健保藥價調整對醫院換藥及照護結果之影響：以糖尿病口服用藥為例

Abstract

背景：糖尿病為國人常見之慢性病，盛行率已上升至百分之八，長年高居國人第四大死因。為控制藥費支出，中央健康保險署自2000年4月至2011年12月已調整過七次藥價，每次健保藥價調整皆會影響醫院藥品選用決策。 目的：1) 探討健保藥價調整，在2009年10月1日新核定藥價生效前後各觀察期間，口服降血糖藥品換藥人數比率變化之趨勢。2) 探討研究期間皆在同家醫院內就診之第二型糖尿病人用藥廠牌更換有無對於糖尿病照護結果之影響。 方法：使用國家衛生研究院提供之2009年至2010年之糖尿病特殊需求檔，觀察期間為新藥價核定前半年、以及核定後一年。第一部分進行病人層級與醫院層級之換藥/換廠現象之描述。第二部分以羅吉斯迴歸在控制病人特性及就醫場所特性後，探討病人用藥廠牌更換與否對於糖尿病相關急診之影響。 結果：藥價調整前中後三期，以病人為單位時，換藥比率分別為19.43%、26.70%、和18.68%，以醫院為單位時，換廠比率大於20%之醫院家數占比，分別為17.3%、29.09%和20.64%。公立醫院、區域醫院、中區業務組之醫院平均換廠人數比率在藥價調整期間占比最高。分析各別三期病人用藥廠牌更換與否與糖尿病相關急診之關聯後，發現並無統計上之顯著影響。 結論：藥價調整期間換藥、換廠人數比率增加，但病人用藥廠牌更換對糖尿病照護結果並未發現統計上顯著影響。

Background: Diabetes mellitus in Taiwan is common, the prevalence rate has gone up to 8%; and it has been the fourth rank of the 10 major cause of death. From 2000 to 2011, the National Health Insurance Administration in Taiwan has implemented the drug price adjustments (DPA) for seven times to control the growth of pharmaceutical expenditures. DPA may lead to changes in drug selection among hospitals.

Objectives: This study intended to examine the effect of the 6th drug price adjustments in 2009 on hospital drug brand switching and health care outcome. This study employed oral hypoglycemic agents as an example. Methods: Using the NHI claims data from 2009 to 2010, this study described the trend of drug brand switching at patient level and hospital level half year before and one year after the price adjustment. Then, this study used logistic regression to examine the relationship between drug brand switching and diabetes-related ED visits at patient level. Results: During the pre-adjustment, immediate-adjustment, and after-adjustment periods, the drug switching rate was 19.43%, 26.70% and 18.68% respectively at patient level. At hospital level, the proportion of higher brand switching ratio (>20%) was 17.3%, 29.09% and 20.64% respectively. Furthermore, the average brand switching rate was higher in public hospitals, regional hospitals and the hospitals in central division of NHI. After accounting for potential confounders, including the characters of patients and hospitals, we found no significant association between patients’ drug brand switching and diabetes-related ED visits. Conclusions: During the 6th DPA, the rate of drug and brand switching increased significantly at both patient and hospital levels. This study found no significant relationship between drug brand switching and health care outcome.