緊急醫療資料庫之時間序列分析

「緊急醫療服務」(Emergency Medical Services, EMS)提供重大傷、病患由事故現場、直到到達醫院急診部門前的緊急救護服務。到院前實施的緊急處置能不僅能夠減少病患的死亡及失能，同時能降低其復健與急救的醫療成本。為了提供有效的緊急醫療服務，在有限的資源內如何適度的配置是一個重要的議題。像某些類型的緊急醫療事件的發生具有週期性的發生高峰是可以預見的，而醫療資源也應對這些週期的配置。舉例來說，溺水與中暑的發生在夏季時達到發生率的高點，而中風則相反，好發於冬季。本研究最大的目的是對緊急醫療資料庫中共32個求救原因進行全面性的分析，希望找出緊急醫療事件中顯著的週期性模式。研究方法上，採取了一個新穎的分析方法，傅立葉─高斯分解法，能對於輸入的時間序列進行週期性分析，並產生出高闡述性的圖像結果，並且使用傳統的統計分析來提供使用者可靠的統計指標。在研究結果顯示，週期性模式是廣泛的出現在不同類型的緊急醫療事件中。

關鍵字

時間序列分析；緊急醫療服務；醫療資料庫；週期模式

並列摘要

Emergency Medical Services (EMS) provide medical cares to seriously injured or illed patients, while they are being transported from the incident sites to the hospitals. Out-of-hospital acute medical interventions aims not only to reduce the mortality rate and the degree of disability of patients but also to minimize patients’ rehabilitation and medical costs. One critical issue in providing effective EMS is the quantities of resources to be allocated. It is conceivable that some types of EMS incidents should have periodical patterns and the public health agents should allocate the EMS resources accordingly. For example, drowning and hyperthermia incidents should peak during summer seasons, while stoke incidents should peak during winter seasons. The primary objective of the study presented in this thesis is to conduct comprehensive time series analyses based on the records in an EMS database in order to identify significant periodical patterns of EMS incidents. In this study, a novel analysis method designed to provide the user with a highly interpretable picture of the periodical patterns in the input time series has been employed. The analyses were then followed by carrying out the conventional statistical tests to provide the user with solid statistical metrics. The results derived from this study reveal that periodical patterns are commonly present in many types of EMS incidents.

並列關鍵字

Time series analysis ； Emergence medical services ； Medical database ； Periodical patterns

參考文獻

[4] 陳品良, “一個新的時間序列分析法及其在大型醫學資料庫的應用,” 台灣大學, 台北市, 2014.

[6] Robert B. Cleveland, William S. Cleveland, Jean E. McRae, and Irma Terpenning, “STL: A Seasonal-Trend Decomposition Procedure Based on Loess,” Journal of Official Statistics, pp. 3-73, 1 1990.

[1] Bray, J.E., Straney, L.D.J., Barger, B., Finn, J.C., "Effect of public awareness campaigns on calls to ambulance across Australia," Stroke, pp. 1377-80, 5 2015.

[2] Arntz H-R, Willich SN, Schreiber C, Bruggemann T, Stern R, Schultheiss H-P., “Diurnal, weekly and seasonal variation of sudden death. Population-based analysis of 24,061 consecutive cases.,” European Heart Journal, pp. 315-20, 2 2000.

[3] Brennan PJ, Greenberg G, Miall WE, Thompson SG, “Seasonal variation in arterial blood pressure.,” British Medical Journal, 10 1982.

國際替代計量

緊急醫療資料庫之時間序列分析

全文下載

主題瀏覽