目標：基於台灣居高不下的剖腹產率，健保局於2005年5月將自然產支付點數加倍和剖腹產點數一致，期望藉由加倍自然產給付來降低剖腹產率。本研究分析該給付調整是否達成其政策目的。方法：本文利用2003至2007年健保資料庫頭胎生產案件，估計上使用邏輯機率模型
(logistic model)、固定效果(fixed-effect)和多層次(multi-level model)模型來分析給付調整對降低剖腹產利用的影響，並觀察長(2003-4 v.s 2006-7)、短(2004 v.s 2006)期有無不同效果。此外，本文將剖腹產進一步依原因區分為一般性剖腹(scheduled cesarean sections)及緊急剖腹(emergency cesarean sections)兩種，觀察政策效果是否依剖腹緊急程度而有不同。結果：在控制產婦年紀、醫師接生時點，和醫師特性的考量等因素下，各種模型均顯示提高自然產給付對頭胎樣本不論是短期或長期時，一般性和緊急剖腹產率上皆無顯著變化。結論：這結果隱含醫師接生所得可能不是決定生產方式的主要因素，健保局需考慮其他方法來降低剖腹產。

Objectives: In light of the high cesarean section (c-section) rate in Taiwan, in May 2005 the Bureau of National Health Insurance (BNHI) doubled the payment for vaginal delivery to the same amount it paid for delivery by c-section. This study investigated whether this payment
change effectively reduced the c-section rate in Taiwan. Methods: We obtained information about all obstetric cases between 2003 and 2007 from National Health Insurance data. Logistic, fixedeffect,and multilevel models were utilized to determine if the payment increase lowered the short term (2004 vs. 2006), or long term (2003-4 vs. 2006-7) c-section rates for first-borns. Additionally, we separated c-sections into two groups based on their causes, scheduled or emergency, to examine if the payment increase produced different effects on these two groups. Results: After controlling for the women age at delivery, birth order of the child, provider characteristics, and time of delivery, results of all models indicated that the payment change produced almost no effect on the reduction in scheduled or emergency c-section rates for first-borns, in either the short or long term. Conclusions: Our findings indicated that the reimbursement scheme for deliveries might not be the key factor for obstetricians in determining the use of c-section. BNHI might consider other policy instruments in seeking to lower the c-section rate.