External quality assessment of microbial counts from water: To score or not to score for proficiency

Scoring schemes were first introduced in water-chemistry inter-laboratory p roficiency tests. Such scoring does not behave in a comparable manner when applied to microbiology external quality assessment. Conventional scores ca n become a lottery. This poses a challenge to (a) find helpful statistical methods which will look beyond this natural variation and (b) attempt to id entify any poor performance by the participants. The aim is to provide usef ul management tools. In the Public Health Laboratory Service, water external quality assessment scheme participants have detected methodology problems. Formal analyses are carried out on cumulated results in two stages, first demonstrating whethe r there is any systematic difference between laboratories of microbiologica l as well as statistical significance and only then allocating participants to categories of acceptable, or poor, performance. The scheme has been ope rating for several years and performance improvements are evident.