Institute for Digital Health and Innovation, University of Arkansas for Medical Sciences, Little Rock, Arkansas, USA.
Department of Biomedical Informatics, University of Arkansas for Medical Sciences, Little Rock, Arkansas, USA.
Popul Health Manag. 2022 Dec;25(6):798-806. doi: 10.1089/pop.2022.0189. Epub 2022 Nov 25.
This study evaluated relationships between county-level social vulnerability and broadband access using spatial clustering and regression approaches. County-level broadband availability (Federal Communications Commission [FCC] and Microsoft; 2019-2020), social vulnerability (COVID-19 Community Vulnerability Index [CCVI]; 2020), and primary care access (Area Health Resource File; 2019-2020) data sets were used. Two measures of broadband availability were considered: (1) Microsoft system-reported proportion of county population with broadband and (2) difference in FCC-reported and Microsoft-reported proportions of county population with broadband. Cluster maps were constructed using local Moran's I, and spatial Durbin models were estimated using primary care shortage designation and CCVI themes (socioeconomic status, minority status, housing/transportation/disability, epidemiological risk, health care system, high-risk environment, and population density). Among 3102 counties, county-level broadband coverage varied widely between Microsoft (0.39) and FCC (0.84), with greater coverage in the East and West, and larger discrepancies between FCC and Microsoft data in the South and Appalachia. In spatial regressions, a one-point increase in socioeconomic status vulnerability (0-least; 10-most vulnerable), was associated with a 2.0 percentage point (pp) reduction in broadband access ( < 0.001). Similar inverse relationships were observed with housing, epidemiological, and health care system variables. There were greater divergences between FCC and Microsoft measures with each one-point increase in socioeconomic status (1.4 pp), epidemiological risk (0.6 pp), and health care system (0.7 pp) vulnerability. More vulnerable counties had lower broadband and larger divergences between FCC and Microsoft data. Broadband is necessary for utilizing telehealth services; careful considerations in measuring broadband access can facilitate policies that improve equitable access to care.
本研究采用空间聚类和回归方法评估了县级社会脆弱性与宽带接入之间的关系。使用了县级宽带可用性（联邦通信委员会[FCC]和微软；2019-2020 年）、社会脆弱性（COVID-19 社区脆弱性指数[CCVI]；2020 年）和初级保健服务可及性（区域卫生资源文件；2019-2020 年）数据集。考虑了两种宽带可用性衡量标准：（1）微软系统报告的县人口中拥有宽带的比例，以及（2）联邦通信委员会报告的和县人口中拥有宽带的比例与微软报告的比例之间的差异。使用局部 Moran's I 构建聚类图，并使用初级保健短缺指定和 CCVI 主题（社会经济地位、少数族裔地位、住房/交通/残疾、流行病学风险、医疗保健系统、高危环境和人口密度）估计空间 Durbin 模型。在 3102 个县中，微软的县一级宽带覆盖率（0.39）和联邦通信委员会的覆盖率（0.84）差异很大，东部和西部的覆盖率较高，而南部和阿巴拉契亚地区联邦通信委员会和微软数据之间的差异较大。在空间回归中，社会经济地位脆弱性增加一个点（0-最脆弱；10-最脆弱），与宽带接入减少 2.0 个百分点（ < 0.001）相关。与住房、流行病学和医疗保健系统变量观察到类似的反向关系。随着社会经济地位（1.4 个点）、流行病学风险（0.6 个点）和医疗保健系统（0.7 个点）脆弱性每增加一个点，联邦通信委员会和微软措施之间的差异更大。脆弱性更高的县宽带接入率更低，联邦通信委员会和微软数据之间的差异也更大。宽带是利用远程医疗服务的必要条件；在衡量宽带接入时谨慎考虑可以促进制定改善公平获得医疗服务的政策。
