台灣版診斷關聯群權重影響因素之探討 --- 以某醫學中心為
例
雖然全民健保業已推行 10 年，不僅獲致民眾高度滿意的評價亦是社會安全網中極重要的一環，然在
開辦初期推行 ?? 計酬的給付制度不免造成醫師誘發醫療需求的行為增加，而產生醫療浪費之情事。
故漸以前瞻性支付制度取代，期與醫療服務提供者共同承擔財務風險責任，以達到全民健保「確保醫
? 資源使用效 ? 」的目標。
加護病房是為醫療資源耗用密度極高的單位，對收治急重症患者為對象之醫學中心更是如此；美國在
實施診斷關聯群組之後，出現該類病患為沒人收治的醫療人球，加護病房的利用率因此減少 1/3 。
為避免前述情狀發生，本研究之目的在於檢討我國在導入此支付制度時，是否尚需將此類患者列入分
類考量，。首先，依據文獻所指出，疾病嚴重度、加護病房住院天數及呼吸器使用天數是反映其資源
（費用）耗用的重要變項，遂將以上變項納入研究分析，檢視「中央健康保險局第三版住院診斷關聯
群」（以下簡稱為 Tw-DRGⅢ ）之相對權值 (Relative Weight ， RW) （以下簡稱為權重）是否反應以
上資源耗用之差異。
本研究以 Teradata SQL-Queryman 為資料管理軟體，以 SPSS 10.0 為統計分析之研究工具。採回溯性
資料分析，以某醫學中心 93 年 7 月至 94 年 6 月期間住院病患為研究對象，由個案醫院健保申報資料
中的健保醫事檔描述個案醫療耗用情形。依全民健康保險住院診斷關聯群支付通則草案第三版規定，
以同一研究對象住院日期進行歸人，排除不符 Tw-DRGⅢ 支付之個案，透過健保局住院診斷關聯群編
審網站下載編審單機版進行分類編派，除探討個案醫院住院病人診斷關聯群結構外，以病患特質、資
源耗用特質及住院費用，瞭解其與 Tw-DRGⅢ 之權重之相關性，本研究之結果如下 :
一、 TW-DRGⅢ 權重與住院病人特質有正相關（年齡、病患疾病嚴重度）；經不同性別、病患來源
、出院狀態、科別分析，顯示男性、非急診、死亡者其權重較高，且達統計上的差異。
二、 TW-DRGⅢ 權重與病人資源耗用特質有正相關（住院天數、加護病房住院天數、呼吸器使用天
數）；經分析使用加護病房、呼吸器及開刀房處置與否，顯示有使用加護病房、呼吸器及開刀房處置
者其 TW-DRGⅢ 權重較高。
三、 TW-DRGⅢ 權重與病人醫療費用有正相關。
四、病患特性、資源耗用特質及醫療費用經強迫進入迴歸分析顯示，年齡、病患來源、出院狀況、疾
病嚴重度、就醫科別、住院天數、呼吸器使用、呼吸器使用天數、加護病房使用、加護病房使用天數
、開刀房處置及醫療費用等預測變數對 TW-DRGⅢ 權重會產生影響，解釋力達 55.8% ，相較於僅以
醫療費用為權重預測變數的解釋力 36% 為高。
An Exploration of Factors Influencing Tw-DRG Relative
weights － The Example of A Medical Center
The universal health care system was implemented in Taiwan over ten years.It acquires the people's good evaluation an d become an important component of the social safety net. It is also an inevitable problem that fee-for-service reimbursement system led to physician induced demand and inefficient distribution of medical resources, so The Bureau of the National Healt h Insurance (BNHI) replace the payment system by prospective payment system, to expects. Healthcare providers bear the fin ancial risks to get the goal of utilizing the medical resources for health care more efficient.
Patient care of ICU is the most consumption for medical resources , especially to the medical center.As the experience of DR Gs/PPS implemented in U.S. utilization of ICU were.reduce about 1/3, and DRG dumping was happen to the patients who sic k severity.To avoided the same sitaution in the future ， this study tried to examine difference between the Relative Weight of Tw-DRG and related medical resources utilizing factors （ The characteristics of patient demography ； resources utilizatio n ； Medical expenditures ） by DRGs the 3nd version.
This study design is based on hospitalized secondary retrospective data.Since July, 2004 to June, 2005 from a medical center.
The data anaysis by SPSS 10.0 version and Teradata SQL-Queryman.First, the inpatients data were screened by who should p ay by DRGs payment system and grouped individual into DRGs by the 3nd version software of medical information service s ystem released from BNHI.The results for this research as followings:
1.The effect of on RW of Tw-DRG and patients characteristics Ⅲ
There were significant positive correlations between RW of Tw-DRG with age, Charlson Comorbidity Index Ⅲ ； The differe nce on RW of Tw-DRG is different due to different gender, source of patient, discharge status, medical departments. Ⅲ 2.The effect of on health resource consumption characteristics on RW of Tw-DRGⅢ ：
There were significant positive correlations between RW of Tw-DRG with length of stay, length of stay in ICU, length of st Ⅲ ay on ventilator ； The difference on RW of Tw-DRG is different due to operation procedure, ventilator using, ICU using. Ⅲ 3.The effect of Medical expenditures on RW of Tw-DRG Ⅲ
There were significant positive correlations between RW of Tw-DRG with Medical expenditures. Ⅲ
4.From the results of multiple regression analyses, patients age,Charlson Comorbidity Index, source of patient, discharge statu
s, medical departments, length of stay, length of stay in ICU, length of stay on ventilator, operation procedure, ventilator usin
g, ICU using operation procedure, ventilator using, ICU using were significantly associated with higher RW of Tw-DRG , w Ⅲ
ith the adjusted R square of 0.558.
