School of Social Work and School of Public Health, Colorado State University, Campus Delivery 1586, Fort Collins, CO, United States.
School of Social Work and School of Public Health, Colorado State University, Fort Collins, CO, United States.
Child Abuse Negl. 2018 Dec;86:358-367. doi: 10.1016/j.chiabu.2018.08.010. Epub 2018 Aug 27.
This national study of US counties (n = 2963) investigated whether county-level drug overdose mortality is associated with maltreatment report rates, and whether the relationship between overdose mortality and maltreatment reports is moderated by a county's rural, non-metro or metro status. Data included county-level 2015 maltreatment reports from the National Child Abuse and Neglect Data System, modeled drug-overdose mortality from the Centers for Disease Control, United States Department of Agriculture Rural-Urban Continuum Codes, US Census demographic data and crime reports from the Federal Bureau of Investigation. All data were linked across counties. Zero-inflated negative binomial (ZINB) regression was used for county-level analysis. As hypothesized, results from the ZINB model showed a significant and positive relationship between drug overdose mortality and child maltreatment report rates (χ = 101.26, p < .0001). This relationship was moderated by position on the rural-urban continuum (χ=8.76, p = .01). For metro counties, there was a 1.9% increase in maltreatment report rate for each additional increment of overdose deaths (IRR=1.019, CI=[1.010, 1.028]). For non-metro counties, the rate of increase was 1.8% higher than for metro counties (IRR=1.018, CI=[1.006, 1.030]); for rural counties, the rate of increase was 1.2% higher than for metro counties (IRR=1.012, CI=[0.999, 1.026]). Additional research is needed to determine why the relationship between drug overdose mortality and maltreatment reports is stronger in non-metro and rural communities. One potential driver requiring additional inquiry is that access to mental and physical health care and substance use treatment may be more limited outside of metropolitan counties.
这项针对美国各县的全国性研究（n=2963）调查了县一级药物过量死亡率是否与虐待报告率有关，以及县的农村、非大都市或大都市地位是否会调节过量死亡率与虐待报告之间的关系。数据包括来自国家儿童虐待和忽视数据系统的 2015 年县一级虐待报告、疾病控制中心的药物过量死亡率模型、美国农业部农村-城市连续体代码、美国人口普查的人口统计数据以及联邦调查局的犯罪报告。所有数据都在县一级进行了链接。零膨胀负二项式（ZINB）回归用于县一级的分析。正如假设的那样，ZINB 模型的结果显示，药物过量死亡率与儿童虐待报告率之间存在显著的正相关关系（χ=101.26，p<0.0001）。这种关系受到农村-城市连续体位置的调节（χ=8.76，p=0.01）。对于大都市县，每增加一个额外的过量死亡，虐待报告率就会增加 1.9%（IRR=1.019，CI=[1.010, 1.028]）。对于非大都市县，增长率比大都市县高 1.8%（IRR=1.018，CI=[1.006, 1.030]）；对于农村县，增长率比大都市县高 1.2%（IRR=1.012，CI=[0.999, 1.026]）。需要进一步研究以确定为什么药物过量死亡率与虐待报告之间的关系在非大都市和农村社区更强。一个需要进一步调查的潜在驱动因素是，大都市县以外的心理健康和身体保健以及药物使用治疗的获取可能更有限。
