Department of Gastroenterology and Hepatology, Weill Cornell Medicine, New York, NY, United States.
Division of Digestive and Liver Disease, Columbia University Medical Center, New York, NY, United States.
JMIR Public Health Surveill. 2020 Jul 17;6(3):e19354. doi: 10.2196/19354.
Coronavirus disease (COVID-19) is a novel viral illness that has rapidly spread worldwide. While the disease primarily presents as a respiratory illness, gastrointestinal symptoms such as diarrhea have been reported in up to one-third of confirmed cases, and patients may have mild symptoms that do not prompt them to seek medical attention. Internet-based infodemiology offers an approach to studying symptoms at a population level, even in individuals who do not seek medical care.
This study aimed to determine if a correlation exists between internet searches for gastrointestinal symptoms and the confirmed case count of COVID-19 in the United States.
The search terms chosen for analysis in this study included common gastrointestinal symptoms such as diarrhea, nausea, vomiting, and abdominal pain. Furthermore, the search terms fever and cough were used as positive controls, and constipation was used as a negative control. Daily query shares for the selected symptoms were obtained from Google Trends between October 1, 2019 and June 15, 2020 for all US states. These shares were divided into two time periods: pre-COVID-19 (prior to March 1) and post-COVID-19 (March 1-June 15). Confirmed COVID-19 case numbers were obtained from the Johns Hopkins University Center for Systems Science and Engineering data repository. Moving averages of the daily query shares (normalized to baseline pre-COVID-19) were then analyzed against the confirmed disease case count and daily new cases to establish a temporal relationship.
The relative search query shares of many symptoms, including nausea, vomiting, abdominal pain, and constipation, remained near or below baseline throughout the time period studied; however, there were notable increases in searches for the positive control symptoms of fever and cough as well as for diarrhea. These increases in daily search queries for fever, cough, and diarrhea preceded the rapid rise in number of cases by approximately 10 to 14 days. The search volumes for these terms began declining after mid-March despite the continued rises in cumulative cases and daily new case counts.
Google searches for symptoms may precede the actual rises in cases and hospitalizations during pandemics. During the current COVID-19 pandemic, this study demonstrates that internet search queries for fever, cough, and diarrhea increased prior to the increased confirmed case count by available testing during the early weeks of the pandemic in the United States. While the search volumes eventually decreased significantly as the number of cases continued to rise, internet query search data may still be a useful tool at a population level to identify areas of active disease transmission at the cusp of new outbreaks.
冠状病毒病（COVID-19）是一种新型病毒性疾病，已在全球迅速传播。尽管该疾病主要表现为呼吸道疾病，但多达三分之一的确诊病例报告了胃肠道症状，如腹泻，并且患者可能有轻微症状，不会促使他们寻求医疗。基于互联网的传染病学提供了一种在人群水平上研究症状的方法，即使是那些没有寻求医疗的人也是如此。
本研究旨在确定美国与胃肠道症状相关的互联网搜索与 COVID-19 确诊病例数之间是否存在相关性。
本研究选择了腹泻、恶心、呕吐和腹痛等常见胃肠道症状作为分析的搜索词。此外，发烧和咳嗽被用作阳性对照，便秘被用作阴性对照。从 2019 年 10 月 1 日至 2020 年 6 月 15 日，从 Google Trends 获得了美国所有州的选定症状的每日查询份额。这些份额分为两个时间段：COVID-19 之前（3 月 1 日之前）和 COVID-19 之后（3 月 1 日至 6 月 15 日）。从约翰霍普金斯大学系统科学与工程数据存储库获得了 COVID-19 确诊病例数。然后，将每日查询份额的移动平均值（归一化为 COVID-19 之前的基线）与确诊疾病病例数和每日新增病例进行分析，以建立时间关系。
许多症状的相对搜索查询份额，包括恶心、呕吐、腹痛和便秘，在整个研究期间仍接近或低于基线；然而，发烧和咳嗽以及腹泻等阳性对照症状的搜索量显著增加。这些每日发烧、咳嗽和腹泻搜索查询量的增加早于病例数的快速上升，大约提前 10 到 14 天。尽管累计病例数和每日新增病例数持续上升，但这些术语的搜索量在 3 月中旬后开始下降。
在大流行期间，症状的谷歌搜索可能早于病例和住院人数的实际上升。在当前的 COVID-19 大流行期间，本研究表明，在美国 COVID-19 大流行的早期，通过现有检测方法，发烧、咳嗽和腹泻的互联网搜索查询在确诊病例数增加之前就有所增加。尽管随着病例数的继续增加，搜索量最终显著下降，但互联网查询搜索数据仍然是人群水平上识别新爆发前疾病传播活跃地区的有用工具。
