COMMENTS ON THE ROUTINE TESTING OF WATER SAMPLES FOR LEGIONELLA SP

Water samples analysed for Legionella sp. by a commercial company were studied over a 6 month period. Samples yielding legionella were re-ex amined to determine the efficiency of heat or acid pre-treatment in is olating legionella compared to untreated specimens. In addition, more than one colony of each colony type were tested serologically to deter mine whether serotypes in addition to those originally identified were present. Significantly (P < 0.01) higher legionella colony numbers we re found on the heat pre-treatment plates than on the acid pre-treatme nt plates. Legionellae were identified from only the untreated plate i n 5.1% of positive water samples; in these instances the concentration s of legionellae tended to be low. Six out of 18 positive samples were found to contain more than one serogroup of Legionella pneumophila wh en more than one colony of each colony type was tested. It is recommen ded that the procedures used in legionella screening be rationalised t o facilitate maximum detection without excessive workload.