Abstracts

Rationale: People with epilepsy who do not respond adequately to 2 different regimens of anti-seizure medications (ASMs) are considered to have DRE. While relevant guidelines recommend considering neuromodulation and other non-pharmacologic alternatives for people with DRE, instead of continued trials of ASMs, available data suggest that such treatments are often deferred for several years following DRE onset. We sought to summarize the healthcare resource use of Commercial (COM), Medicare (MCARE), and Medicaid (MCAID) enrollees with DRE during the 2-year period before neuromodulator implantation.

Methods:

We used data from the Merative Marketscan and US Centers for Medicare and Medicaid Services (CMS) to identify COM, MCARE, MCAID enrollees with DRE who underwent neuromodulator implantation. The earliest date on which implantation was noted was designated as the index date. People without an epilepsy diagnosis on index date were excluded, as were those without any ASM claims in the year prior to implantation and those not continuously enrolled for the two-year period before and including index date ("pre-index"). All available information during pre-index was used to assess selected demographic and clinical characteristics, as well as patterns of use and cost of healthcare services and pharmacotherapy. Analyses alternatively employed an all-cause and epilepsy-related perspective; the latter was defined as all medical care (inpatient and outpatient) resulting in diagnoses of epilepsy and all ASM dispenses.

Results: A total of 6,877 people (COM N=860; MCARE N=3,548; MCAID N=2,469) with DRE met all study selection criteria. The mean age (SD) for each cohort was 26.4 (16.2), 46.6 (13.3), and (14.1) years respectively. The mean Charleson Comorbidity Index (SD) was COM 1 (1.5), MCARE 2.1 (2.3), and MCAID 1.3 (1.5).  The MCARE, and to some extent the MCAID cohort, tended to be frailer than those covered by COM. In the two years before neuromodulation, the cost of epilepsy related healthcare was approximately $92,000 in COM, $45,000 in MCARE and $48,000 in MCAID cohorts.  Over 50% of all people had one or more epilepsy related hospital admissions and over 40% of all people had 5 or more ASMs before neuromodulation. Epilepsy related hospital stays ranged from ~6 days in COM to over 9 days in the MCARE cohort.  In general, MCARE and MCAID people had higher levels of admissions to hospital and ED whilst COM insured people had better access to neurologist and physician office care.  Healthcare resource use in people with DRE is between 2 and 9 times that of the ‘average’ person, depending on the type of insurance coverage.

Conclusions:

Regardless of coverage, people with DRE experienced high levels of healthcare resource use before neuromodulator implantation.  Contrary to best practice guidance, many people experienced extended treatment with unsuccessful ASMs resulting in delays to accessing neuromodulation and avoidable healthcare resource use.  

Funding: LivaNova PLC, Houston, TX