SUMP BAY FEVER - INHALATIONAL FEVER ASSOCIATED WITH A BIOLOGICALLY CONTAMINATED WATER AEROSOL

Objective-To investigate the clinical, serological, and environmental features of a work related inhalational fever associated with exposure to an aerosol generated from a biologically contaminated 130 000 gall on water pool in a building used for testing scientific equipment. Met hod-Cross sectional survey of all exposed subjects (n = 83) by symptom questionnaire, clinical examination, spirometry, and serology for ant ibody to Pseudomonads, pool water extract, and endotoxin. In symptomat ic patients diffusion capacity was measured, and chest radiology was p erformed if this was abnormal. Serial peak flow was recorded in those subjects with wheeze. Bacterial and fungal air sampling was performed before and during operation of the water pool pump mechanism. Endotoxi n was measured in the trapped waters and in the pumps. Serum cotinine was measured as an objective indicator of smoking. Results-Of the 20 s ymptomatic subjects, fever was most common in those with the highest e xposure (chi(2) 42 . 7, P < 0 . 001) in the sump bay when the water wa s (torrentially) recirculated by the water pumps. Symptoms occurred la te in the working day only on days when the water pumps were used, and were independent of the serum cotinine. Pulmonary function was normal in most subjects (spirometry was normal in 79/83, diffusion capacity was low in five subjects, chest radiology was normal). Peak flow recor ding did not a work relation. The bacterial of the aerosol rose from 6 to > 10 000 colony forming units per cubic metre (cfu/m(3)) (predomin antly environmental Pseudomonads) when the pumps were operating. High endotoxin concentrations were measured in the waters and oil sumps in the pumps. Low concentrations of antibody to the organisms isolated we re detected (apart from two subjects with high antibody) but there was no relation to exposure or the presence of symptoms and similar antib ody was found in the serum samples from a nonexposed population. The f ever symptoms settled completely with the simple expedient of changing the water and cleaning the pumps. Conclusion-Given the results of our study, the development of inhalational fever in this unique environme nt and clearly restricted cohort was closely related to the degree of exposure to contaminated aerosol and mainly occurred in the absence of distinct serological abnormality and independent of cigarette smoking .