- 學位論文
SARS-CoV-2 Omicron BA.2 變異株群體免疫：數理模式研究
Herd Immunity to SARS-CoV-2 Omicron BA.2: A Modeling Study
摘要
背景與研究目標：在短期內無法清除COVID-19疫情的情況下，是否能透過全民大規模接種COVID-19疫苗產生群體免疫，以達成與病毒共存的目標（維持社會經濟正常運作，且染疫後死亡率不高於流行性感冒（以下簡稱流感））是當前COVID-19防疫政策上最關鍵的問題。然而，針對此議題，目前尚缺乏具有實證數據佐證的數理模式研究。本研究擬依據臺灣2022年Omicron BA.2變異株疫情流行數據，建立COVID-19傳播數理模式，以分析是否可透過大規模COVID-19疫苗接種達成群體免疫，同時探討將COVID-19重症死亡率降低到接近流感水準的可行性。 方法：本研究建立考慮中重症率及死亡率之SARS-CoV-2 Omicron BA.2變異株SEIRS傳播動態模型。Omicron BA.2傳播動態流行病學參數及疫苗保護力參數皆依據系統性文獻回顧實證數據。本研究分為兩部分：第一部分為根據臺灣2022年疫苗接種數，以數理模式擬合2022/3/10 至 2022/6/30 指揮中心公布每日新增通報確診人數、中重症人數及死亡人數，估計臺灣民眾口罩配戴率（與他人接觸時全程配戴口罩）隨時間變化趨勢、染疫後中重症率及中重症死亡率，以此為基礎，估計大規模疫苗接種對預防感染、中重症及死亡的效果，再進一步推估 2022/7/1至2022/12/31新確診及Rt值（effective reproductive number）變化趨勢，探討是否能夠將Rt值壓低到群體免疫閾值以下（Rt值<1且持續降低到趨近於0）。敏感度分析為以模擬不同比例（25%、50%、75%）確診者通報及接受隔離對民眾口罩配戴率、疫苗預防效果及中重症率的影響及不同染疫後保護期平均持續時間（90天、180天、314天）對Rt值及未來疫情的影響，計算其敏感度範圍（sensitivity range）。第二部分為以年齡結構模型（Age-structure model）（分成：≤ 19 歲、20~59 歲、≥60 歲）探討在理想疫苗接種率（第一劑：99%、第二劑：97.5%、追加劑：95%）情況下將COVID-19重症死亡率降低到接近流感水準的可行性，敏感度分析考慮不同防疫策略（配戴口罩/確診者通報及接受隔離）及具免疫逃脫特性的新變異株（傳播力為Omicron BA.2之1.34倍）流行等各種情境。 結果：擬合後本模型預測2022/3/10~6/30間通報確診人數、中重症人數及死亡人數均非常接近指揮中心公布之數據。透過模型擬合，可估計得到此波疫情本土流行病學參數如下：（1）民眾口罩配戴率先由75% (65%~80%) (2022/3/10) 逐步下降到55% (51%~60%) (2022/4/26) 再逐步提升至85% (83%~87%) (2022/6/30) ；（2）染疫後中重症率為2.3% (1.55%~2.75%)；（3）中重症死亡率為33%。大規模疫苗接種在族群層次對預防感染、中重症、死亡的效力分別為55.6% (36.2% ~ 62.2%)、91.4% (87.4% ~ 92.7%)、92.0% (88.2% ~ 93.3%)。若口罩配戴率為0%，僅憑大規模疫苗接種無法達到群體免疫，到2022/6/30前累積確診數將達9,325,123（6,626,819 ~ 10,801,246）、中重症數將達48,088 (33,093 ~ 56,877)、死亡數將達18,936 (13,012 ~ 22,408)，Rt值雖能短暫降到0.1 (2022/4/19 ~ 2022/4/25)，但隨著染疫後免疫力快速衰退，Rt值在2022/6/30會再上升到1.2 (1.1 ~ 1.3)，到2022年底仍在1.0，降不下來，而陷入持續流行（endemic）狀態。2022/7/1~12/31的疫情發展取決於口罩配戴率。若民眾口罩配戴率維持在85%，可望疫情能夠逐步趨緩到年底（2022/12/31新增通報確診數可降到358）；若民眾口罩配戴率在7/1後逐步放鬆至55%（2022/4/26水準），則疫情預期將在9月反彈，於2022/9/30達到高峰（2022/9/30新增通報確診數60,712）。敏感度分析顯示染疫後保護期平均持續時間對Rt值及未來疫情趨勢影響很小，在染疫後平均保護期為314天的最佳情況下，若全民皆不配戴口罩，僅憑大規模疫苗接種加上疫情爆發後的自然感染， Rt值會從0.9（2022/7/1）上升至1.4（2022/9/1~9/14）再逐步下降至0.9（2022/12/31），Rt值均在群體免疫閾值（1.0）上下波動，疫情均會持續流行，無法穩定地讓Rt值下降到趨近於0而達成群體免疫。即使在理想疫苗接種率的情況下，僅憑疫苗接種，在沒有新變異株出現的情況下，2023年COVID-19重症及死亡人數將達30,699人為2018年流感併發重症1,196人的25倍。必須加上口罩配戴率85%及50%確診者通報及接受隔離才能將COVID-19重症及死亡人數控制在接近流感的水準。然而若具有免疫脫逃能力的新變異株進入國內造成流行，僅憑疫苗接種，2023年COVID-19重症及死亡人數將高達36,142人（新變異株免疫逃脫能力0%）；113,901人（新變異株免疫逃脫能力50%）；235,069人（新變異株免疫逃脫能力100%）為2019年（Influenza vaccine-mismatched）流感併發重症2,325人的15~100倍。即使加上口罩配戴率85%及50%確診者通報及接受隔離，在面對新變異株的情況下一樣無法將COVID-19重症及死亡人數控制在接近流感的水準，2023年COVID-19重症及死亡人數將達3,144人（新變異株免疫逃脫能力50%）；41,106人（新變異株免疫逃脫能力100%）。 結論：本研究顯示僅憑全民大規模接種COVID-19疫苗無法將Rt值壓低到1以下，即使加上大規模疫情爆發導致的大量自然感染，Rt值仍會在1.0上下波動而使疫情陷入持續流行狀態，無法達成群體免疫而成功與病毒共存，且台灣社會還要付出大量中重症及死亡數的重大代價。若要將染疫後重症死亡率降低到接近流感的水準，全民必須嚴格遵守防疫規範，包括配戴口罩以及確診者通報及接受隔離以避免傳染他人，若防疫強度放鬆則必然導致週期性COVID-19疫情爆發及死亡潮，因此持續地採取高強度口罩配戴率是必要的。但即使如此，未來若無法成功研發及全面接種保護力更強的次世代疫苗，面對具更加強大免疫逃脫能力的新變異株從國外傳入國內造成社區傳播，仍無法避免爆發嚴重疫情，面對比流感高十到百倍的重症及死亡人數。
