目標：經由分析急診病患特性與其門診利用情形，瞭解發生「可避免急診」病患之特性， 並探討可避免急診與初級照護品質的關係。方法：本研究以2010年至2013年發生「可避免急 診」之病患為主要研究對象，與同年未發生可避免急診之門診患者進行配對。使用「照護連續 性」作為測量初級照護品質的指標，分析「病患特性」、「照護連續性」、「門診主要就醫機 構層級」對於發生可避免急診的影響。結果：研究結果發現，男性、投保金額低的病患，發生 可避免急診的機率較高。多變量分析結果發現，在控制其他變項後，照護連續性越低的組別， 發生可避免急診機率越高（OR=1.924~3.220）。相較於診所層級，區域醫院（OR=3.543）、醫 學中心（OR=2.916）與地區醫院（OR=2.148）發生可避免急診機率較高。結論：本研究發現 門診的照護連續性高則發生可避免急診之機率較低，因此建議應提昇民眾的照護連續性。此外 建議可將「可避免急診」作為監控門診照護品質之指標。（台灣衛誌 2017；36(5)：473-486）

Objectives: This research aims to describe the characteristics and healthcare utilization of patients with avoidable emergency department (ED) visits and to explore the relationship between avoidable ED visits and the quality of primary care. Methods: Patients who had avoidable ED visit between 2010 and 2013 was selected as the case-group in this study and were matched to patients who had outpatient visits but without any avoidable ED visit in the same year. The continuity of care index (COCI) was used to measure the quality of primary care. Multivariate logistic regression analysis was applied to explore whether avoidable ED visits are associated with patients’ characteristics, the continuity of care, and the type of providers as their usual source of primary care. Results: The study found that patient who was male or with lower income class were associated with higher probability of having avoidable ED visits. The result of multivariate analysis further revealed that the odds of avoidable ED visits were higher for patients with lower continuity of care (OR = 1.924 to 3.22), and for whom received primary care from regional hospitals (OR = 3.543) and medical centers (OR = 2.916). Conclusions: Higher continuity of care was found to significantly lower the risk of avoidable ED visits. Health policy stakeholders are therefore encouraged to improve continuity of care. We also suggest that avoidable ED visits should be used to monitor the quality of primary care routinely. (Taiwan J Public Health. 2017;36(5):473-486)