Rev. sci. tech. Off. int. Epiz., 1994, 13 (3), 687-699  Use of combined Shewhart-CUSUM control  charts in internal quality control of  enzyme-linked immunosorbent assays for  the typing of foot and mouth disease  virus antigen  S.D. BLACKSELL *, L.J. GLEESON *, R.A. LUNT * and CHANPEN CHAMNANPOOD **  Summary: An enzyme-linked immunosorbent assay (ELISA) for the typing of foot and mouth disease virus (FMDV) antigen was employed for the routine laboratory diagnosis of FMD at a regional veterinary laboratory in northern Thailand. An objective procedure was developed to monitor the test performance of the ELISA, using absolute test control limits in a Shewhart- CUSUM (cumulative sum) control chart method. The procedure detected significant data trends and 'beyond control limit' situations for each antigen typing system (types O, A and Asia I), using an assay variable (yi). Retrospective analysis using Shewhart-CUSUM control charts of data from 42 ELIS As demonstrated that control limits were exceeded in two assays for FMDV type A. The Shewhart-CUSUM control chart is a simple and reliable internal quality control method for the detection of significant random and systematic variation in assays. KEYWORDS: ELISA - Foot and mouth disease - Quality control - Shewhart-CUSUM control charts.  INTRODUCTION  Detection and serotyping of foot and mouth disease virus (FMDV) antigen in field  samples can be performed rapidly and with high sensitivity in the laboratory using an  enzyme-linked immunosorbent assay (ELISA) (9). With minor modifications, this  ELISA has been used for routine FMD diagnosis in a variety of situations (5,11), and  may comprise up to seven separate antigen detection assays (for FMDV serotypes O, A, C,  Asia 1, SAT 1, SAT 2 and SAT 3) to be performed on a single test sample. The inherently  complex nature of the ELISA can lead to results containing random error (e.g. reagent  dilution error in a single test) or systematic variation (17) (e.g. gradual loss of assay  sensitivity caused by peroxidase lability), and therefore periodic standardisation of  reagents and internal quality control to monitor assay performance are desirable.  * Commonwealth Scientific and Industrial Research Organisation, Australian Animal Health  Laboratory, Institute of Animal Production and Processing, P.O. Bag 24, Geelong, Victoria 3220,  Australia.  ** Department of Livestock Development, Northern Veterinary Research and Diagnostic Centre,  Hang Chat, 52190 Lampang, Thailand.