研究背景:過去研究探討藥價調整對於醫師處方行為的影響,多著重在原廠藥與學名藥以及不同成分間互換的情形,較少提及藥價調整幅度不同時,同成分不同廠牌藥物的更換或續用情形。 研究目的:探討2006年11月1日第五次藥價調整前後,在第二型糖尿病主要用藥上使用量的變化情形,以及控制醫院特性及病患特質後,分析藥價調整幅度高低對於第二型糖尿病病患長短期續用糖尿病藥物情形的影響。 研究方法:本研究使用2005~2007年國家衛生研究院發行之100萬人承保抽樣歸人檔,選取在2005~2007年期間每年因糖尿病就診兩次以上及每年拿第二型糖尿病用藥兩次以上之糖尿病患者,屬次級資料分析。第一部份分析2006年9月至2007年4月各成份糖尿病藥物在2006年11月1日第五次藥價調整前後的使用量變化情形。第二部份則以迴歸模型控制醫院特性和病患特質,區分病患使用藥價調整幅度大於等於30%及病患使用藥價調整幅度小於30%者,分析藥價調整對短期及長期續用藥物的影響。本研究為確保資料的穩健性,另外嘗試以藥價調整幅度32%、28%及23%作為藥價調整幅度的分界點進行敏感度分析。 研究結果:2006年9月至2007年4月期間各成份藥物在各月的使用量成長率方面,在2006年11月1日第五次藥價調整前後各成份使用量成長率皆略有增減,在2006年12月以前大致上呈現上升的趨勢,在2007年1月至2007年4月則呈現波動狀態。在控制醫院特性及病患特質後,無論是長期或短期續用情形,病患使用藥價調整幅度大於等於30%藥物之續用情形,均比病患使用藥價調整幅度小於30%藥物之續用情形低。 結論:由第一部分研究結果顯示在2006年11月1日第五次藥價調整前後的各成份第二型糖尿病藥物使用量上無明顯變化。由第二部份結果顯示藥價調整幅度高低會影響到民眾長、短期續用藥物情形,顯示藥價調整幅度高低的確會影響到醫師處方行為。健保局應監控在藥價調整前後,醫師的處方行為改變情形,並應進一步確保病患用藥的品質。
Background Previous studies have focused on the changes between brand-name drug and generic drug or between different elements drugs after the NHI drug price adjustment. Few studies focus on the drug changes of same element but different companies’ drug when the price adjustment rate is different. Objectives To examine the trend of drug amount for type II diabetes before and after the fifth price adjustment on November 1, 2006. Furthermore, to evaluate the short term and long term effects of different price adjustment levels on the continually using diabetes drug while controlling for hospitals’ characteristics and patients’ characteristics. Method Using 2005~2007 National Health Insurance Registry of Beneficiaries and their Claims Data, we chose those who conducted physician visit for diabetes two times or more in every year and received diabetes drugs two times every year during 2005~2007. Part I: Analyze the trend and amount of diabetes drugs’ changes before and after the fifth drug price adjustment. Part II: Define the price adjustment level as high or low using 30% as a cut-off point and define the dependent variable as “continuously use the same drug”. Long term or short term was defined by using the cut-off point on April 1, 2007 after the fifth adjustment on November 1, 2006. This study employed regression model to examine the short and long term effect of price adjustment level on the change of diabetes drugs. In order to increase the robustness of the analysis, the study also used drug adjustment rate of 32%, 28% and 23% for sensitive analysis. Outcome Part I: In term of the amount of diabetes drugs used between September,2006 and April,2007 there is an increasing trend before December,2006, but no clear trend afterward. Part II: After controlling for hospitals’ and patients’ characteristic, the regression results showed that the likelihood of “continuously using same drug” tended to be lower in the group with higher than 30% price adjustment compared with those in the group with price adjustment lower than 30%. While using drug price adjustment rate of 28% or 23% as demarcation point, we found the results were similar to those presented above. Conclusion The NHI fifth drug price adjustment did not affect the total amount of diabetes drug utilization. However, the magnitude of price adjustment had significant effect on the changing drug behavior. The Bureau of National Health Insurance should monitor doctors’ prescribing behavior before and after drug price adjustment and assure the quality of drugs.