目的 通过星点设计 - 效应面法优化替硝唑结肠靶向生物黏附片处方。 方法 以海藻酸钠用量 (<>X₁) 、羟丙基甲基纤维素用量（ <> X ₂ ）及包衣增重（ <> X ₃ ）为考察因素，以 3 和 12 h 的累积释放度为考察指标 , 分别用多元线性模型、二次多项式模型描述考察指标和 3 个考察因素之间的数学关系 , 根据模型绘制效应面和等高线图 , 通过重叠等高线图确定优化处方 , 最后进行验证。 结果 根据二次多项式模型 , 发现 3 个考察因素和 2 个考察指标之间存在可信的定量关系；优化处方各设定指标的预测值和测定值非常接近；二次多项式模型比多元线性模型置信度高。 结论 采用星点设计 - 效应面法 , 得到了基于二次多项式模型的替硝唑结肠靶向生物黏附片处方优化模型 , 实现了该结肠靶向生物黏附片的处方优化。

OBJECTIVE To optimize the formulation of tinidazole colon-targeted bioadhesive tablets by the central composite design-response surface methodology (RSM plus CCD) . METHODS In the formulation design using RSM plus CCD , independent variables were the amounts of Alginic Acid and HPMC, and weight gain. The percentages of in vitro cumulative releases at 3 and 12 h were dependent variables. Multilinear and quadratic models were used to estimate the relationship between the dependent and the independent variables, and to delineate RSM and overlay contour plots in order to select the optimal formulations with the application of hypothesized in vitro cumulative releases (%) at 2 and 24 h. RESULTS The quantitative relationships between three factors and two evaluation indexes were characterized. In vitro release test of one selected optimal formulation indicated that were high approximation between the observed and estimated values. Moreover, quadratic model showed better prediction capability than multilinear model. CONCLUSION Quadratic model was preferred in the optimization of formulation due to the statistical confidence. The multi-objective simultaneous optimization of tinidazole colon-targeted bioadhesive tablets formulation could be achieved by the central composite design and response surface methodology.