Duffy S, Churey J, Worobo R W, Schaffner D W
Food Risk Analysis Initiative, Rutgers University, New Brunswick, New Jersey 08901, USA.
J Food Prot. 2000 Nov;63(11):1587-90. doi: 10.4315/0362-028x-63.11.1587.
Raw data from validation studies of UV tubes used for nonthermal pathogen reduction in apple cider underwent comprehensive statistical analysis. Data from each tube that demonstrated at least a 5-log reduction of Escherichia coli ATCC 25922, a surrogate for E. coli O157:H7, in each of three trials were used in the analysis. The within- and between-tube variability was calculated for 70 tubes. The mean log reductions of the tubes fit a Beta distribution (Kolmogorov-Smirnov test, 0.0246), and the between-replicate variability followed a logistic distribution (Kolmogorov-Smirnov test, 0.0305). These two distributions can be used together to model UV cider treatment as part of an overall E. coli O157:H7 in cider risk assessment. Examples of codes from @RISK and Analyticato describe these distributions, such as one would find in a quantitative risk assessment, are included.
对用于苹果酒非热减菌的紫外线灯管验证研究的原始数据进行了全面的统计分析。分析中使用了来自每个灯管的数据，这些灯管在三项试验的每一项中均显示出对大肠杆菌ATCC 25922（一种大肠杆菌O157:H7的替代菌）至少有5个对数级的减少。计算了70个灯管的管内和管间变异性。灯管的平均对数减少量符合β分布（柯尔莫哥洛夫-斯米尔诺夫检验，0.0246），重复间变异性遵循逻辑分布（柯尔莫哥洛夫-斯米尔诺夫检验，0.0305）。这两种分布可共同用于将紫外线处理苹果酒建模为苹果酒中大肠杆菌O157:H7总体风险评估的一部分。文中包含了来自@RISK和Analyticat的代码示例，用于描述这些分布，就如同在定量风险评估中会看到的那样。
