Value of central analysis of leucocyte depletion quality control data within the National Blood Service, England

The results of quality monitoring leucocyte counts were analysed nationally for the first 2 years of universal leucocyte depletion (LD), spanning the time-period before and after standardization of the counting and LD methods. The objectives were twofold: first to determine whether the implementation strategy was effective in achieving the LD specification (< 5 x 10(6) leucocytes in 99% of components with 95% statistical confidence); and second, whether quality monitoring was able to detect potential non-conformance.

Residual leucocytes were counted using Beckman Coulter or Becton Dickinson (BD) flow cytometers and reagents. Data were collected into standardized analysis software (NWA Quality Analyst) for local trend analysis and checks for conformance to process specification, and collated centrally. Analysis was performed for six time-periods between January 1999 and March 2001. Specification failures were analysed to determine the likelihood of extreme failure. Statistical process monitoring was adjusted to suit LD processes.

Data from red cells in optimal additive solution (OAS), filtered either as whole blood or red cell concentrates, and platelet pools improved significantly over the 2-year period with specification failures falling from 0.35%, 0.48% and 0.56%, respectively, in January-June 1999 to 0.06%, 0.01% and 0.04% in January-March 2001. Specification failures in red cells in OAS LD for the period January-December 2000 showed only 0.02% with a leucocyte count of > 30 x 10(6)/unit. Extreme failures are now very rare. Monitoring methods have been effective in detecting process change and drift.

LD performance varies between different LD systems, but monitoring has proved sufficiently robust to detect processes that perform poorly. The chosen specification has been both achievable and appropriate to the systems in use. Standardization of the counting method is central to the ability to monitor and analyse results effectively across the whole service.