【摘要】 目的 研究探讨应用线性规划(linear regression)趋势拟合模型，预测分析承德市献血人群血清学检测量变化和效果评价。方法 根据2011年1月至2021年12月承德市中心血站献血人群常规血清学检测与HIV阳性(初筛与确证)样本量三类检测数据统计资料，用趋势拟合法分别对三类检测数据，建立对时间的线性回归模型，对2022年三类检测数据进行预测，参数估计方法为最小二乘估计。结果 线性拟合经检验模型精度为优，拟合效果良好，预测变化趋势和准确值。2022年常规血清学检测预测值为44 251人份，HIV初筛反应性数预测值为81人份，HIV确证阳性数预测值为12人份。结论 承德市中心血站献血者常规血清学检测与HIV初筛反应性样本量二类检测数据有逐年持续上升趋势，HIV确证阳性样本数量变化不明显(或无显著性变化)。线性规划(linear regression)趋势拟合模型可以作为预测并强化血站实验室血液样本检测量科学管理的有效方法和可行性工具。更多还原

【Abstract】 Objective To predict and analyze the change in serological detection and effect evaluation on blood donors in Chengde by exploring the application of linear regression trend fitting prediction model. Methods From January 2011 to December 2021, the three testing data which included routine serological sample, HIV-positive initial reactivity sample and HIV-positive confirmed sample of blood donors in Chengde Central Blood Station were collected. A trend fitting prediction model was applied to set up a linear regression model of time-series of these data, and the three types of detection data in 2022 were predicted with the parameter estimation method as the least squares estimation.Results The linear regression model exhibited the excellent accuracy and good fitting effect, which can be used to describe changing trends and accurate values. The 2022-year predictive values were 44,251, 81, 12 in routine serological tests, HIV-positive initial reactivity number, HIV-positive confirmed number, respectively. Conclusion The routine serological tests, and HIV-positive initial reactivity number(second-class test data) are increasing year by year, but HIV-positive confirmed number is not significant(or less change). The linear regression trend fitting prediction model is eligible to be used as an effective method and feasibility tool for strengthening the scientific management of laboratory blood sample in the blood station.更多还原