概率单位平行线法在乙肝疫苗效力试验数据统计中的应用

采用概率单位平行线法对7个厂家7年间生产的血源乙型肝炎疫苗的效力试验结果进行统计分析，并与现行规程的Reed－Muench法计算结果比较，前法可进行线性、平行性检验及可信限率等分析，有效地区分试验误差，准确客观地评价疫苗效力。用此法时，剂量应不少于4组；各剂量反应率应落在经验区间内；EDS50可信限率应≤60％，相对效价可信限率应≤150％；可用合格、不合格或复试三个标准予以评价。本文对疫苗效力的稳定性及试验的一致性进行了质量分析，计得共同回归系数为1．5896、总体相对效价为2．31（2．20，2．42）。上述方法已用BASIC语言编制了程序，可用于常规检定。

Application of Probit Parallel Line Assay to Statistics of Potency Test of HB Vaccine

Data of potency test of HB vaccine manufactured by 7 manufacturers during 7 years were calculated by probit parallel line assay (PP) and compared with that calculated by current Reed-Muench method (RM). Linear and parallel examination and confidence-limit rate analysis could be performed by PP. It could also effectively distinguish experimental error and estimate vaccine efficacy exactly and objectively. The doses analysed by PP should not less than 4 groups, and each dose- response rate should be within the em-pirical interval.The confidence-limit rate of ED50 and of relative potency should be≤60% and ≤150%, respectively. The results could be evaluated as qualified', 'disqualified' and 'need reexamination'. The stability of vaccine efficacy and consistency of experiments were analysed. It has been established that common regression coefficient and total relative potency were 1. 5896and 2.31 (2. 20, 2. 42)respectively. The programme of PP was organized by BASIC for routine use.