| 摘要: |
| 为探究基于城市污水中新型冠状病毒（新冠病毒）浓度，采用线性拟合模型预测新冠病毒感染人数与医院报告感染人数的一致性，于2023年3月—2024年3月，在A、B两市污水处理厂设置采样点，每周采集1～2次水样，共采集478和882个水样，经聚乙二醇沉淀法富集浓缩后，用实时荧光反转录聚合酶链式反应定量检测，结合污水流量，计算新冠病毒周平均总拷贝数；以线性拟合模型预测A、B两市新冠病毒感染人数，采用一致性指数（d）、效率系数（E）、均方根误差（RMSE）和斯皮尔曼（Spearman）相关系数（r）进行一致性检验，评价模型预测精度。研究结果显示，A、B两市污水中新冠病毒检测阳性率分别为51.9%和86.3%；污水中新冠病毒周平均拷贝数与医院报告每日新增感染人数周平均值的总体变化趋势一致，且预测的新冠病毒感染人数与医院报告感染人数的波动变化一致；A、B两市预测值与报告值的d为0.854 0和0.853 4，E为0.566 6和0.588 0，RMSE为38.943 4和32.688 4，r为0.856 4和0.734 2，其中d、E、r一致性检验较好，而RMSE较差；A、B两市的预测值与报告值平均相差1.15和2.46倍。该方法对新冠病毒感染人数的预测结果与医院报告值相近，且在不同城市表现出相近的倍数关系，可为将来快速准确地预测新冠病毒感染人数提供预测框架。 |
| 关键词: 污水 新冠病毒 线性拟合模型 精度评价 一致性指数 效率系数 |
| DOI：DOI：10.3969/j.issn.1674-6732.2025.01.001 |
| 分类号:X832；R511 |
| 基金项目:江苏省卫健委医学科研重点项目（ZD2021021） |
|
| Inference of the Number of Community-level Infected Individuals Based on the Concentration of SARS-CoV-2 in Urban Wastewater |
|
WANG Rui1， PENG Shifu2， DING Zhen1，2*
|
|
1.School of Public Health， Nanjing Medical University， Nanjing， Jiangsu 211100， China； 2. Jiangsu Provincial Center for Disease Control and Prevention， Nanjing， Jiangsu 210003， China
|
| Abstract: |
| This study utilized a linear fitting model to predict the number of COVID-19 infections based on the concentration of SARS-CoV-2 in urban wastewater， and evaluated the consistency between the predicted values and hospital-reported infection numbers. From March 2023 to March 2024， sampling points were established at wastewater treatment plants in Cities A and B， with water samples collected 1~2 times weekly， totaling 478 and 882 samples， respectively. Polyethylene glycol precipitation was used for concentration， followed by quantitative detection using real-time fluorescent reverse transcription polymerase chain reaction（RT-qPCR）. In conjunction with wastewater flow rates， the weekly average total copy number of SARS-CoV-2 was calculated. The linear fitting models were then used to predict the number of COVID-19 infections in Cities A and B. Predictive accuracy was evaluated through consistency tests， including the consistency index（d）， efficiency coefficient（E）， root mean square error（RMSE）， and Spearman's rank correlation coefficient（r）. The research results showed that the positive detection rates of SARS-CoV-2 in wastewater from Cities A and B were 51.9% and 86.3%， respectively. The weekly average copy number of SARS-CoV-2 in wastewater showed a consistent overall trend with the weekly average of daily new infections reported by hospitals， and the predicted number of COVID-19 infections exhibited consistent fluctuation trends with the hospital-reported numbers. For Cities A and B， the d values between the predicted and reported values were 0.854 0 and 0.853 4，E values were 0.566 6 and 0.588 0， RMSE values were 38.943 4 and 32.688 4， and r values were 0.856 4 and 0.734 2， respectively. Among these，d，E， and r demonstrated good consistency， while RMSE indicated weaker performance. The predicted values were， on average， 1.15 times and 2.46 times the reported values in Cities A and B， respectively. The predictive results of this method for the number of COVID-19 infections were similar to those reported by hospitals and showed comparable multiplicative relationships across different cities， providing a predictive framework for rapid and accurate future predictions of COVID-19 infection numbers. |
| Key words: Wastewater SARS-CoV-2 Linear fitting model Accuracy assessment The index of agreement The coefficient of efficiency |
