摘要: | 研究背景:醫療費用持續成長,但因資源有限,如何有效率使用醫療資源是醫療體系重要的課題。我國的全民健保論病例計酬DRGs給付制度即將全面實施,醫療機構的成本控制更顯重要,面對醫療的不確定性,醫師與醫院的特性如何影響醫療資源耗用是相當值得探討的問題。
研究目的:1. 輕度急性胰臟炎的住院醫療資源耗用的分布情形。 2. 醫師與醫院的特性和住院醫療資源耗用的相關性探討。
研究對象:民國93年1月至94年6月期間在台灣中部的醫學中心級、區域教學醫院、與地區醫院各1家私立醫院住院,出院主診斷為輕度急性胰臟炎(Ranson criteria≦2)的所有個案,同一病人的不同次住院視為不同個案;總共有八位消化內科主治醫師參與本研究。
研究方法:採回溯性設計,由單一研究者進行病歷回顧,收集病人的人口學特性與臨床資料,醫療資源耗用以住院日數和醫療費用衡量,醫療費用以健保申報費用清單為依據,將總醫療費用整理分類為固定費用(包含論日計算的病房費與診療費二項費用)、檢查與專科處置費用、放射線診療費、與藥品費用共四大類,另外調查常用的特定檢查的利用情形。醫師與醫院的特性因素由研究醫院提供。以敘述性統計呈現輕度急性胰臟炎的住院醫療資源耗用的分布情形。利用分類迴歸樹狀分析法(classification and regression tree, CART)探討醫師與醫院的特性因素對住院日數、檢查與專科處置費用、和藥品費用的影響;投入CART的自變項包含醫師、醫院、和病人的特性因素。醫師特性部份包含年齡、最高學歷、專科年資、是否擔任教職、每週門診次數、與收入來源等六項;醫院因素則有評鑑級別與有無實習醫師二項;病人因素包括性別、年齡、住院途徑、Charlson Comorbidity Index、Ranson criteria、致病原因、有無發燒與急性胰臟炎的併發症等八項。
結果:輕度急性胰臟炎的住院醫療資源耗用的分布範圍很大,總醫療費用介於4,601-104,169元,平均值為22,659元,中位數為17,242元;住院日數介於1-21,平均值為5.63,中位數為5;固定費用介於1,000-33,849元,平均值為7,966元,中位數為6,822元;檢查與專科處置費用介於512-34,040元,平均值為6,120元,中位數為3,739元;放射線診療費用介於0-37,155元,平均值為3,250元,中位數為760元;藥品費用介於632-50,329元,平均值為5,259元,中位數為2,666元。腹部超音波是最常被使用的特殊檢查,其次是腹部CT/MRI。住院醫療費用結構,以固定費用(與住院日數相關)和檢查與專科處置費用所占比例最大。分類迴歸樹狀分析法發現影響住院日數最重要的因素是病人特性中的併發症,其次是發燒與否,醫師與醫院特性中,醫師的重要性大於醫院。影響檢查與專科處置費用最重要的因素是病人因素中的致病原因(膽結石)與併發症,對於沒有併發症的非膽結石引起的急性胰臟炎,影響檢查與專科處置費用最重要的因素是醫師的專科年資。影響藥品費用最重要的是病人因素中的併發症與發燒,其次是醫師特性中的每週門診次數、專科年資、與年齡,醫院因素的影響不明顯。
結論:輕度急性胰臟炎的住院醫療資源耗用在不同的醫師與醫院之間有顯著的變異性,醫學中心級的醫院有較高的資源耗用。住院醫療費用結構,以固定費用(與住院日數相關)和檢查與專科處置費用所占比例最大,我國的住院日數高於美國。影響輕度急性胰臟炎的住院醫療資源耗用最重要的因素是病人的特性,其次是醫師因素;醫院的影響並不顯著。後續應該有更大規模與廣泛性的研究,深入探討影響醫療資源耗用的因素。
Background and Aim: Increasing health care costs have stimulated interest in cost-effective medicine and have fueled the debate regarding the relative roles of physicians and hospitals in generating costs. The aims of this study were to investigate the association of physicians and hospitals’ characteristics with medical resource utilization and to examine the distribution of hospital charges in mild acute pancreatitis.
Methods: The study design was a retrospective secondary data analysis in three private hospitals (8 gastroenterologists in total). All patients (n=170) admitted due to mild acute pancreatitis between January 2004 and June 2005 were recruited for the study. Patient clinical information was obtained from medical charts which were reviewed by a single researcher. Medical resource utilization was assessed using total and sorted medical charges obtained from the registration data files of hospitals. Descriptive analyses were performed to understand the distribution of hospital charges. Classification and regression tree (CART) analysis was used to examine the association of physicians and hospitals’ characteristics with medical resource utilization while controlling for the patients’ factors. The dependent variables were as follows: length of stay (LOS); checkup and management charge; and medication charges.
Results: The medical resource utilization varied considerably. On average, the patient’s LOS was 5.63 days (range 1-21; median 5), total hospital charge TWD 22,659 (range 4,601-104,169; median 17,242), fixed-daily charge TWD 7,966 (range 1,000-33,849; median 6,822), checkup and management charge TWD 6,120 (range 512-34,040; median 3,739), radiology charge TWD 3,250 (range 0-37,155; median 760), and medication charge TWD 5,259 (range 632-50,329; median 2,666). CART analysis indicated that patients’ factors account for the most of differences. Physicians’ factors, especially age, also play an essential role.
Conclusion: The medical resource utilization varied considerably in mild acute pancreatitis, mostly resulted from patients’ factors, some from physicians’. Further studies leading to a more cost-effective health care system is necessary. |