[Optimization of sample splitting in a centralized clinical chemical institute (author's transl)].

Mechanized sample splitting machines controlled by a laboratory data processing system have been realized in only a few centralized laboratories. Bottlenecks and mistakes in sample processing are avoided by means of direct machine readable identification and parallel splitting of the secondary tubes. Our previous experience has shown that a strategy for sample splitting has to go far beyond these basic functional requirements. The splitting process must be suited to the organization of the particular laboratory, and it must be adjusted to deal with problems of individual samples containing analytically interfering substances, or variable splitting may be required in cases of inadequate sample volumes. In addition to sample identification, the secondary tube has to be coded with the date and the type of material. This allows cumulative on-line processing and series of analyses of different materials. A suitable positional arrangement of containers for control material must be born in mind for quality control performances. We have realized these additional requirements by means of a consequent mutual adaptation in the layout of the request form (marking of priority and additional information), the file structure of the data processing system and the control program of the sample splitting machine.