The major sources of infection for Legionnaires'
Disease, identified by study of outbreaks, are hot water
systems and cooling towers. However, most cases are not
part of outbreaks and, for these, the source of infection is
rarely traced. The principle aim of this study was to help
understand the source of non-outbreak infection by examining
the epidemiology of the disease in Scotland.
Of the. recognised cases which met the study casedefinition,
366 were ill between 1978 to 1986 giving a mean
annual incidence rate of 7.9 per million. The annual
incidence varied in Scotland (range 3.1 to 20.2) and within
health boards. Geographical variations were demonstrated by
health board, by city and within cities, particularly for
non-travel infection. For example, the cumulative incidence
rate per million for non-travel, non-outbreak disease in
Greater Glasgow Health Board (GGHB) was 130 compared to 45
for the whole of Scotland, and 11, 33 and 50 in Tayside,
Lanarkshire and Lothian Health Boards respectively. Of 16
postcode sectors with a high incidence of disease in
Scotland, 14 were in GGHB. In GGHB, the residence of nontravel,
non-outbreak cases (but not of travel-related ones)
was clustered in central areas. Previously unrecognised
clustering was also found in other health boards.
These variations were not fully explained by differences
in the population's exposure to diagnostic tests, as
indicated by the number of serology tests reguested by
Scottish hospitals; the diagnostic service and approach of
bacteriology laboratories; and the approach of hospital
consultants to the diagnosis of Legionnaires' Disease.
Differences in host susceptibilty, as reflected by
socio-economic status and the incidence of other respiratory
disease, were small and did not explain the variation.
In the City of Glasgow, many cooling towers were not
maintained in accord with recommendations and posed a
theoretical risk of infection. The location of residence of
non-travel cases was associated with the location of
premises with cooling towers, the incidence of non-travel
Legionnaires' Disease being more than three times higher in
areas of Glasgow within 0.5 kilometres of a cooling tower
than in areas more than one kilometre away.
The best explanation for these observations is that
cooling towers were a major source of non-travel, nonoutbreak
infection. Hence, for the investigation and
prevention of such infection, the emphasis should be on
cooling tower maintenance. Close surveillance of apparently
sporadic disease is recommended as the basis for disease
control and future research.