Interhospital spread of multiply resistant klebsiella.

Strains studied were from (1) an outbreak of infection at St Thomas's Hospital in 19771; (2) an outbreak at the Luton and Dunstable Hospital in 1979, affecting at least 48 patients; (3) patients at Hammersmith Hospital in 1980, where a cluster of infections caused by the strain was found during a larger outbreak caused by a different multiply resistant klebsiella. Most isolates were from urine, and two were from blood cultures. No unusual virulence or invasiveness was suggested by the clinical histories. Identification of strains, sensitivity testing, and methods to characterise plasmids were as described.2 All the isolates of serotype K16 (table) were identical in biochemical reactions, and all but two carried plasmids of molecular weights 90 and 65 megadaltons (Md). The plasmid of 90 Md carried no antibiotic resistance and was not transferred. The plasmid of 65 Md was transmissible to Escherichia coli K12, belonged to incompatibility group M (IncM), and determined resistance to ampicillin/carbenicillin/ cephaloridine, tetracycline, chloramphenicol, sulphonamides, mercuric chloride, neomycin/kanamycin, gentamicin, streptomycin/spectinomycin, and trimethoprim. Its trimethoprim resistance was transposable from one genetic locus to another. The same transposon has been found in other plasmids and bacteria.3 Plasmids of the different K16 isolates were indistinguishable in molecular weights, resistance linkages, incompatibility grouping, endonuclease cleavage, and transposable DNA sequence. The 1979 K16 isolates, from Luton, carried a third plasmid, of 40 Md. Its resistances were masked by the 65 Md, IncM plasmid but it was isolated from two gentamicin-sensitive strains that lacked that plasmid. It was IncX, determined resistance to ampicillin, sulphonamides, streptomycin/spectinomycin, and trimethoprim, and had the same transposon as the IncM plasmid, suggesting that transposition had occurred. On the evidence of serotype, biotype, and the carriage of the 90 Md and 65 Md IncM plasmids we conclude that the same clone of bacteria was responsible for infections in different hospitals. The isolates from Luton had acquired an extra plasmid, whose absence from the 1980 Hammersmith isolates shows that the route of transmission of the strain was not from Luton to Hammersmith. Four of the K16 strains examined were resistant to nalidixic acid, at least two of them being from patients treated with that drug. As this character can be acquired by a single-step mutation, it does not alter our conclusions on the relation of the strains. Plasmid transfer between bacteria occurred during the 1977 outbreak.1' In the 1979 outbreak discussed here, klebsiella strains of serotypes K20, K61, and K62 were found with the same IncM plasmid as the K16 strains, evidence for resistance transfer between bacteria in the hospital.