[Reported cases of Legionella pneumonia in the Netherlands, 1987-2000]

To describe the incidence and mortality of legionnaires' disease (LD) in the Netherlands on the basis of registration, with a focus on possible sources of infection, and geographic and seasonal variations.

Retrospective, descriptive.

The incidence and mortality rate of LD from July 1987 until December 2000 was described using registration data from the Dutch Health Inspectorate. Possible sources of infection were also listed (patients involved in the Bovenkarspel outbreak (1999) were excluded). Geographic variations in occurrence of LD were investigated by incidences computed per province. To gain insight into the possible existence of seasonal variations the mean numbers of patients per month were compared.

During the period of interest a total of 806 persons with LD were reported, 36 of whom were excluded. The mean incidence rose from 0.27 cases per 100,000 inhabitants in the 1987-1998 period to 0.91 cases per 100,000 inhabitants in the 1999-2000 period. The male:female ratio was 2.6:1. A total of 96 persons (12.5%) died as a result of LD. There were more deaths in the group of patients with confirmed LD compared to the group of patients with probable LD. After 1998 the case fatality rate was lower than in the previous period (Fisher's exact test: p = 0.005). The potential source of infection was situated in the open population in 92.9% of cases. In 7.1% of cases the potential source was situated in a hospital. The incidence per province varied from 0.13 per 100,000 (Drente) to 0.66 per 100,000 (Limburg). During the summer months more patients were reported, most of whom had contracted the infection abroad. In contrast to the 1987-1998 period, after 1998 the incidence in the Netherlands was higher than the European mean and the incidence of confirmed LD was higher than in the United States.