An outbreak of Legionella longbeachae infection in an intensive care unit?

During a nine-day period, five patients in a 14-bed intensive care unit (ICU) were shown to have seroconverted with a four-fold or greater rise in serum antibody titre to Legionella longbeachae serogroup 1. A further two patients were observed to have high titres consistent with previous exposure but earlier serum samples were not available for comparison. No patients had antibody responses to Legionella pneumophila serogroups 1 and 2. L. longbeachae was not cultured from respiratory secretions from patients or from the environment within the unit. Legionella anisa was recovered from one cooling tower on the ninth floor of the tower block. The ICU is located on the first floor of the same tower and receives external air from two vents, one on the eastern and the other on the western aspect. All patients with serological evidence of L. longbeachae infection were concomitantly infected with multiresistant Staphylococcus aureus, and were located in bays on the eastern side of the unit. A large pigeon nest was discovered within 1-2 m of the eastern vent. Following removal of the birds' nest, no further cases were seen on routine screening of all patients within the unit over the next eight weeks. Alternatively, seroconversion may have been related to demolition of the adjacent nine-storey nurses home. This was begun one month before the first case was diagnosed and was completed four months later. The periodic northerly winds could have carried legionellae from the demolition site directly over the block housing the ICU and may have concentrated them near the eastern air vent. All patients had pneumonia, which,vas probably multifactorial in origin. There is some uncertainty whether the serological responses, seen were an epiphenomenon or were truly indicative of infection with L. longbeachae. (C) 2002 The Hospital Infection Society.