Chemical measurement laboratories that provide a broad range of services rely on efficient and rigorously validated measurement procedures to meet industry demands. The success of these laboratories ultimately hinges upon their ability to produce results that are both fit-for-purpose and affordable. Skillful planning of measurement experiments must therefore provide adequate confidence in results without being overly complex. A statistical procedure is presented that can be used by laboratories to optimize such experiment designs. Inputs to the statistical model are based upon simple criteria such as the number of available samples, expected variabilty of calibration data, and the largest acceptable uncertainty in the result. The outputs of the model guide a measurement scheme that can provide results with as small an uncertainty as possible under specified constraints, such as the desired range of calibration and the number of samples and repeated measurements that should be made. The method and software used to execute this model are demonstrated for planning evaluations of vitamin D in serum that inform clinical medicine diagnostics.

提供广泛服务的化学计量实验室依靠高效且经过严格验证的计量程序来满足行业需求。这些实验室的成功最终取决于其产生既适合目的又可负担得起的结果的能力。因此，熟练地计划测量实验必须对结果提供足够的信心，而又不要过于复杂。提出了一种统计程序，实验室可以使用该程序来优化此类实验设计。统计模型的输入基于简单的标准，例如可用样品的数量，校准数据的预期可变性以及结果中最大的可接受不确定性。模型的输出指导一种测量方案，该方案可以在指定的约束条件下提供尽可能小的不确定性结果，例如所需的校准范围，样品数量和应进行的重复测量。演示了用于执行此模型的方法和软件，可用于规划血清中维生素D的评估，从而为临床医学诊断提供依据。