北市到院前緊急醫療系統最佳路徑之探討

Abstract

緊急醫療的最大價值就是在意外發生後，能迅速給予各種可能的急救處理，因此時間的爭取是非常重要的。Cumnius R.O.發現，每延遲一分鐘開始急救，病人急救成功的機率便減少7-10%。因此在院前的救護以及救護車的運送時間對病人的生命安全有絕對的重要性。目前救護車輛前往事故地點路徑選擇，是靠駕駛人員的人為經驗判斷，若是對前往搶救的路徑上的交通狀況、道路設施而延遲搶救時間，或在送到醫院後才發現病床不足、醫護設備不足的現象，導致時間的浪費。如何減少上述狀況而造成時間延遲，是緊急救護上一大課題。空間分布的查詢與展示，是最容易懂及最實用的資料建構方式。因此希望能運用地理資訊系統呈現所規劃之最適路徑的緊急救護模型，提供給醫院或消防分隊派遣救護車時的決策考量。 本研究於資料庫中加入醫院的緊急醫療資源，並考慮交通即時狀況，避開擁擠和阻塞路段，規劃消防分隊前往事故地點以及從事故地點到最適當醫院的最佳路徑選擇模式。並且隨機抽取個案來做使用程式前後的模擬(傳統地圖上的空間直線最短路徑與使用本研究程式所運算的最短路徑)，並用SPSS軟體做推論統計比較兩者的優劣。 將收集到的台北市消防分隊救護資料以Excel的檔案格式加以建檔整理，經過核對原始資料無誤後，隨機抽取真實案例資料；上午尖峰(7:00-9:00)時間64筆資料，下午尖峰(17:00-20:00)時間138筆資料，共計202筆資料來進行程式的模擬。而本研究不考慮使用程式前後在離峰時間所產生的差異。模擬結果利用paired t test來作推論性統計分析。 結論發現，paired t test所進行的推論性統計分析達到顯著差異，可見，本研究程式在尖峰時間對於縮短在院前的救護以及救護車的運送時間有很大的幫助。

When people are caused to injury in an accident, one of the most important things of all is to move them to a hospital in time. It is obvious that time is crucial to rescue persons with serious injury. So far, the transportation is often decided by human beings. Sometimes, according to not familiar with the traffic condition or hospital facilities, such as whether enough beds or being able to offer such supports or not, we often lose the golden hours on emergency rescue. This thesis presents an adaptable model to find a proper path for Emergency Medical evacuation. Geographical Information Systems (GIS) are involved in our design to provide a friendly interface to illustrate the path for evacuation. A system integrated with GIS to offer EMS organization to plan a path from stations to accident places and a path from accident places to hospitals by taking account of the condition of traffic, and avoiding the crowed roads. We consider the time and EMS resource of hospitals as parameters of the program, and put them into database of the system. We also select some cases randomly for simulating in order to evaluate the program, and use SPSS for inferential statistics. The data which were gathered from stations in Taipei were made up by Excel. We selected 202 cases in peak hour, 64 cases in 7:00-9:00, the others in 17:00-20:00. The results of our simulation are compared by using inferential statistics. A significant improvement can be archived if an adaptable path of rescue can be obtained prior to a fluctuated event happened. After comparing by using inferential statistics such as paired t test, we can conclude that it is great helpful to save the time from stations to accident places and from accident places to hospitals.