疾病就診率係針對固定期問內，患有疾病而到全民健康保險特約之各級醫療院所就診之人數予以統計，其分析可供規劃醫療保健服務與修訂全民健保制度之參據，因此應成為經常性之統計工作。本文以適合連續性調查的輪換抽樣設計為原則，運用調查當期及較早期之資訊以建搆複合估計式，並推導該估計式之期望值與變異數，據以此比較與簡單估計式之相對效率。為評估複合估計方法運用於疾病就診率之效益，遂選定某些疾病進行分析，並研究在何種權數組合下，複合估計式能減低變動情形及其相對效率。 關鍵字：疾病就診率、輪換抽樣法、複合估計式、相對效率 / The treatment rate is a statistical analysis of numbers of people who had taken treatment in the fixed period. It can be a reference of planning medical service, revising medical laws and so on. So its analysis is important and should be an occasionally statistical work. In this paper, we adopt rotation sampling method, combine current and earlier information to construct composite estimator, and derive its expected value and variance formula, so as to compare with simple estimator. To evaluate the benefit of adopting composite estimator on the treatment rate, we choose some diseases as objectives, and determine the optimal weighted parameter so that composite estimator can reduce the variance and its effects on the relative efficiency. Key words: treatment rate, rotation sampling, composite estimator, relative efficiency