Legionella infection has been confirmed in 13 people (Legionella pneumophila serogroup 1 was isolated from five), investigations are continuing for three, and one of the cases who died had shown clinical symptoms suggestive of legionellosis but laboratory confirmation could not be obtained. Three of the four cases who died had a bacteriologically confirmed diagnosis of legionellosis.

The epidemic curve shows that one case occurred at the beginning of September, and the remaining cases during the three weeks from 2 to 23 October (figure 1). All but the first case had stayed at the hotel between 23 September and 16 October (figure 2). To date, no cases who stayed at the hotel after 16 October (the date when the hotel was identified as the probable source of the infection) have been identified.

Figure 1: Cases of respiratory symptoms among people who stayed at the hotel Festival by week of onset of clinical signs

Figure 2: Cases of respiratory symptoms by week of stay at the hotel Festival, Istanbul, Turkey

According to a tour operator, the hotel could have accommodated 3300 people between the 1 September and 22 October 1997, an average of 470 people each week. The 12 cases confirmed during a three week exposure period at the hotel imply an attack rate in the order of 1% (the attack rate in legionellosis epidemics is usually between 0.1% and 5% [2]).

Nine people who stayed at the hotel in September or October have reported nonspecific symptoms and are being followed up. The exact number of cases associated with this outbreak cannot be known precisely yet. Results of serological tests will be known within two weeks and the possibility that these will be positive and that further cases may come to light cannot be excluded.

This report is also available in French at the Réseau National de Santé Publique (http://www.b3e.jussieu.fr:80/rnsp/).