The use of beads (microspheres) added into samples for absolute leucocyte enumeration by single platform (SP) flow cytometry is well established. International guidelines generally recommend at least duplicate or triplicate CD3 assays to ensure adequate quality control (QC) and reproducibility of the SP-derived cell count. Cost constraints of replicate testing, especially in a resource-poor setting, results in laboratories frequently only performing a single tube. The QC/reproducibility benefit is lost and the precision and accuracy relies heavily on the operator correctly pipetting the beads and blood. In April 2004, single-tube SP PLG CD4 testing (1) (using FlowCARE PLG CD4 reagents, the automated whole blood preparation system Immunoprep FlowCount beads, and sample analysis by XL-MCL flow cytometers, all from BCI, Miami, FL, USA) was implemented in 22 new district laboratories (without prior flow cytometry/pipetting experience) to support the South African National Anti-Retroviral (ARV) Treatment program. Because automated pipettes were not supplied to ARV sites, the introduction of QC system was necessary to monitor possible manual pipette error and ensure the reliability and reproducibility of single-assay CD4 results across the organisation. Although the stabilised fixed blood product (e.g., Immunotrol, BCI) was also implemented for once-daily QC of operating procedure, reagents, and pipetting across laboratories, this was not regarded as adequate to ensure correct pipetting of beads/blood in any given sample. The concept that the known quantities of beads added to samples can be used for internal QC of the pipetting step, by monitoring a bead count rate (BCR), was introduced by Bergeron et al. (2) and also described in recent Centers for Disease Control and Prevention guidelines for SP lymphocyte enumeration (2, 3). We implemented and further extended this concept by monitoring the BCR on sequentially prepared specimens and representing the sequential BCR in sequence plots. Examples shown in Figures 1 and 2 are taken from actual assessment of new South African National Health Laboratory Services (NHLS) ARV testing laboratories, where staff performing the testing were undergoing training and were essentially unfamiliar with the technique of flow cytometry and accurate reverse pipetting. The average BCR is also included on the plot, which is different for each instrument and dependent on the flow cytometric flow-rate setting (high, medium, or low). However, the expected 5% coefficient of variation (CV) limits can be set for all sites, irrespective of the instrument flow rate (data not shown but based on data accumulated from South African ARV sites). The average BCR can be used to easily calculate the standard deviation (limits shown on the plots) and identify any outliers (samples with incorrect pipetting). Sequence plot of BCR of 50 samples recorded in the sequence in which they were prepared at a new flow cytometry laboratory, site A. Patterns in the sequence can be noted. Curved lines indicate a wavelike trend (pattern) indicating that a stepper pipette was used in preparing the specimens. This site was warned of the inaccuracy associated with the use of stepper pipettes. The wavelike pattern of the BCR plot was a tangible measurement for the laboratory to visualise the poor pipetting performance. This initial pattern is followed by poorly reproducible BCR shown by several outliers in the circle. These are grouped together and not randomly distributed throughout the 50 specimens. This sudden change from the previous pattern was found to be due to a different person (less experienced with pipetting) preparing specimens from number 42 onwards. This change is easily detectable and can be used to monitor new staff members before they are allowed to perform routine duties. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.] Sequence plot of BCR of 50 samples recorded in the sequence in which they were prepared at a new laboratory, site B. This pattern shows wide scatter in the first 30 samples prepared by an inexperienced technologist. After training, the BCR variability is reduced as shown in the rectangular box of the remaining 20 samples showing improved performance. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.] An additional form of presenting the BCR separate from the sequence plot information, but which may also be useful for overall laboratory and instrument performance, is shown in Figure 3. The BCR is plotted against the time taken in seconds to analyse each specimen and includes the standard deviation limits used to identify outliers (calculated using the average BCR and expected percentage of CV). BCR plotted against time (in seconds) taken to analyse each specimen. The data for this plot are the same as those presented in Figure 2 and show the standard deviation (SD) limits calculated from the average BCR with an expected CV of 5%. Outliers are easily identified outside the SD limits and in this example show that most of the outliers from the test site have below-average BCR and comprise the same outliers noted on the sequence plot (samples analysed on “medium” instrument flow rate). Note that the reference site has a higher BCR (samples analysed on “high” instrument flow rate) with wider SD limits but was calculated using the same expected CV of 5%. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.] In conclusion, we have found the BCR sequence plots useful for monitoring training needs and assessing performance of staff in new NHLS ARV laboratories. Further, it has proved to be an invaluable tool for ensuring accuracy of CD4 counts in our setting and has been useful for monitoring longitudinal performance of NHLS ARV laboratories. By linking the BCR monitoring to our NHLS laboratory computer management system and by developing a system algorithm related to accessing previously authorised and approved CD4/BCR results, we have been able to automate BCR monitoring to ensure single-sample QC and monitor longitudinal laboratory performance, deriving daily, weekly, or monthly statistics. In practical terms, a CD4/BCR result that falls outside the 5% CV is rejected by the laboratory computer management system, thus highlighting possible pipetting error. An instruction is subsequently noted for the operator to re-prepare and re-analyse the sample before release of the final result. Reagent “shrinkage” from a business management perspective can also be monitored based on the numbers of “rejected” results. Moreover, BCR can aid in identifying faulty pipettes and further enables identifying faults in the flow cytometer. Lesley Erica Scott*, Deborah Kim Glencross , * Department of Molecular Medicine and Haematology, University of the Witwatersrand, Johannesburg, South Africa, National Health Laboratory Services, Johannesburg, South Africa.