OBJECTIVES: The aim of the study was to provide the information about potential impacts on the patient distribution and national expenditure of a treatment flow optimization toward specialists in epilepsy care.

METHODS: We developed a Markov model to simulate a flow of epilepsy patients who refer to specialists from non-specialists, and to surgery from specialists for estimation of current patient distributions and expenditure. The patients were assumed to initiate treatments with a single drug by non-specialists and progress to combination therapy with second, third and fourth drugs, on necessity. The patients were assumed to refer to the specialist at each treatment line transition with a certain probability. The treatment failure rates and the transition latency were set by literature and expert opinion. The referral rates for specialists/surgeries were calibrated to reproduce the annual number of surgeries in the Japanese national statistics. The patients could return to the single drug treatment if the seizures were controlled after surgery. The incidence of newly-patients were estimated from epidemiological data. BIA was also performed for the optimized treatment flow with increased referral rates to specialists from the current condition.

RESULTS: The number of newly-pediatric-epilepsy patients in Japan was estimated to be approximately 7,200/year and 0.6 million pooled population were analyzed. In the current treatment flow, the number of refractory patients treated with >4 drugs by non-specialists were estimated as 0.1 million and yielded 61 billion JPY in annually. However, in the optimized treatment flow, the number of refractory patients in non-specialists was significantly decreased, and the costs were reduced by 30 billion JPY and cost-saving of 0.4 billion JPY in total national expenditure would be expected.

CONCLUSIONS: This study highlights that any policy decision making for referral optimization to specialists in appropriate epilepsy patients would be feasible with a very few budget impacts