Abstract：
In a recent monitoring study of Minnesota\'s public supply wells, Cryptosporidium was commonly detected with 40% of the wells having at least one detection. Risk factors for Cryptosporidium occurrence in drinking water supply wells, beyond surface water influence, remain poorly understood. To address this gap, physical and chemical factors were assessed as potential predictors of Cryptosporidium occurrence in 135 public supply wells in Minnesota. Univariable analysis, regression techniques, and classification trees were used to analyze the data. Many variables were identified as significant risk factors in univariable analysis and several remained significant throughout the succeeding analysis techniques. These factors fell into general categories of well use and construction, aquifer characteristics, and connectedness to the land surface, well capture zones, and land use therein, existence of potential contaminant sources within 200-feet of the well, and variability in the chemical and isotopic parameters measured during the study. These risk categories, and the specific variables and threshold values we have identified, can help guide future research on factors influencing Cryptosporidium contamination of wells and can be used by environmental health programs to develop risk-based sampling plans and design interventions that reduce associated health risks.
摘要：
在最近对明尼苏达州公共供应井的监测研究中，通常检测到隐孢子虫，其中40％的孔至少有一次检测。饮用水供水井中隐孢子虫发生的危险因素，超越地表水的影响，仍然知之甚少。为了解决这个差距，在明尼苏达州的135个公共供应井中，物理和化学因素被评估为隐孢子虫发生的潜在预测因子。单变量分析,回归技术，并使用分类树来分析数据。在单变量分析中，许多变量被确定为重要的风险因素，并且在随后的分析技术中，有几个变量仍然很重要。这些因素属于油井使用和建设的一般类别，含水层特征，和陆地表面的连通性，油井捕获区，以及其中的土地用途，在井的200英尺内存在潜在的污染源，以及研究过程中测量的化学和同位素参数的可变性。这些风险类别，以及我们确定的特定变量和阈值，可以帮助指导有关影响井隐孢子虫污染的因素的未来研究，并且可以被环境卫生计划用于制定基于风险的采样计划和设计减少相关健康风险的干预措施。
