Details of the datasets.

BackgroundPlasmodium falciparum cases are rising in China due to the imported malaria cases from African countries. The main goal of this study is to examine the impact of imported malaria cases in African countries on the rise of P. falciparum cases in China before and during the COVID-19 pandemic.MethodsA generalized regression model was used to investigate the association of time trends between imported malaria cases from 45 African countries and P. falciparum cases in 31 provinces of China from 2012 to 2018 before the COVID-19 pandemic and during the COVID-19 pandemic from October 2020 to May 2021. Based on the analysis, we proposed a statistical and deep learning hybrid approach to model the resurgence of malaria in China using monthly data of P. falciparum from 2004 to 2016. This study builds a hybrid model known as the ARIMA-GRU approach for modeling the P. falciparum cases in all provinces of China and the number of malaria deaths in China before and during the COVID-19 pandemic.ResultsThe analysis showed an emerging link between the rise of imported malaria cases from Africa and P. falciparum cases in many provinces of China. Many imported malaria cases from Africa were P. falciparum cases. The proposed deep learning model achieved a high prediction accuracy score on the testing dataset of 96%.ConclusionThe study provided an analysis of the reduction of P. falciparum cases and deaths caused by imported P. falciparum cases during the COVID-19 pandemic due to the control measures regarding the limitation of international travel in China. The Chinese government has to prepare the imported malaria control measures after the normalization of international travel, to prevent the resurgence of malaria disease in China.背景：受非洲国家输入性疟疾病例影响，中国境内的恶性疟原虫（Plasmodium falciparum）感染病例呈上升趋势。本研究的核心目标为探究2019冠状病毒病（COVID-19）大流行前后，非洲国家输入性疟疾病例对中国境内恶性疟原虫感染病例增长的影响。方法：本研究采用广义回归模型，分别针对2012-2018年（COVID-19大流行前）与2020年10月至2021年5月（COVID-19大流行期间）两个时段，分析45个非洲国家的输入性疟疾病例与中国31个省份的恶性疟原虫感染病例的时间趋势关联。基于上述分析，本研究提出一种统计模型与深度学习相结合的混合建模方案，利用2004-2016年的月度恶性疟原虫感染数据，对中国境内疟疾反弹态势进行建模。本研究构建了ARIMA-GRU混合模型，用于对中国各省的恶性疟原虫感染病例数，以及COVID-19大流行前后中国的疟疾死亡人数开展建模分析。结果：分析结果显示，非洲输入性疟疾病例的增长与中国多省份的恶性疟原虫感染病例增长之间存在逐渐显现的关联；且非洲输入性疟疾病例中多数为恶性疟原虫感染病例。本研究提出的深度学习模型在测试集上取得了96%的高预测准确率。结论：本研究分析了COVID-19大流行期间，中国因实施国际旅行限制防控措施，使得输入性恶性疟原虫感染病例及相关死亡人数下降的情况；并指出中国政府需在国际旅行常态化后制定输入性疟疾防控方案，以防范中国境内疟疾疫情反弹。