Abstracts

Rationale: A previous analysis compared insurance claims data on healthcare costs in patients with probable Lennox−Gastaut syndrome (LGS) before and after initiation of clobazam (CLB) treatment. This method did not control for increasing cost trends in the patient population prior to initiation, thereby potentially underestimating the overall budget impact of CLB initiation. This study compares the previous methodology to two alternative analyses, one that controls for preexisting cost trends in the probable LGS population and another that evaluates the impact of CLB availability on the probable LGS population regardless of CLB use.  Methods: In Studies 1 and 2, annual costs in probable LGS patients were compared during the year before and after CLB initiation; Study 2 also included a time-series analysis comparing area under the curve (AUC) of monthly costs after CLB initiation to extrapolated costs in the same population if CLB had not been initiated, thereby controlling for cost trends prior to CLB initiation (Table 1). In Study 3, average costs per patient per month (PPPM) were compared before and after the commercial availability of CLB in probable LGS patients, regardless of CLB use or the amount of time enrolled before and after CLB availability.  Results: There was a significant increase in annual all-cause total costs in the year after CLB initiation compared to pre-CLB in Study 1 ($73,319 v $81,389, P<.001) and Study 2 ($74,864 v $89,161, P=.004, Table 2). Annual all-cause medical costs were also increased following CLB initiation in both studies (Study 1: $57,090 v $59,292, P<.411; Study 2: $65,791 v $77,450, P=.028). Annual epilepsy-related medical costs were decreased following CLB initiation in Study 1 ($23,740 v $19,958, P=.004), and were slightly decreased in Study 2 ($23,839 v $22,018, P=.216). When increasing pre-CLB costs were controlled by calculating the difference between observed post-CLB costs and extrapolated costs if the pre-CLB trend had continued into the post-CLB period, there were reductions of $1,539 and $2,236 in all-cause and epilepsy-related annual medical costs. In Study 3, there was a significant decrease in PPPM all-cause total costs following the commercial availability of CLB, regardless of CLB use ($3,752 v $2,879, P < .001), and a decrease in PPPM all-cause ($3,403 v $2,584, P < .001) and epilepsy-related ($818 v $710, P < .001) medical costs. Conclusions: Studies 1 and 2 found increased annual costs in LGS patients after initiation of CLB; however this method did not take into account increasing costs in the population before CLB initiation. In a time-series analysis in Study 2, which controlled for increasing costs before CLB initiation, there were cost reductions following CLB initiation. Study 3 found that there was a decrease in costs following CLB availability in LGS patients, regardless of CLB usage. Alternative methodologies such as those used in Studies 2 and 3 may be less susceptible to underestimating the cost benefits of CLB introduction in the LGS population. Funding: Funded by Lundbeck LLC