楊志良2006-07-252018-06-292006-07-252018-06-292002-07-31http://ntur.lib.ntu.edu.tw//handle/246246/5022全民健康保險實施多年來，醫療院所對於醫療審查執行是否適當，仍有相當的質疑與 不滿，且與健保局間有非常大的爭議，而國內目前尚無許多臨床醫療參考基準，能參考之 文獻亦十分有限，因而突顯「入院適當性」研究之重要性。 本研究分別以信度、效度在歐美均被肯定之Appropriateness Evaluation Protocol (AEP)及 Intensity Severity Discharge(ISD)為發展之基礎，經國內修訂完成之Modified AEP 及Severity of illness/Intensity of service (SI/IS )為入院審查工具，以盲技術(Blind Technique)分派病歷給具 醫、護背景且經訓練之審查者(reviewers)，於樣本醫院進行門、急診入院適當性審查，並由 臨床專科醫師確認工具之效度，再經專家座談討論及統計分析，有以下之發現： 審查者使用適當性判定工具之結果，四位審查者間mAEP 之一致性普遍較SI/IS 高；門診 及急診kappa 值約在0.4 與 0.7 之間，達到Landis 和koch 所訂再現性佳(good reproducible) 之水準，若經審查者進一步討論不同意見後再計，kappa 值可顯著提高；比較審查者與專科 醫師(效標參考值)之判定，二工具門診敏感度均達91%以上；而急診敏感度與特定度分別在 87%及93%以上，由於效標參考值訂定不易，若未來用於保險審查，此部份須再予努力。以 mAEP 審核病歷所花時間，平均值約在3 分鐘以內，少於國外研究的7-10 分鐘顯示mAEP 為 一審核迅速且簡易省時之工具；另審查者使用否決(overrides) mAEP 判定之比率在4%以下， 與美國的5%以下相近，顯示mAEP 有良好的完整性。不適當入院原因探討上，門診和急診均 以「任何診斷處置或（和）治療應可於門診完成」之比率最高，結果亦與歐美研究相符。在 探討入院適當性相關變項上，以專科醫師判定為依變項時，門診僅有「科別」一項顯著；而 急診有「性別」、「年齡」、「有無重大傷病」、「入院前有無重大檢查」為顯著變項；由 於此發現僅為本研究初步結果，其對適當性之解釋效力仍待後續研究予以證實。 本研究發現原版AEP 需予以修訂(modified)方適合國內使用，而結果顯示經修訂之mAEP 可用於國內入院適當性之判定，且隨著醫學科技進步及醫療環境的改變，mAEP 應適時予以 增修，以符合不同時空之需要。由於臨床基準判定工具之建立，將耗費十分可觀之人力、財 力、物力等成本，且非一朝一夕可完成，本研究期望mAEP 短期內能予以電子化，對全民健 保之專業審查將有相當助益，更期待藉本研究之? 磚引玉，國內後續有更多相關研究投入。Although National Health Insurance (NHI) has been implemented for many years, most of the hospitals are still unsatisfied with the medical review. At the same time, there are few clinical guidelines and references available in Taiwan. Therefore, the importance of the study on appropriateness of admission is self-evident. Appropriateness Evaluation Protocol (AEP) and Intensity Severity Discharge(ISD) are two most recommended admission review instruments which the reliability and validity were proved in United States and Europe countries. We adapted these two tools and the modified AEP(mAEP) and Severity of illness/Intensity of service (SI/IS) were used in this study. First of all, the physician and nurse reviewers who were well trained to review the emergency room (ER) and outpatient (OPD) medical records to judge admission appropriateness by blind technique. Then the gastrologists and general surgeons justified the validity of tools and clinical experts discussed the result. Finally, the data was analyzed by statistical methods. The findings are as follows: General speaking, the consistency of mAEP is better than SI/IS. In reliability test, the reviewers’ Cohen’s kappa are between 0.4 and 0.7 which met the guideline of good reproducible by Landis and Koch. If we recalculate the consistency after the reviewers revised their discordant decision, the kappa would be significantly higher. In validity test, the OPD’s sensitivity of mAEP and SI/IS are over 91%. The ER’s sensitivity and specificity are over 87% and 91% . Due to the difficulty of finding the gold standard, it has to be much careful if the mAEP was used for utilization review in the future. The mean time for admission review was about 3 minutes. To compare with western countries’ 7-10 minutes, our study took less time. This also suggested that mAEP was a fast, easy and time saving utilization review instrument. The override of mAEP in our study is 4%, which is closed to the 5% in US, and it showed that the completeness of mAEP is good. The number one reason of both outpatient and ER inappropriate admission are ”Any needed diagnostic procedures and/or treatment can be done on an outpatient basis”. It is as same as western countries’ studies. To analyze the related factors of admission appropriateness, when the specialists’ decision is the dependent variable, we found only the physician’s specialty was significant in OPD, but sex, age (demographic variables), major casualty, major examinations before admission (disease variables) were significant in ER. Due to these findings are only the preliminarily result, so it is required more studies to prove the power of explanation. We found the original AEP has to be modified in order to meet the native needs. Although the sensitivity and specificity of mAEP were proved in Taiwan, it has to be revised periodically due to the new medical technologies and changed environment. From the experience of western studies, we knew that invent a utilization review tool needs lots of manpower, budgets, materials and also time consuming, so it could not be developed in a short time. Therefore, we expect that the mAEP can be computerized for efficient medical review in short times. We also believe that our study is just a beginning, and we hope more researches will join in together in the near future.application/pdf231686 bytesapplication/pdfzh-TW國立臺灣大學公共衛生學院醫療機構管理研究所reporthttp://ntur.lib.ntu.edu.tw/bitstream/246246/5022/1/902416H002040.pdf