並列摘要
Background: Whether it is possible to achieve herd immunity through mass COVID-19 vaccination is presently the most critical issue of COVID-19 prevention and control. Waning of immunity and new variants breakthrough infection raise questions on the feasibility for achieve herd immunity to COVID-19, similar to herd immunity to measles and influenza. However, empirical data-validated modeling studies are lacking. Thus, this study aimed to evaluate whether it is theoretically feasible for mass COVID-19 vaccination to decrease Rt value of SARS-CoV-2 to herd immunity threshold through a mathematical modeling. Methods: We established a SEIRS model of SARS-CoV-2 Omicron BA.2 transmission. The parameters of transmission of Omicron BA.2 and COVID-19 vaccines were based on the empirical studies. Our study was divided into two parts. The first part was about estimating the proportion of mask-wearing (Masking), the proportion of moderate/severe cases, and mortality among moderate/severe cases by fitting the model to outbreak data from March 10th, 2022 to June 30th, 2022 in Taiwan. In addition, we assessed the impact of different interventions on outbreak control in terms of the trajectory of the epidemic from July 1st, 2022 to December 31th, 2022, and the effective reproductive number (Rt). We conducted sensitivity analyses to assess the impact of different proportion of notification isolation on the proportion of mask-wearing, the proportion of moderate/severe cases, and population-level intervention effect, and different duration of natural immunity on Rt and epidemic trajectory. In the second part, we used age-structure model to assess the feasibility of reducing the mortality rate of COVID-19 to near influenza under ideal vaccination coverage (First dose : 99%; Second dose: 97.5%; Booster dose: 95%). The sensitivity analysis considered various scenarios with different interventions (mask-wearing/ notification isolation) and the emergence of new variant with the stronger ability to evade immunity from COVID-19 vaccines. Results: The epidemiological parameters of this outbreak in Taiwan, which could be estimated through model fitting were as follows: (1) the proportion of mask-wearing gradually decreased from 75% (65%~80%) (March 10th, 2022) to 55% (51%~60%) (April 26th, 2022) and then increased sharply to 85% (83%~87%) (June 30th, 2022); (2) the proportion of moderate/severe cases was 2.3% (1.55%~2.75%); (3) mortality among moderate/severe cases was 33%. Population-level intervention effectiveness of COVID-19 vaccines against infection, moderate/severe disease, and death were 