2010-2014台灣醫院整合醫療主治醫師之特徵、趨勢與臨床表現

Abstract

研究背景:近20年來，醫院整合醫療主治醫師（hospitalist）的出現改變了美國住院醫療的面貌，台灣則2000年初逐漸開始引進此一制度用於照顧住院病患，本研究的目的是探究2010-2014年，台灣醫院整合醫療主治醫師的背景、特徵與趨勢，同時分析其照護一般內科疾病住院病患的結果。 方法:本研究利用衛福部健康資料加值中心之全民健康保險申報檔，以全人口檔進行分析。依照文獻及本土資料驗證，定義「醫院整合醫療主治醫師」如下：每年至少有100筆以上之門住診療費申報，且住院診療申報佔總診療申報的80%以上，排除重症加護醫師（intensivist）後，定義為醫院整合醫療主治醫師。於2010至2014年期間，使用描述性統計分析此類醫師的人數、人口學特徵、專業背景、服務醫院層級與地區等。在照護結果方面，選擇一般內科四大常見住院疾病（肺炎、泌尿道感染、慢性阻塞性肺病急性發作、腸胃道出血），比較醫院整合醫療主治醫師，與非醫院整合醫療主治醫師（對照組）之照護結果。本研究選擇之照護結果包含四大項：住院死亡率、總住院天數、總住院申報費用、出院後30日內死亡率。推論式統計採用廣義估計方程式（generalized estimating equations）以處理重複住院與醫師病患朝狀結構之問題，並校正包含病患人口學特質、經濟狀況、共病症指數、就醫層級、區域及醫師服務量等。P值小於0.05定義為統計顯著。 結果: 依據全國專科醫師登錄，共篩選13207內科、家庭醫學科、神經內科專科醫師. 符合定義為醫院整合醫療主治醫師者，在2010年有93人，在2014年有73人，人數與佔比呈現下降趨勢。同一時期中，住院診療申報佔總診療申報低於20%者，佔比呈現逐上升趨勢，住院診療申報佔總診療申報介於60%-80%者，佔比呈現逐漸下降趨勢，顯示整體執業趨勢仍偏向門診醫療。醫院整合醫療主治醫師多為男性（佔70%以上），背景以內科專科醫師居多（佔80%以上），平均年齡為40-45歲之間，以上三特質於2010-2014年期間無明顯趨勢變化。醫院整合醫療主治醫師大多執業於北部與南部都會區，且呈現於醫學中心執業之趨勢，比例由2010的34.4%上升到2014年的49.3%。病患照護結果經迴歸校正後，相較於對照組，醫院整合醫療主治醫師在照護肺炎病患時呈現較低的死亡率，但於泌尿道感染、慢性阻塞性肺病急性發作、腸胃道出血則無差異。肺炎、泌尿道感染、慢性阻塞性肺病急性發作三者的住院天數經校正後無統計差異，但腸胃道出血病患的住院天數在接受醫院整合醫療主治醫師照護組顯著較長。總住院申報費用方面，醫院整合醫療主治醫師在肺炎與泌尿道感染呈現顯著較低的申報費用，在慢性阻塞性肺病急性發作、腸胃道出血則無費用差異。在出院後指標上，泌尿道感染病患由醫院整合醫療主治醫師照護的出院後30日死亡率降低，其他三疾病則無統計差異。 結論: 台灣醫院整合醫療主治醫師的背景以內科醫師為主，但於2010至2014年並未呈現上升趨勢。醫院整合醫療主治醫師照護肺炎與泌尿道感染住院病患，呈現住院或短期死亡率下降並且申報費用降低之結果。然而住院天數並無顯著縮短，反而在腸胃道出血病患呈現延後出院現象。本研究顯示醫院整合醫療主治醫師在台灣的全民健康保險及給付制度上有潛力改善住院醫療品質並降低成本，但降低住院天數的效率不如預期。

Introduction: The emergence of hospitalist has dramatically changed the face of hospital care in the US, and has distributed to Taiwan since early 2000. This study aims to investigate the background of hospitalist care in Taiwan, including demographics and characteristics of hospitalists who cared for hospitalized patients in Taiwan from 2010 to 2014, and also the clinical performances of hospitalists in caring for hospitalized general medicine patients. Methods: This study is conducted with a retrospective design using National Health Insurance (NHI) whole-population claims data from the Collaboration Center of Health Information Application(CCHIA), Ministry of Health and Welfare in Taiwan. We defined that a physician who had at least 100 evaluation-and-management codes in a given year and carried inpatient claim amount for ≧80% of their total inpatient and outpatient claims in a given year, except for practicing as an intensivist, was a hospitalist. Trend and characteristics of hospitalists from 2010 to 2014 were analyzed. The main care outcomes included hospital mortality , length of stay, hospitalization cost and post-discharge 30-day mortality. Outcomes of hospitalized patients with pneumonia, urinary tract infection (UTI), chronic obstructive pulmonary disease (COPD) and upper gastrointestinal bleeding (UGIB), cared by hospitalists were compared to the non-hospitalist group. To deal with repeated admission and clustering of patients under physicians, generalized estimating equations (GEE) was constructed on which in-hospital outcomes and post-discharge outcomes were used as the endpoint variables. Statistical significance was set at a two-sided P<0.05 Results: A total of 13207 internal medicine, family medicine and neurology specialists were identified. The distribution showed that the number and percentage of specialists decreased gradually from 0-9% to 90-100%. the number of hospitalists declined from 108 in 2010, to 84 in 2014. Most hospitalists were male (more than 70%). The average age of hospitalists were 40-45. Over 80% of hospitalists were internal medicine specialists. Hospitalists tended to practice in metropolitan areas of Taiwan, and a trend toward affiliation to medical centers. After controlling for patient-level and institutional factors, multivariate regression analysis showed that hospitalized patients with pneumonia handled by hospitalists had lower hospital mortality, while patients with UTI, COPD and UGIB had no statistical difference in hospital mortality compared to those handled by non-hospitalist. Adjusted LOS of patients with pneumonia, UTI and COPD were similar, but LOS of patients with UGIB was longer when cared by hospitalists compared that of non-hospitalists. Adjusted hospitalization costs were lower for hospitalists in pneumonia and UTI, but insignificant difference in COPD and UGIB. Post-discharge 30-day morality was lower for UTI patients handled by hospitalists, while no significant difference in the other three diseases. Conclusion: The background and practice pattern of hospitalists in Taiwan were similar with the US, but the growth trend was not evident from 2010 to 2014. The performance of hospitalists showed favorable survival benefit and cost saving in caring for pneumonia and UTI patients. However, LOS of hospitalists care was not uniformly shortened. These findings suggested that hospitalist care in Taiwan has potential to improve survival and save cost, but the efficiency in throughput warrant further study.