中央健保局為控制國內醫療費用的成長並有效運用醫療資源，將逐步實施總額預算制度，而「診斷關聯群前瞻性支付」(簡稱DRG/PPS)是新制度中醫療費用分配與支付的基準。為求執行上的公平與正確，如何有效審查進而抑制醫療服務提供者在申請給付時有意或無意間產生的DRGs取巧行為，即成為醫療服務管理的重要課題。此類資訊密集的審查作業目前仍以人工審查為主，實有導入資訊科技的必要。本研究以整合模擬退火演算法與適應性類神經模糊推論系統的方式來建構「尋優適應性類神經模糊推論模式」，並以此智慧型機制進行DRGs取巧行為是否存在的自動檢測。實例驗證的結果顯示：本研究所提模式不僅有高於其他功能相似系統的平均分類正確率(90.10%).更在決策支援方面有較高的透明度及可信賴度。

In order to control the growing costs of healthcare spending, the Bureau of National Health Insurance is adopting the Global Budget System. In this new system, Diagnostic Related Groups/Prospective Payment System (DRG/PPS) is the mechanism for payment evaluation and control. Many healthcare frauds, also known as DRGs creeps, have been found in other countries that have already implemented the similar payment systems. For a fair and effective execution of such payment system, elimination of these frauds becomes an important task for healthcare administration. Detecting DRGs creeps manually is costly. However, costs to tailor packages to fit the detection needs and integrate them into an existing environment are outrageous. In this research, a hybrid soft computing model that integrates simulated annealing and adaptive-network-based fuzzy inference system (ANFIS) is proposed for the automated detection of DRGs creeps. The proposed approach is affordable for healthcare units at all levels. Implementation and evaluation results demonstrate that the proposed hybrid model has improved overall performance in identifying DRGs creeps.