55.6% (36.2% ~ 62.2%), 91.4% (87.4%~92.7%), and 92.0% (88.2%~93.3%), respectively. Vaccination alone with large numbers of people continued to be infected, Rt could temporarily decreased to 0.1 (2022/4/19 ~ 2022/4/25), however, due to the waning of natural immunity, Rt raised to 1.2 (1.1 ~ 1.3) on June 30th, 2022, and remained 1.0 by the end of 2022. The trajectory of the epidemic from July 1st, 2022 to December 31th, 2022 depended on the proportion of mask-wearing. If masking could maintain at 85%, the number of new confirmed cases would decrease greatly to 358 on December 31th, 2022. In contrast, if masking gradually relaxed to 55%, the epidemic would rebound in September and the number of new confirmed cases would reach 60,712 on September 30th, 2022. Sensitivity analysis showed that the duration of natural immunity had little impact on Rt. Even under the most optimal scenario (duration of natural immunity was 314 days and vaccination coverage continued to increase), vaccination alone would not suppress Rt to less than herd immunity threshold (1.0). Ideal vaccination alone and no new variants appears in 2023, the number of severe cases and deaths from COVID-19 in 2023 will reach 30,699, which is 25 times higher than that from influenza (1,196) in 2018. Public health interventions, including mask-wearing (85%) and notification isolation (50%), must be kept indefinitely to let severe cases and deaths from COVID-19 close to that from influenza. However, if a new variant with the stronger ability to evade immunity from COVID-19 vaccines appears in 2023, the number of severe cases and deaths from COVID-19 in 2023 will reach 36,142 (immune escape: 0%), 113,901 (immune escape: 50%), and 235,069 (immune escape: 100%), which is 15 to 100 times higher than that from influenza (2,325) in 2019. Even with mask-wearing (85%) and notification isolation (50%), the number of severe cases and deaths from COVID-19 will reach 3,144 (immune escape: 50%), and 41,106 (immune escape: 100%). Conclusions: Our study shows that herd immunity to COVID-19 can not be achieved, even after increasing vaccination coverage, Rt will not suppress to less than 1. Public health interventions, including mask-wearing, and notification isolation must be kept indefinitely to prevent the massive increase in COVID-19 severe cases and deaths. If epidemic preventions begins to relax, the new outbreak of COVID-19 seems inevitably, thus, it is necessary to take high-level of epidemic preventions, but even so, if there emerges new variant with the stronger ability to evade immunity from COVID-19 vaccines, we will experience massive increase in severe cases and deaths from COVID-19, which is 10 to 100 times higher than that from influenza.