group also decreased use of newer AEDs (p[lt].01). All groups increased use unrelated to seizures after evaluation (all p[lt].05). This was relatively small (mean increase in 6 month use = $26), but it excluded non-AED medication use, which could not be captured completely. Complete seizure control after surgery for TLE is associated with a decline in seizure-related health care use by 2 years after surgery, due to lower inpatient and AED use, particularly of newer AEDs. Seizure-related use does not change in unoperated subjects or in those who have seizures after surgery. All groups increased unrelated use over time, but this appears to account for a small proportion of overall costs. (Supported by RO1 HS09986 [amp] RO1 NS32375.)