Abstract

Sixty staff working in a hospital operating theatre, where a case of humidifier fever had been identified, were studied together with 49 subjects working in other parts of the hospital. They each had a blood test for serology, a skin test, and a chest radiograph and completed a questionnaire. The theatre staff also had pulmonary function tests. The theatre humidifier was found to contain several organisms including amoebae and antigens cross-reacting highly with those implicated in previous outbreaks of humidifier fever. Of the 60 exposed subjects, 25 had developed antibodies, nine had probable symptoms of humidifier fever, and six possible symptoms. There was a strong association between symptoms and antibodies (p = 6 x 10(-5) by Fisher's exact test). The development of antibodies was also related to duration of exposure in the theatre (p less than 0.01 by X2 test for trend) and inversely to smoking (p = 0.0073 by Fisher's exact test) but not to history of atopy. Because of the presence of antigens and because certain biocides added were ineffective in controlling antigenic build-up the humidifier was switched off. Eight months later specific IgG levels in the theatre staff sera, estimated by an enzyme-linked immunosorbent assay technique, had fallen on average by 25%. Inhalation challenge with humidifier water was performed in eight subjects. Four subjects reacted to the challenge, including both those with antibodies and previous symptoms.