森林遊樂區風險管理研究：溪頭自然教育中心意外事件分析

Abstract

為了減少戶外遊憩風險並維護遊客的安全，管理單位可以從先前意外事件案例中學習，了解事件統計記錄、事故頻率或死亡人數等描述性統計數據，並進一步了解人為因素以及環境因素對於事件發生的影響。本研究以溪頭自然教育園區為研究地點，分析了人為因素和環境因素等潛在影響因素對所遊客意外事件的影響，可作為森林遊樂區風險管理規劃之用。 本論文採用回顧性分析，以了解2012年7月至2018年6月台灣溪頭自然教育中心（XNEA）發生的疾病和傷害事件。 276例患者的傷病病例的流行病學分析結果， 56.2% 為女性，平均發病率呈下降趨勢，約為0.024 例 / 1,000人-日，即每 100,000 名遊客約 3 例病例。大部分病例發現於大學池、巨型觀景台、溪頭自然教育中心入口（二號入口）等著名旅遊景點。超過一半的患者 (55.4%) 有固有的健康狀況。最普遍的症狀是頭暈（47.8%）。大多數的病例發生在早晨和夏季。患者的平均年齡為 52 歲（範圍 2-93歲）。老年遊客（≥60 歲）最容易出現健康問題。通過卡方檢驗，我們發現發生暈厥的男性多於女性。患有高血壓的遊客出現頭暈、呼吸問題和暈厥的風險更大。糖尿病患者容易虛弱。有心臟相關問題的人容易出現胸痛或胸悶。 除了對疾病病例的研究外，本論文還對傷害案例進行了研究，發現2012年7月至2018年6月期間記錄了957起傷害案例。超過一半的傷者（59.6%）為女性。受傷訪客的平均年齡約為 49 歲（範圍 1-92 歲）。他們中的大多數 (71.5%) 因滑倒、絆倒或從高處墜落而受傷。夏季月份和下午時間是受傷案件最頻繁的時期。分析的傷害率約為每年 0.090 例 / 1,000 人-日，或每 100,000 名訪客約 9 例傷害案例。從熱點圖來看，大部分的受傷案例都被發現在一些熱門的旅遊景點，如遊客中心、大學池和溪頭的入口處。與其他景點相比，這些景點海拔較低且相當平坦。天氣變項和遊客傷害風險的地形資訊已納入Poisson和負二項式回歸模型中。 Poisson 和負二項式回歸模型都證明了類似的發現，即每日遊客傷害率與關於平均氣溫、相對濕度和降雨條件的天氣共變量呈正相關。模型選擇標準和適配度表明，負二項式模型比Poisson回歸模型提供了更好的模擬。受傷率風險與平均氣溫、相對濕度和降雨條件的增加有關，發病率比 (IRR) 分別為 1.019、1.020 和 1.438。關於傷害率的地形變量與 Poisson 和負二項式回歸的關係表明，平均海拔和傷害發生地點的平均坡度對傷害率在統計上不顯著。 本論文的結果證實，過往病史對戶外參與者構成了固有風險，而這種在戶外休閒期間病發的關係證實了特定患者群體的風險更大。在森林中遭受的傷害與天氣因素有關，尤其是下雨。這些研究結果都為未來的事故預防計劃以及旅行者提供了具體的證據和資訊以正確地認識事故風險。

To develop a proactive manner for visitor risk management, it is imperative for the outdoors organization to learn prior incident cases not only the descriptive statistics such as statistical records of incidents, frequency of accidents, or deaths but also the correlation analysis. Different from conventional incident reports, this dissertation analyzes the effect of potential contributing factors regarding human factors and environmental factors to the incident experienced. This dissertation applied the retrospective analysis in order to provides the comprehension of incident cases both illness and injury encountered in Xitou Nature Education Area (XNEA) of Taiwan between July 2012 and June 2018. Epidemiologic analysis results of illness cases indicated of the 276 patients assessed, 56.2% were female with a decreasing trend of the average rate of about 0.024 per 1,000 person–days, or around 3 illness cases per 100,000 visitors. Most of the illness cases encountered were found to some of the famous tourist spots, such as the University Pond, Giant Scenic Platform and XNEA’s entrances (Second Pass). More than half of ill participants (55.4%) had preexisting health conditions. The most general symptom was dizziness (47.8%). Most health-related cases took place in the morning and during summer period. The average age of ill participants was 52 years (range 2–93). Elderly visitors (≥60 years) were the most susceptible to illness. By using chi-square test, we found that more men than women encountered syncope. Participants with high blood pressure had a greater risk of experiencing dizziness, respiratory problems, and syncope. People with diabetes were susceptible to weakness. People with cardiac-related problem were vulnerable to chest pain or tightness in their chest. Apart from studying in illness cases, this dissertation also study injury cases and found that 957 injury cases were documented between July 2012 and June 2018. Over the half of injury patients (59.6%) were female. The average age of injured visitors was about 49 years (range 1−92 y). Most of them (71.5%) experienced an injury because of a slip, a trip, or a fall from a height. Summer months and afternoon time were the most frequent periods of injury cases. Injury rate analyzed was about 0.090 per 1,000 person−day a year of injury cases, or around 9 injury cases per 100,000 visitors. From the hot spot map, most of the injury cases were discovered to some of the popular tourist spots, such as the visitor center, University Pond, and XNEA’s entrances. These spots are at lower altitudes and quite flat compared to other tourist spots. Weather variables and topographical information on visitor injury risks have been fitted in Poisson and negative binomial regression models. Both Poisson and negative binomial regression models demonstrated a similar finding that daily visitor injury rate was positively related to weather covariates regarding average air temperature, relative humidity, and rain condition. Model selection criteria and goodness−of−fit revealed that the negative binomial model provides the better fit than the Poisson regression model. Risk of injury rate was associated to an increase in average air temperature, relative humidity, and rain condition with the incidence rate ratio (IRR) at 1.019, 1.020, and 1.438, respectively. The relation of topographic variables on injury rate fitting in Poisson and negative binomial regression indicated that average elevation, and average slope of injury locations were statistically insignificant for injury rate. The results of this dissertation confirmed that preexisting medical conditions establish an inherent risk in outdoor participants, and this relationship with the development of illness during outdoor recreation confirmed greater risks in specific patient groups. Injury experienced in forest was associated with weather factors particularly the occurrence of rain. All of these provides concrete evidence and information for future incident prevention programs as well as for the right perception of incident risk for future travelers.