The effect of population mobility on COVID-19 incidence in 314 Latin American cities: a longitudinal ecological study with mobile phone location data.
The effect of population mobility on COVID-19 incidence in 314 Latin American cities: a longitudinal ecological study with mobile phone location data.
复制标题
DOI：
10.1016/s2589-7500(21)00174-6
复制
发表时间：
2021-11
期刊：
影响因子：
--
通讯作者：
Diez Roux AV
中科院分区：
文献类型：
--
作者：
Kephart JL;Delclòs-Alió X;Rodríguez DA;Sarmiento OL;Barrientos-Gutiérrez T;Ramirez-Zea M;Quistberg DA;Bilal U;Diez Roux AV
关于郊区流动性的变化对随后的新冠肺炎发病率的影响，人们知之甚少，这在拉丁美洲尤其相关，因为那里存在大量障碍，阻碍了获得新冠肺炎疫苗的机会，而非药物干预对缓解努力至关重要。我们的目标是在大量拉丁美洲城市的郊区水平上检验人口流动和新冠肺炎发病率之间的纵向联系。在这项纵向生态研究中，我们收集了汇总的手机位置数据、每日确诊的新冠肺炎病例以及城市和社会环境的特征，以分析阿根廷、巴西、哥伦比亚、危地马拉和墨西哥人口超过10万 000的城市在郊区层面的人口流动和新冠肺炎发病率，时间为2020年3月2日至8月29日。联合国开发计划署在拉丁美洲和加勒比地区和Grandata提供了空间聚合的手机数据；新冠肺炎确诊病例来自各国政府报告，人口和社会经济因素来自每个国家的最新全国人口普查。我们使用混合效应负二项回归进行时间序列分析，在调整城市环境和社会经济因素(不变的教育程度、居住过度拥挤、人口密度[均在郊区水平])之后，检验从基线(人口流行前一周，2020年3月2日至9日)到随后的郊区新冠肺炎发病率(滞后1-6周)之间的纵向关联。我们包括1031个郊区，代表314个拉丁美洲城市，分别位于阿根廷(107个郊区)、巴西(416个)、哥伦比亚(82个)、危地马拉(20个)和墨西哥(406个)。在主要调整的模型中，我们观察到前一周新冠肺炎每单位日志发生的发生率比(与基线相比)增加2.35(95%可信区间2.12-2.60)。因此，每周活动能力降低10%与接下来一周新冠肺炎发生率降低8.6%(95%可信区间7·6~9·6)相关。随着移动性和新冠肺炎发病率之间的滞后增加，这种关联逐渐减弱，并且在滞后6周时与零没有区别。郊区人口流动的减少与该郊区居民新冠肺炎发病率的降低相关。在郊区一级减少人口流动的政策可能是一种有效的新冠肺炎缓解战略，尽管它们应该与减轻流动性减少对最弱势群体造成的任何不利社会和经济后果的战略相结合。惠康信托公司。摘要的西班牙语译本见补充资料部分。
Little is known about the effect of changes in mobility at the subcity level on subsequent COVID-19 incidence, which is particularly relevant in Latin America, where substantial barriers prevent COVID-19 vaccine access and non-pharmaceutical interventions are essential to mitigation efforts. We aimed to examine the longitudinal associations between population mobility and COVID-19 incidence at the subcity level across a large number of Latin American cities. In this longitudinal ecological study, we compiled aggregated mobile phone location data, daily confirmed COVID-19 cases, and features of urban and social environments to analyse population mobility and COVID-19 incidence at the subcity level among cities with more than 100 000 inhabitants in Argentina, Brazil, Colombia, Guatemala, and Mexico, from March 2 to Aug 29, 2020. Spatially aggregated mobile phone data were provided by the UN Development Programme in Latin America and the Caribbean and Grandata; confirmed COVID-19 cases were from national government reports and population and socioeconomic factors were from the latest national census in each country. We used mixed-effects negative binomial regression for a time-series analysis, to examine longitudinal associations between weekly mobility changes from baseline (prepandemic week of March 2–9, 2020) and subsequent COVID-19 incidence (lagged by 1–6 weeks) at the subcity level, adjusting for urban environmental and socioeconomic factors (time-invariant educational attainment, residential overcrowding, population density [all at the subcity level], and country). We included 1031 subcity areas, representing 314 Latin American cities, in Argentina (107 subcity areas), Brazil (416), Colombia (82), Guatemala (20), and Mexico (406). In the main adjusted model, we observed an incidence rate ratio (IRR) of 2·35 (95% CI 2·12–2·60) for COVID-19 incidence per log unit increase in the mobility ratio (vs baseline) during the previous week. Thus, 10% lower weekly mobility was associated with 8·6% (95% CI 7·6–9·6) lower incidence of COVID-19 in the following week. This association gradually weakened as the lag between mobility and COVID-19 incidence increased and was not different from null at a 6-week lag. Reduced population movement within a subcity area is associated with a subsequent decrease in COVID-19 incidence among residents of that subcity area. Policies that reduce population mobility at the subcity level might be an effective COVID-19 mitigation strategy, although they should be combined with strategies that mitigate any adverse social and economic consequences of reduced mobility for the most vulnerable groups. Wellcome Trust. For the Spanish translation of the abstract see Supplementary Materials section.