Objective: To analyze the correlation between loss of smell/taste and the number of real confirmed cases of coronavirus disease 2019 (COVID-19) worldwide based on Google Trends data, and to explore the guiding role of smell/taste loss for the COVID-19 prevention and control. Methods: "Loss of smell" and "loss of taste" related keywords were searched in the Google Trends platform, the data were obtained from Jan. 1 2019 to Jul. 11 2021. The daily and newly confirmed COVID-19 case number were collected from World Health Organization (WHO) since Dec. 30 2019. All data were statistically analyzed by SPSS 23.0 software. The correlation was finally tested by Spearman correlation analysis. Results: A total of data from 80 weeks were collected. The retrospective analysis was performed on the new trend of COVID-19 confirmed cases in a total of 186 292 441 cases worldwide. Since the epidemic of COVID-19 was recorded on the WHO website, the relative searches related to loss of smell/taste in the Google Trends platform had been increasing globally. The global relative search volumes of "loss of smell" and "loss of taste" on Google Trends was 10.23±2.58 and 16.33±2.47 before the record of epidemic while 80.25±39.81 and 80.45±40.04 after (t value was 8.67, 14.43, respectively, both P<0.001). In the United States and India, the relative searches for "loss of smell" and "loss of taste" after the record of epidemic were also much higher than before (all P<0.001). The correlation coefficients between the trend of weekly new COVID-19 cases and the Google Trends of "loss of smell" in the global, United States, and India was 0.53, 0.76, and 0.82 respectively (all P<0.001), the correlation coefficients with Google Trends of "loss of taste" was 0.54, 0.78, and 0.82 respectively (all P<0.001). The lowest and highest point of loss of smell/taste search curves of Google Trends in different periods appeared 7 to 14 days earlier than that of the weekly newly COVID-19 confirmed cases curves, respectively. Conclusions: There is a significant positive correlation between the number of newly confirmed cases of COVID-19 worldwide and the amount of keywords, such as "loss of smell" and "loss of taste", retrieved in Google Trends. The trend of big data based on Google Trends might predict the outbreak trend of COVID-19 in advance.
目的： 基于互联网大数据，对嗅觉丧失、味觉丧失与全球新型冠状病毒肺炎（简称新冠肺炎）真实确诊例数变化趋势进行相关性分析，探究嗅觉味觉改变对新冠肺炎疫情防控的指导意义。 方法： 以“loss of smell”（嗅觉丧失）和“loss of taste”（味觉丧失）等相关词组为关键词，获取2019年1月1日至2021年7月11日间的谷歌趋势数据，并在世界卫生组织（WHO）网站获取自2019年12月30日以来的新冠肺炎确诊病例数据。所有数据均采用SPSS 23.0软件进行统计学分析。组间相关性检验最终采用Spearman相关性分析。 结果： 本研究采集了80周的数据，对累计186 292 441例全球新冠肺炎确诊病例的新增变化趋势进行了回顾性分析。自WHO网站有疫情记录以来，全球范围内关于嗅觉丧失和味觉丧失的谷歌相对检索量不断升高，嗅觉丧失和味觉丧失的相对检索量在有疫情记录前分别为10.23±2.58和16.33±2.47，在有疫情记录后分别为80.25±39.81和80.45±40.04（t值分别为8.67、14.43，P值均<0.001）；有疫情记录后美国和印度的嗅觉丧失、味觉丧失的谷歌相对检索量也远高于有疫情记录前（P值均<0.001）。全球、美国和印度周新增新冠肺炎确诊例数变化趋势与嗅觉丧失的谷歌趋势的相关系数分别为0.53、0.76和0.82（P值均<0.001），与味觉丧失的谷歌趋势的相关系数分别为0.54、0.78和0.82（P值均<0.001）。在不同时间段内，嗅觉丧失和味觉丧失搜索量曲线的最低点和最高点出现的时间均分别比每周新冠肺炎新增确诊例数曲线的最低点和最高点早7~14 d。 结论： 全球新冠肺炎新增确诊例数与谷歌趋势中嗅觉丧失、味觉丧失等相关关键词的检索量呈显著正相关。基于互联网大数据的嗅觉味觉丧失变化趋势可提前预测新冠肺炎疫情走势。.