因應高齡多重慢性病及急診擁塞，台灣於2015年試行醫院整合醫學制度（Hospitalist，簡稱HP），已有部分醫院參與。儘管先前國際研究HP表現逾20年，但關於HP能否改善照護品質仍有分歧，本文藉由文獻回顧的綜合發現，探討台灣HP模式對住院照護品質之影響。作者搜尋PubMed資料庫、華藝線上圖書館以及衛福部HP計畫成果發表會議研究摘要，發表或陳現台灣HP效率、臨床結果之初級資料被納入，排除概念性文章，23份文獻中17份符合準則納入分析，並根據Donadedian品質三面向（結構－過程－臨床結果）衡量，摘錄品質資料以評估HP之表現。結果顯示台灣HP病房平均醫師人力為7人、病床46床，HP排值模式為二班（85.71%）或三班（14.29%），醫師以內科專科領域為主（65.71%）。與傳統病房模式比較，47項品質衡量中31項呈現HP表現較佳（65.96%）、16項HP表現較差（34.04%）。整體而言，HP模式能降低轉入加護病房率（50%）、醫糾件數（100%），但病人滿意度降低（66.67%）；HP模式能縮短住院天數（50%）、醫療費用降低（100%）、急診待床比率降低（50%）；HP模式能降低14天/30天再住院率（87.50%）、48小時重返急診（100%），但死亡率提高（71.43%）。整體而言，相較於傳統住院照護，HP是有效率的提供者（住院日數、醫療費用減少），HP的臨床品質與傳統模式相當（再入院率降低），然而，死亡率、病人滿意度尚待改善。多數文獻沒有校正干擾因素可能影響研究結果的外部效度。

Coping with aging, multiple chronic diseases, and hospital emergency congestion, Taiwan implemented a trial program of hospitalist (HP) care provision in some hospitals in 2015. Although more than two decades of international research has examined the performance of HP models, disagreement remains regarding whether HP improves inpatient care quality. This literature review summarizes findings from comparative evaluations to explore the effects of HP models on care quality in Taiwan. We searched PubMed databases, Airiti Library, and Ministry of Health and Welfare national conference reports published or presenting primary data on efficiency and clinical outcomes in Taiwan’s HP systems. A total of 23 articles were identified; conceptual articles were excluded. The remaining 17 articles meeting the inclusion criteria were reviewed. According to Donabedian’s three-concept measures of structure, processes, and outcome, comparative evaluations of quality between HP systems and attending physicians were analyzed. Of publications reviewed, average manpower was seven physicians per 46 beds in HP wards. HP duty scheduling was either three shifts (57.14%) or two shifts (42.86%). Internal medicine was the most common specialty (65.71%). HP models outperformed the traditional ward model in 31 of the 47 quality measures (65.96%); however, 16 evaluations revealed worse performance (34.04%). The reviewed articles demonstrated that HP care reduces admissions to intensive care units (50%), medical malpractice (100%), patient satisfaction (66.67%), average length of stay (50%), medical expenditures (100%), bed waiting times (50%), postdischarge 14-day and 30-day readmission rates (87.5%), and 48-hour returns to the emergency department (100%); however, mortality was higher (71.43%). In summary, Taiwan HP wards are more efficient providers of inpatient care than traditional wards are based on reductions in total medical expenditures and length of stay. The clinical quality of HP wards is comparable to that provided by traditional wards; however, patient satisfaction and mortality are not uniformly improved. Failure of some of the reviewed studies to adjust for confounders may affect the external validity of our results.