目的：探討CT/MRI利用之適當性及其影響因素。 方法：隨機抽取健保局南區分局94年4月至6月間重複執行CT及MRI二次（含）以上的腦阻塞或顱内傷案件共359件，由專業審查醫師評估執行CT/MR的臨床適當性。以SPSS12.0套裝軟體進行卡方、t檢定及羅吉斯迴歸分析。 結果：CT/MRI首次執行適當的個案佔96.94％，然而再次執行則下降為52.65％，再次執行不適當的主要原因以『病情不適合』為最多。就疾病而言，腦阻塞再次執行不適當的原因以『前次執行已適當』最多；顱内傷則以『病情不適合』為最多。醫院特性權屬別、層級別、病床數及市場佔有率分別於首次CT/MRI執行之適當性有統計上顯著關聯，而與再次執行之適當性有統計上顯著關聯的是市場佔有率與個案年齡。在控制醫院層級別、病床數、市場佔有率及個案年齡下，權屬別與首次執行適當性有統計上顯著關聯，公立醫院適當之勝算為財團法人醫院之0.143倍(p<0.05)。再次執行再納入專業審查變數被控制下，跨院與再次執行CT/MRI之適當性有統計上顯著關聯，跨院適當之勝算為非跨院適當之0.031倍(p<0.01)。 結論：CT/MRI跨院再次執行有不適當的重複利用情形，健保相關單位應加強稽核跨院重複使用的案件，藉此減少影像醫療費用支出。

Objectives: To assess the appropriateness of computed tomography and magnetic resonance imaging and explore factors affecting the appropriateness of imaging. Methods: The numbers of patients of repeated imaging with occlusion/stenosis of cerebral arteries 193 cases and intracranial injury 166 cases encountering to the contract hospitals of South BNHI during April to June 2005 were acquired by random sampling. The appropriateness of initial and subsequent imaging during the period of three-month was audited by neurologists and neurosurgeons through retrospective charts review. Results: Of 359 examinations, 96.94 percent and 52.65 percent were performed with the true need for initial and subsequent imaging respectively. The main reason of inappropriate imaging for all subsequent examination was "no clinical indication", while "positive prior test" and "no clinical indication" were with respect to occlusion/stenosis of cerebral arteries artery and intracranial injury. The characteristics of hospital were significantly associated with the appropriateness of initial imaging; otherwise the market share and patient's age were significantly associated with the appropriateness of subsequent imaging. Multivariate logistic regression showed that hospital ownership was a factor affecting the appropriateness of initial imaging examination (public vs. nonprofit, odds ratio= 0.143, p 0.05), the variable "across hospital" was significantly associated with subsequent appropriateness of imaging examination (across vs. non-across, odds ratio=0.031, p 0.01). Conclusions: We observed the inappropriateness of repeated imaging examination performed across hospitals. Therefore, decreased the inappropriate use of diagnostic imaging should be a priority in efforts of Bureau of National Health Insurance that focus on controlling imaging expenditures.