Gavin Herbert Eye Institute, Department of Ophthalmology, University of California, Irvine School of Medicine, Irvine, CA.
Department of Computer Science, University of California, Irvine, Irvine, CA; and.
Cornea. 2021 Mar 1;40(3):282-291. doi: 10.1097/ICO.0000000000002579.
Our study fills the spatiotemporal gaps in dry eye disease (DED) epidemiology by using Google Trends as a novel epidemiological tool for geographically mapping DED in relation to environmental risk factors.
We used Google Trends to extract DED-related queries estimating users' intent from 2004 to 2019 in the United States. We incorporated national climate data to generate heat maps comparing geographic, temporal, and environmental relationships of DED. Multivariable regression models were constructed to generate quadratic forecasts predicting DED and control searches.
Our results illustrated the upward trend, seasonal pattern, environmental influence, and spatial relationship of DED search volume across the US geography. Localized patches of DED interest were visualized in urban areas. There was no significant difference in DED queries across the US census regions (P = 0.3543). Regression model 1 predicted DED queries per state (R2 = 0.61), with the significant predictor being urban population [r = 0.56, adjusted (adj.) P < 0.001, n = 50]; model 2 predicted DED searches over time (R2 = 0.97), with significant predictors being control queries (r = 0.85, adj. P = 0.0169, n = 190), time (r = 0.96, adj. P < 0.001, n = 190), time2 (r = 0.97, adj. P < 0.001, n = 190), and seasonality (winter r = -0.04, adj. P = 0.0196, n = 190; spring r = 0.10, adj. P < 0.001, n = 190).
Our study used Google Trends as a novel epidemiologic approach to geographically mapping the US DED. Importantly, urban population and seasonality were stronger risk factors of DED searches than temperature, humidity, sunshine, pollution, or region. Our work paves the way for future exploration of geographic information systems for locating DED and other diseases through online population metrics.
我们使用谷歌趋势作为一种新颖的流行病学工具，从空间和时间上填补了干眼症 (DED) 流行病学的空白，以研究 DED 与环境危险因素之间的关系。
我们使用谷歌趋势从 2004 年到 2019 年在美国提取了与 DED 相关的查询，以估计用户的意图。我们结合了全国气候数据，生成了热图，比较了 DED 的地理、时间和环境关系。构建了多变量回归模型，以生成预测 DED 和对照搜索的二次预测。
我们的结果说明了美国地理上 DED 搜索量的上升趋势、季节性模式、环境影响和空间关系。在城市地区，可视化了 DED 兴趣的局部斑块。在美国人口普查区之间，DED 查询没有显著差异（P = 0.3543）。回归模型 1 预测了每个州的 DED 查询量（R2 = 0.61），具有显著预测作用的是城市人口[ r = 0.56，调整（adj.）P < 0.001，n = 50]；模型 2 预测了随时间的 DED 搜索（R2 = 0.97），具有显著预测作用的是对照查询（r = 0.85，adj. P = 0.0169，n = 190）、时间（r = 0.96，adj. P < 0.001，n = 190）、时间 2（r = 0.97，adj. P < 0.001，n = 190）和季节性（冬季 r = -0.04，adj. P = 0.0196，n = 190；春季 r = 0.10，adj. P < 0.001，n = 190）。
我们的研究使用谷歌趋势作为一种新颖的流行病学方法，从空间上绘制了美国 DED 的地图。重要的是，城市人口和季节性是 DED 搜索的更强危险因素，而温度、湿度、阳光、污染或地区则不是。我们的工作为通过在线人口指标定位 DED 和其他疾病的地理信息系统的未来探索铺平了道路。
