Introduction to control charts in the analytical laboratory. Tutorial review

Abstract

The use of control charts in the analytical laboratory is introduced. Examples are provided to illustrate the use of control charts in assessing the performance of a measurement system; these are followed by a discussion of the assumptions and theory underlying the charts and of rules that are used to signal measurement problems. Next, some practical issues involved in setting up control charts are outlined. In particular, the problems arising in calculating an appropriate standard deviation, for establishing control limits for a chart, are considered: among these are the need to avoid partial replication and to include between-analyst, between-instrument and short-term temporal variation, where appropriate, in the estimate of random error. Aspects of the routine use of charts are then considered, including the recording of extra information about the assay on the control chart, so that it can be used for retrospective analyses of system performance and as a basis for assay improvement studies. Reference is made to the use of control charts as a kind of laboratory database, where the information required for planning future experimental studies is stored. Finally, brief reference is made to other types of control chart not discussed in the paper