緊急救護院外時間與 Covid-19 大流行的時間延滯分析

Abstract

在新冠肺炎疫情的影響下，許多國家的醫療資源都面臨能量不足的危機。先前的研究顯示，在疫情爆發期間，院外救護時間有被拉長。然而，至今仍沒有探討疫情爆發對緊急救護衝擊之後續效應的研究。因此，本研究針對院外時間受疫情尖峰期間的影響來分析，疫災是否對到院前時間有領先或滯後的影響。 本研究的目的是透過新冠肺炎之確診病例數與院外時間之時間序列，以及相關的防疫政策規定，探討救護系統的改變與恢復和疫情爆發與冷卻之間，是否存在「領先效應」與「滯後效應」。院外時間包括反應時間、現場時間以及送醫時間，分別是接獲報案電話至救護車抵達現場的時間間隔、救護車在救護現場停等時間，以及救護車離開現場至抵達醫院的時間間隔。 本研究使用兩個方法，找出院外時間受疫情影響的時間點。首先，使用多週期時間序列分解法（MSTL）從五年的每日院外救護時間中，去除與疫情無關的季節性波動訊號。接著使用貝氏變點檢測法（Bayesian Change Point Detection）從剩下的時間訊號中，找出變化點。本研究認為，去除與疫情無關的季節性波動訊號，可以讓我們更清楚地獲得變數受到疫災影響的日期。 受限於無法找到新冠肺炎之每日確診病例數的分布，本研究Covid-19的三個高峰期是由PPE著裝政策和2020年7月27日疫情警戒降級定義，分別從2020年2月13日至2020年6月1日、2020年12月22日至2021年3月19日和從2021年5月14日至2021年7月27日。優點是可以取得更多與整個台灣的Covid-19疫災相關的信息。缺點是使用政府推行的政策，可能也會與疫情真正的爆發與冷卻日期有時間差，而且不是每一波疫情都會馬上影響到台北市的緊急救護。 本研究以台北市地區2017至2020年的資料進行案例研究。院前時間受變化的日期與每個高峰期的爆發與冷卻之間存在時間差，包括9個領先效應和11個滯後效應。大部分受到改變的到院前時間，在疫情爆發時呈現滯後效應，而在疫情冷卻時呈現領先效應。然而，在第一次和第二次疫情的爆發期間，分別因為疫情前期的提早準備工作與訓練，以及冬天增加的案件數量，使一部分的案件的到院前時間被提早被改變。另外，第二次疫情冷卻時，反應時間的恢復有11天的滯後效應，隱含著救護人員對未知疫災的心理恐懼。 從數據分析中還有以下發現：第一，我們根據防護規定跟降級政策，反應時間、院前總時間跟其他非尖峰時段比起來，數值有顯著升高。第二，有送醫案件的現場時間在2020年1月之後，穩定的提高約3至4分鐘。第三，第二波疫情在台北市在疫情爆發的邊緣。可以觀察到，即使台北市確診數量很少，救護的防疫政策還是會影響到反應時間。第四，因為台北市是第三波疫情爆發的主要起源地，所以爆發的時間點對於到院前時間的影響相對劇烈，到院前時間也因此在較短時間內受到改變，特別是反應時間和送醫時間。推測到院前時間在第三波疫情之下的改變，可能與台北市救護需求上升相關。 本研究致力於使政府和EMS系統可以更清楚知道Covid-19疫情爆發對緊急救護衝擊之後續效應，並且使他們可以從到院前時間受到疫情以及其他因素影響的分析，更好的調整政策。

Under the impact of the Covid-19, medical resources are exhausted. Previous research has shown that during the outbreak of Covid-19, not only the number of cases received by the EMS system were affected but also the EMS response time were elongated. However, there is currently no literature studied on whether there is a lead or lag effects of Covid-19 on EMS prehospital time. The purpose of this study is to observe the possible “lead effect” or “lag effect” of the EMS system and Covid-19 through the time series of the prehospital time of EMS and the peak periods of Covid-19. Prehospital time includes response time, scene time, and transport time, which are the time interval between receiving a report call and the arrival of the ambulance at the scene, the time the ambulance stopped at the scene, and the time interval between the ambulance leaving the scene and arriving at the emergency department (ED) of hospital. This study used two methods to find out the time points at which prehospital time of EMS was affected by Covid-19. First, seasonal fluctuations unrelated to Covid-19 are removed using 5-year daily time series of each prehospital time by Multiple Seasonal Trend Decomposition using Loess (MSTL). After that, Bayesian Change Point Detection (BCPD) method is adopted to detect change points from the remaining time signal in prehospital times. Limited by the inability to find the distribution of daily confirmed cases of Covid-19, the three peak periods of Covid-19 in this study were defined by the PPE dressing policy and the downgrade of Epidemic Alert Level on July 27, 2020. These peak periods are from February 13 to June 1 in 2020, from December 22 in 2020 to March 19 in 2021, and from May 14 to July 27 in 2021. The advantage is that more information about the Covid-19 epidemic in entire Taiwan can be obtained. In contrast, the drawback is that the use of policies implemented by the government may also have a time lag with the actual outbreak and cooling date of the epidemic, and not every peak periods will immediately affect EMS in Taipei City. Taipei City’s EMS of OHCA patients from 2017 to 2021 serves as the case study in this work. There are time lags between prehospital times being changed and time points of each peak period, including 9 lead effects and 11 lag effects. Lag effect can be mostly found between the outbreaks and prehospital time changes, while lead effect can be mostly found between the cooldowns and prehospital time recoveries. However, during the first and second outbreaks, lead effects of Covid-19 outbreak on prehospital times can be found due to the early preparations and training, as well as the increase in the number of cases in winter. In addition, the 11-day lag effect of the second cooldown was found on response time, which implied the psychological pressure of the ambulance personnel for the unknown epidemic. The following findings were also found from the data analysis. First, response time and total prehospital time have significantly increased compared with other non-peak periods. Second, scene time transported to ED has steadily increased by about 3 minutes after January 2020, while scene time without transported to ED did not change significantly because of time waiting for the police's arrival is almost the same. Third, there are no confirmed Covid-19 cases in Taipei City during the second peak period, however, response time is still affected. Fourth, because Taipei City is at the center of the third outbreak, so the time of the outbreak has a relatively severe impact and quick changes on prehospital times, especially response time and transport time. It is speculated that quick change of prehospital time may be related to the instantaneous increase in the demand for EMS in Taipei City during the third peak period. The results potentially provide to the government and EMS organizers a better understanding the aftermath of the Covid-19 outbreak on ambulance dispatch and the resilience of EMS system and allowing them to better adjust policy of EMS in response to potential indirect and direct factor of Covid-19 which may affect the prehospital time.